diff --git a/checkpoint-19250/config.json b/checkpoint-19250/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fe7dbf7d28e62328ecc2df097527bd51ec161f9d --- /dev/null +++ b/checkpoint-19250/config.json @@ -0,0 +1,73 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 6, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 512, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 64, + "intermediate_size": 8960, + "layerwise_start_idx": 21, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 7, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f128c512-mix-1.0-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-19250/generation_config.json b/checkpoint-19250/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-19250/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-19250/model-00001-of-00002.safetensors b/checkpoint-19250/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5b1ba2aa473a2f23ba641584ae6b22e62a7b235f --- /dev/null +++ b/checkpoint-19250/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f31bd8000cbeff1fe129339d3ef980b64efda7f31feaf437d629ff4fdc78f8c8 +size 4754815448 diff --git a/checkpoint-19250/model-00002-of-00002.safetensors b/checkpoint-19250/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c76cf00e7a9c7ce6c6d8a7a347afcbf18b661e2d --- /dev/null +++ b/checkpoint-19250/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d096d5260908edb86a73ee04889ac23d52b4054c4adabbc4a0e6ec76854d6f +size 840447856 diff --git a/checkpoint-19250/model.safetensors.index.json b/checkpoint-19250/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea3b7f98260439a7c863a0290604b077271f337 --- /dev/null +++ b/checkpoint-19250/model.safetensors.index.json @@ -0,0 +1,1076 @@ +{ + "metadata": { + "total_size": 5595140920 + }, + "weight_map": { + "embed_tokens.weight": "model-00001-of-00002.safetensors", + "img2text.bias": "model-00001-of-00002.safetensors", + "img2text.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.weight": "model-00001-of-00002.safetensors", + "layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.gate": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.22.gate": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.23.gate": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.24.gate": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.25.gate": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.26.gate": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.27.gate": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "norm.weight": "model-00001-of-00002.safetensors", + "patch_embedder.proj.bias": "model-00001-of-00002.safetensors", + "patch_embedder.proj.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.bias": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_out.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.conv_out.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.encoder.conv_in.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_in.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-19250/optimizer.pt b/checkpoint-19250/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..d360d83c4cdd076040e585f1df275b9d0a88f37f --- /dev/null +++ b/checkpoint-19250/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7bbdd6b50d5f97d2cb9fdd8d0eb53a0a8f70998a2a59321d167ca3790d119476 +size 5620161007 diff --git a/checkpoint-19250/rng_state_0.pth b/checkpoint-19250/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..f786bd4996bee4d5db558682bd4ed3aaf23c11a1 --- /dev/null +++ b/checkpoint-19250/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1497bf043c91e1049c3d77e5a816a2ae7dfc0cb5ebea49f517ec5ba64782aad7 +size 15024 diff --git a/checkpoint-19250/rng_state_1.pth b/checkpoint-19250/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..985a2d523af9a8af2ecb9e70f5c3469838aed57d --- /dev/null +++ b/checkpoint-19250/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a1bd9b1fd78c3896f010719176a5cf50b834ea4b0e40930b7554f0bf36bceb9d +size 15024 diff --git a/checkpoint-19250/rng_state_2.pth b/checkpoint-19250/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..95aacd2bb0401f5ac53339f40a51902f60d75a3f --- /dev/null +++ b/checkpoint-19250/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd2495d973bf2f172c7bacf5678fa2df86435591f77f4abaf56a935abc6f181f +size 15024 diff --git a/checkpoint-19250/rng_state_3.pth b/checkpoint-19250/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..0473b9ffeb1001c4633c3ae2381e49bf77c83bad --- /dev/null +++ b/checkpoint-19250/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3905d346f4fabeb4ae8896b4cc7e5f94f9617e1c9918b5517febcd98d16f224 +size 15024 diff --git a/checkpoint-19250/scheduler.pt b/checkpoint-19250/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..ad349e18a5f66c8b52b9d47a0261b67f6f407ed1 --- /dev/null +++ b/checkpoint-19250/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f05e855310a7bfa40316df7452e3307b5608794551da30f7f731560bd86b8ce +size 1064 diff --git a/checkpoint-19250/trainer_state.json b/checkpoint-19250/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3cb0f601f6776b5466a8404bf0e72e29041e484e --- /dev/null +++ b/checkpoint-19250/trainer_state.json @@ -0,0 +1,134783 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.690771074710348, + "eval_steps": 500, + "global_step": 19250, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00039952057530962844, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 3.1824, + "step": 1 + }, + { + "epoch": 0.0007990411506192569, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.9778, + "step": 2 + }, + { + "epoch": 0.0011985617259288853, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 2.7824, + "step": 3 + }, + { + "epoch": 0.0015980823012385138, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 2.6074, + "step": 4 + }, + { + "epoch": 0.001997602876548142, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.4209, + "step": 5 + }, + { + "epoch": 0.0023971234518577705, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 2.2482, + "step": 6 + }, + { + "epoch": 0.002796644027167399, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 2.1017, + "step": 7 + }, + { + "epoch": 0.0031961646024770275, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 2.0034, + "step": 8 + }, + { + "epoch": 0.003595685177786656, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.9567, + "step": 9 + }, + { + "epoch": 0.003995205753096284, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.9417, + "step": 10 + }, + { + "epoch": 0.0043947263284059125, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.9526, + "step": 11 + }, + { + "epoch": 0.004794246903715541, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.9613, + "step": 12 + }, + { + "epoch": 0.0051937674790251695, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.9467, + "step": 13 + }, + { + "epoch": 0.005593288054334798, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.9215, + "step": 14 + }, + { + "epoch": 0.0059928086296444265, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.9154, + "step": 15 + }, + { + "epoch": 0.006392329204954055, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.9064, + "step": 16 + }, + { + "epoch": 0.0067918497802636835, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.888, + "step": 17 + }, + { + "epoch": 0.007191370355573312, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.8729, + "step": 18 + }, + { + "epoch": 0.0075908909308829405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8425, + "step": 19 + }, + { + "epoch": 0.007990411506192568, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.8468, + "step": 20 + }, + { + "epoch": 0.008389932081502197, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.8418, + "step": 21 + }, + { + "epoch": 0.008789452656811825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.825, + "step": 22 + }, + { + "epoch": 0.009188973232121454, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8176, + "step": 23 + }, + { + "epoch": 0.009588493807431082, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.7827, + "step": 24 + }, + { + "epoch": 0.00998801438274071, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.7695, + "step": 25 + }, + { + "epoch": 0.010387534958050339, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.78, + "step": 26 + }, + { + "epoch": 0.010787055533359968, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.7588, + "step": 27 + }, + { + "epoch": 0.011186576108669596, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.743, + "step": 28 + }, + { + "epoch": 0.011586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.7357, + "step": 29 + }, + { + "epoch": 0.011985617259288853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.7271, + "step": 30 + }, + { + "epoch": 0.012385137834598482, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.7053, + "step": 31 + }, + { + "epoch": 0.01278465840990811, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 32 + }, + { + "epoch": 0.013184178985217739, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.6831, + "step": 33 + }, + { + "epoch": 0.013583699560527367, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 34 + }, + { + "epoch": 0.013983220135836995, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6682, + "step": 35 + }, + { + "epoch": 0.014382740711146624, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6573, + "step": 36 + }, + { + "epoch": 0.014782261286456252, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.6507, + "step": 37 + }, + { + "epoch": 0.015181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.6424, + "step": 38 + }, + { + "epoch": 0.01558130243707551, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.6251, + "step": 39 + }, + { + "epoch": 0.015980823012385136, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.6215, + "step": 40 + }, + { + "epoch": 0.016380343587694766, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.6226, + "step": 41 + }, + { + "epoch": 0.016779864163004393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.6005, + "step": 42 + }, + { + "epoch": 0.017179384738314023, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.6224, + "step": 43 + }, + { + "epoch": 0.01757890531362365, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5901, + "step": 44 + }, + { + "epoch": 0.01797842588893328, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5962, + "step": 45 + }, + { + "epoch": 0.018377946464242907, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 46 + }, + { + "epoch": 0.018777467039552537, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5697, + "step": 47 + }, + { + "epoch": 0.019176987614862164, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 48 + }, + { + "epoch": 0.019576508190171794, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.5531, + "step": 49 + }, + { + "epoch": 0.01997602876548142, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.554, + "step": 50 + }, + { + "epoch": 0.02037554934079105, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.5539, + "step": 51 + }, + { + "epoch": 0.020775069916100678, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.5591, + "step": 52 + }, + { + "epoch": 0.02117459049141031, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.545, + "step": 53 + }, + { + "epoch": 0.021574111066719935, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5221, + "step": 54 + }, + { + "epoch": 0.021973631642029565, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.5234, + "step": 55 + }, + { + "epoch": 0.022373152217339192, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.5186, + "step": 56 + }, + { + "epoch": 0.022772672792648822, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.5273, + "step": 57 + }, + { + "epoch": 0.02317219336795845, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5201, + "step": 58 + }, + { + "epoch": 0.02357171394326808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5108, + "step": 59 + }, + { + "epoch": 0.023971234518577706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.5078, + "step": 60 + }, + { + "epoch": 0.024370755093887336, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.5026, + "step": 61 + }, + { + "epoch": 0.024770275669196963, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.5076, + "step": 62 + }, + { + "epoch": 0.025169796244506593, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4956, + "step": 63 + }, + { + "epoch": 0.02556931681981622, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4925, + "step": 64 + }, + { + "epoch": 0.02596883739512585, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.5045, + "step": 65 + }, + { + "epoch": 0.026368357970435477, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.4843, + "step": 66 + }, + { + "epoch": 0.026767878545745107, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4807, + "step": 67 + }, + { + "epoch": 0.027167399121054734, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 68 + }, + { + "epoch": 0.027566919696364364, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4784, + "step": 69 + }, + { + "epoch": 0.02796644027167399, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4758, + "step": 70 + }, + { + "epoch": 0.02836596084698362, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4871, + "step": 71 + }, + { + "epoch": 0.028765481422293248, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4753, + "step": 72 + }, + { + "epoch": 0.029165001997602878, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4922, + "step": 73 + }, + { + "epoch": 0.029564522572912505, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4759, + "step": 74 + }, + { + "epoch": 0.02996404314822213, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.4703, + "step": 75 + }, + { + "epoch": 0.030363563723531762, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4666, + "step": 76 + }, + { + "epoch": 0.03076308429884139, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4628, + "step": 77 + }, + { + "epoch": 0.03116260487415102, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.4656, + "step": 78 + }, + { + "epoch": 0.031562125449460646, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4695, + "step": 79 + }, + { + "epoch": 0.03196164602477027, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4517, + "step": 80 + }, + { + "epoch": 0.032361166600079906, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4577, + "step": 81 + }, + { + "epoch": 0.03276068717538953, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4526, + "step": 82 + }, + { + "epoch": 0.03316020775069916, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4476, + "step": 83 + }, + { + "epoch": 0.033559728326008786, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.4499, + "step": 84 + }, + { + "epoch": 0.03395924890131842, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.4549, + "step": 85 + }, + { + "epoch": 0.03435876947662805, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4548, + "step": 86 + }, + { + "epoch": 0.034758290051937674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4525, + "step": 87 + }, + { + "epoch": 0.0351578106272473, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 88 + }, + { + "epoch": 0.035557331202556934, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4343, + "step": 89 + }, + { + "epoch": 0.03595685177786656, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4398, + "step": 90 + }, + { + "epoch": 0.03635637235317619, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4431, + "step": 91 + }, + { + "epoch": 0.036755892928485814, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.438, + "step": 92 + }, + { + "epoch": 0.03715541350379545, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4291, + "step": 93 + }, + { + "epoch": 0.037554934079105075, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.4373, + "step": 94 + }, + { + "epoch": 0.0379544546544147, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4409, + "step": 95 + }, + { + "epoch": 0.03835397522972433, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4305, + "step": 96 + }, + { + "epoch": 0.03875349580503396, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.4337, + "step": 97 + }, + { + "epoch": 0.03915301638034359, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4309, + "step": 98 + }, + { + "epoch": 0.039552536955653216, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 99 + }, + { + "epoch": 0.03995205753096284, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 100 + }, + { + "epoch": 0.040351578106272476, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.42, + "step": 101 + }, + { + "epoch": 0.0407510986815821, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4211, + "step": 102 + }, + { + "epoch": 0.04115061925689173, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4342, + "step": 103 + }, + { + "epoch": 0.041550139832201356, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4191, + "step": 104 + }, + { + "epoch": 0.04194966040751099, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.4234, + "step": 105 + }, + { + "epoch": 0.04234918098282062, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 106 + }, + { + "epoch": 0.042748701558130243, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4308, + "step": 107 + }, + { + "epoch": 0.04314822213343987, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.4189, + "step": 108 + }, + { + "epoch": 0.043547742708749504, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 109 + }, + { + "epoch": 0.04394726328405913, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4278, + "step": 110 + }, + { + "epoch": 0.04434678385936876, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 111 + }, + { + "epoch": 0.044746304434678384, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.4276, + "step": 112 + }, + { + "epoch": 0.04514582500998801, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.4203, + "step": 113 + }, + { + "epoch": 0.045545345585297645, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4313, + "step": 114 + }, + { + "epoch": 0.04594486616060727, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4186, + "step": 115 + }, + { + "epoch": 0.0463443867359169, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4215, + "step": 116 + }, + { + "epoch": 0.046743907311226525, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.4272, + "step": 117 + }, + { + "epoch": 0.04714342788653616, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4262, + "step": 118 + }, + { + "epoch": 0.047542948461845785, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 119 + }, + { + "epoch": 0.04794246903715541, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4071, + "step": 120 + }, + { + "epoch": 0.04834198961246504, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4241, + "step": 121 + }, + { + "epoch": 0.04874151018777467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 122 + }, + { + "epoch": 0.0491410307630843, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4252, + "step": 123 + }, + { + "epoch": 0.049540551338393926, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.424, + "step": 124 + }, + { + "epoch": 0.04994007191370355, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 125 + }, + { + "epoch": 0.05033959248901319, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.412, + "step": 126 + }, + { + "epoch": 0.05073911306432281, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4199, + "step": 127 + }, + { + "epoch": 0.05113863363963244, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 128 + }, + { + "epoch": 0.05153815421494207, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 129 + }, + { + "epoch": 0.0519376747902517, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 130 + }, + { + "epoch": 0.05233719536556133, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4049, + "step": 131 + }, + { + "epoch": 0.052736715940870954, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4134, + "step": 132 + }, + { + "epoch": 0.05313623651618058, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 133 + }, + { + "epoch": 0.053535757091490214, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4156, + "step": 134 + }, + { + "epoch": 0.05393527766679984, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 135 + }, + { + "epoch": 0.05433479824210947, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 136 + }, + { + "epoch": 0.054734318817419095, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 137 + }, + { + "epoch": 0.05513383939272873, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 138 + }, + { + "epoch": 0.055533359968038355, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.428, + "step": 139 + }, + { + "epoch": 0.05593288054334798, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 140 + }, + { + "epoch": 0.05633240111865761, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.4083, + "step": 141 + }, + { + "epoch": 0.05673192169396724, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 142 + }, + { + "epoch": 0.05713144226927687, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 143 + }, + { + "epoch": 0.057530962844586496, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 144 + }, + { + "epoch": 0.05793048341989612, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 145 + }, + { + "epoch": 0.058330003995205756, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 146 + }, + { + "epoch": 0.05872952457051538, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.4168, + "step": 147 + }, + { + "epoch": 0.05912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 148 + }, + { + "epoch": 0.05952856572113464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 149 + }, + { + "epoch": 0.05992808629644426, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 150 + }, + { + "epoch": 0.0603276068717539, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4171, + "step": 151 + }, + { + "epoch": 0.060727127447063524, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 152 + }, + { + "epoch": 0.06112664802237315, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 153 + }, + { + "epoch": 0.06152616859768278, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3966, + "step": 154 + }, + { + "epoch": 0.06192568917299241, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.4121, + "step": 155 + }, + { + "epoch": 0.06232520974830204, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 156 + }, + { + "epoch": 0.06272473032361167, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 157 + }, + { + "epoch": 0.06312425089892129, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 158 + }, + { + "epoch": 0.06352377147423093, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.413, + "step": 159 + }, + { + "epoch": 0.06392329204954054, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4174, + "step": 160 + }, + { + "epoch": 0.06432281262485018, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4133, + "step": 161 + }, + { + "epoch": 0.06472233320015981, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 162 + }, + { + "epoch": 0.06512185377546943, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4128, + "step": 163 + }, + { + "epoch": 0.06552137435077907, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3945, + "step": 164 + }, + { + "epoch": 0.0659208949260887, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 165 + }, + { + "epoch": 0.06632041550139832, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 166 + }, + { + "epoch": 0.06671993607670795, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4104, + "step": 167 + }, + { + "epoch": 0.06711945665201757, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.4144, + "step": 168 + }, + { + "epoch": 0.0675189772273272, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4059, + "step": 169 + }, + { + "epoch": 0.06791849780263684, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 170 + }, + { + "epoch": 0.06831801837794646, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 171 + }, + { + "epoch": 0.0687175389532561, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 172 + }, + { + "epoch": 0.06911705952856573, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 173 + }, + { + "epoch": 0.06951658010387535, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 174 + }, + { + "epoch": 0.06991610067918498, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4097, + "step": 175 + }, + { + "epoch": 0.0703156212544946, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 176 + }, + { + "epoch": 0.07071514182980423, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4052, + "step": 177 + }, + { + "epoch": 0.07111466240511387, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.417, + "step": 178 + }, + { + "epoch": 0.07151418298042349, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 179 + }, + { + "epoch": 0.07191370355573312, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 180 + }, + { + "epoch": 0.07231322413104276, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 181 + }, + { + "epoch": 0.07271274470635238, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 182 + }, + { + "epoch": 0.07311226528166201, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 183 + }, + { + "epoch": 0.07351178585697163, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 184 + }, + { + "epoch": 0.07391130643228126, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 185 + }, + { + "epoch": 0.0743108270075909, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 186 + }, + { + "epoch": 0.07471034758290052, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 187 + }, + { + "epoch": 0.07510986815821015, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 188 + }, + { + "epoch": 0.07550938873351977, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.4132, + "step": 189 + }, + { + "epoch": 0.0759089093088294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.4067, + "step": 190 + }, + { + "epoch": 0.07630842988413904, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 191 + }, + { + "epoch": 0.07670795045944866, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 192 + }, + { + "epoch": 0.07710747103475829, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3964, + "step": 193 + }, + { + "epoch": 0.07750699161006792, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 194 + }, + { + "epoch": 0.07790651218537754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3979, + "step": 195 + }, + { + "epoch": 0.07830603276068718, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 196 + }, + { + "epoch": 0.0787055533359968, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4111, + "step": 197 + }, + { + "epoch": 0.07910507391130643, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 198 + }, + { + "epoch": 0.07950459448661606, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 199 + }, + { + "epoch": 0.07990411506192568, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 200 + }, + { + "epoch": 0.08030363563723532, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 201 + }, + { + "epoch": 0.08070315621254495, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 202 + }, + { + "epoch": 0.08110267678785457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4021, + "step": 203 + }, + { + "epoch": 0.0815021973631642, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 204 + }, + { + "epoch": 0.08190171793847383, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 205 + }, + { + "epoch": 0.08230123851378346, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 206 + }, + { + "epoch": 0.08270075908909309, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.4057, + "step": 207 + }, + { + "epoch": 0.08310027966440271, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4158, + "step": 208 + }, + { + "epoch": 0.08349980023971235, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3898, + "step": 209 + }, + { + "epoch": 0.08389932081502198, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 210 + }, + { + "epoch": 0.0842988413903316, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.4193, + "step": 211 + }, + { + "epoch": 0.08469836196564123, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 212 + }, + { + "epoch": 0.08509788254095085, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 213 + }, + { + "epoch": 0.08549740311626049, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 214 + }, + { + "epoch": 0.08589692369157012, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4098, + "step": 215 + }, + { + "epoch": 0.08629644426687974, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.4015, + "step": 216 + }, + { + "epoch": 0.08669596484218937, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.4113, + "step": 217 + }, + { + "epoch": 0.08709548541749901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 218 + }, + { + "epoch": 0.08749500599280863, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 219 + }, + { + "epoch": 0.08789452656811826, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 220 + }, + { + "epoch": 0.08829404714342788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 221 + }, + { + "epoch": 0.08869356771873751, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.409, + "step": 222 + }, + { + "epoch": 0.08909308829404715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 223 + }, + { + "epoch": 0.08949260886935677, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 224 + }, + { + "epoch": 0.0898921294446664, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 225 + }, + { + "epoch": 0.09029165001997602, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 226 + }, + { + "epoch": 0.09069117059528566, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3904, + "step": 227 + }, + { + "epoch": 0.09109069117059529, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 228 + }, + { + "epoch": 0.09149021174590491, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 229 + }, + { + "epoch": 0.09188973232121454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 230 + }, + { + "epoch": 0.09228925289652418, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 231 + }, + { + "epoch": 0.0926887734718338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 232 + }, + { + "epoch": 0.09308829404714343, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 233 + }, + { + "epoch": 0.09348781462245305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 234 + }, + { + "epoch": 0.09388733519776268, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 235 + }, + { + "epoch": 0.09428685577307232, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 236 + }, + { + "epoch": 0.09468637634838194, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 237 + }, + { + "epoch": 0.09508589692369157, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 238 + }, + { + "epoch": 0.0954854174990012, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 239 + }, + { + "epoch": 0.09588493807431082, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4064, + "step": 240 + }, + { + "epoch": 0.09628445864962046, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 241 + }, + { + "epoch": 0.09668397922493008, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 242 + }, + { + "epoch": 0.09708349980023971, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 243 + }, + { + "epoch": 0.09748302037554935, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 244 + }, + { + "epoch": 0.09788254095085897, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3961, + "step": 245 + }, + { + "epoch": 0.0982820615261686, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4061, + "step": 246 + }, + { + "epoch": 0.09868158210147823, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 247 + }, + { + "epoch": 0.09908110267678785, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 248 + }, + { + "epoch": 0.09948062325209749, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 249 + }, + { + "epoch": 0.0998801438274071, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 250 + }, + { + "epoch": 0.10027966440271674, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 251 + }, + { + "epoch": 0.10067918497802637, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 252 + }, + { + "epoch": 0.10107870555333599, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.4024, + "step": 253 + }, + { + "epoch": 0.10147822612864563, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 254 + }, + { + "epoch": 0.10187774670395526, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 255 + }, + { + "epoch": 0.10227726727926488, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 256 + }, + { + "epoch": 0.10267678785457451, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 257 + }, + { + "epoch": 0.10307630842988413, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 258 + }, + { + "epoch": 0.10347582900519377, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 259 + }, + { + "epoch": 0.1038753495805034, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 260 + }, + { + "epoch": 0.10427487015581302, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 261 + }, + { + "epoch": 0.10467439073112265, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 262 + }, + { + "epoch": 0.10507391130643227, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 263 + }, + { + "epoch": 0.10547343188174191, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 264 + }, + { + "epoch": 0.10587295245705154, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3988, + "step": 265 + }, + { + "epoch": 0.10627247303236116, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 266 + }, + { + "epoch": 0.1066719936076708, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.4079, + "step": 267 + }, + { + "epoch": 0.10707151418298043, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 268 + }, + { + "epoch": 0.10747103475829005, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3954, + "step": 269 + }, + { + "epoch": 0.10787055533359968, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 270 + }, + { + "epoch": 0.1082700759089093, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 271 + }, + { + "epoch": 0.10866959648421894, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 272 + }, + { + "epoch": 0.10906911705952857, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 273 + }, + { + "epoch": 0.10946863763483819, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3899, + "step": 274 + }, + { + "epoch": 0.10986815821014782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 275 + }, + { + "epoch": 0.11026767878545746, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4012, + "step": 276 + }, + { + "epoch": 0.11066719936076708, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 277 + }, + { + "epoch": 0.11106671993607671, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3948, + "step": 278 + }, + { + "epoch": 0.11146624051138633, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4175, + "step": 279 + }, + { + "epoch": 0.11186576108669596, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 280 + }, + { + "epoch": 0.1122652816620056, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 281 + }, + { + "epoch": 0.11266480223731522, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3994, + "step": 282 + }, + { + "epoch": 0.11306432281262485, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 283 + }, + { + "epoch": 0.11346384338793448, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 284 + }, + { + "epoch": 0.1138633639632441, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 285 + }, + { + "epoch": 0.11426288453855374, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 286 + }, + { + "epoch": 0.11466240511386336, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4045, + "step": 287 + }, + { + "epoch": 0.11506192568917299, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 288 + }, + { + "epoch": 0.11546144626448263, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 289 + }, + { + "epoch": 0.11586096683979225, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 290 + }, + { + "epoch": 0.11626048741510188, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 291 + }, + { + "epoch": 0.11666000799041151, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 292 + }, + { + "epoch": 0.11705952856572113, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 293 + }, + { + "epoch": 0.11745904914103077, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3991, + "step": 294 + }, + { + "epoch": 0.11785856971634039, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 295 + }, + { + "epoch": 0.11825809029165002, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 296 + }, + { + "epoch": 0.11865761086695965, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 297 + }, + { + "epoch": 0.11905713144226927, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 298 + }, + { + "epoch": 0.11945665201757891, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 299 + }, + { + "epoch": 0.11985617259288853, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 300 + }, + { + "epoch": 0.12025569316819816, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3998, + "step": 301 + }, + { + "epoch": 0.1206552137435078, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 302 + }, + { + "epoch": 0.12105473431881741, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 303 + }, + { + "epoch": 0.12145425489412705, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 304 + }, + { + "epoch": 0.12185377546943668, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 305 + }, + { + "epoch": 0.1222532960447463, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 306 + }, + { + "epoch": 0.12265281662005593, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 307 + }, + { + "epoch": 0.12305233719536555, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 308 + }, + { + "epoch": 0.12345185777067519, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 309 + }, + { + "epoch": 0.12385137834598482, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 310 + }, + { + "epoch": 0.12425089892129444, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 311 + }, + { + "epoch": 0.12465041949660408, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 312 + }, + { + "epoch": 0.1250499400719137, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 313 + }, + { + "epoch": 0.12544946064722334, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 314 + }, + { + "epoch": 0.12584898122253296, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 315 + }, + { + "epoch": 0.12624850179784258, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 316 + }, + { + "epoch": 0.12664802237315223, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 317 + }, + { + "epoch": 0.12704754294846185, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 318 + }, + { + "epoch": 0.12744706352377147, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 319 + }, + { + "epoch": 0.1278465840990811, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 320 + }, + { + "epoch": 0.12824610467439074, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 321 + }, + { + "epoch": 0.12864562524970036, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3987, + "step": 322 + }, + { + "epoch": 0.12904514582500998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 323 + }, + { + "epoch": 0.12944466640031962, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 324 + }, + { + "epoch": 0.12984418697562924, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 325 + }, + { + "epoch": 0.13024370755093886, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 326 + }, + { + "epoch": 0.1306432281262485, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 327 + }, + { + "epoch": 0.13104274870155813, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 328 + }, + { + "epoch": 0.13144226927686775, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 329 + }, + { + "epoch": 0.1318417898521774, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 330 + }, + { + "epoch": 0.13224131042748702, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 331 + }, + { + "epoch": 0.13264083100279664, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 332 + }, + { + "epoch": 0.1330403515781063, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 333 + }, + { + "epoch": 0.1334398721534159, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 334 + }, + { + "epoch": 0.13383939272872553, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 335 + }, + { + "epoch": 0.13423891330403515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 336 + }, + { + "epoch": 0.1346384338793448, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 337 + }, + { + "epoch": 0.1350379544546544, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 338 + }, + { + "epoch": 0.13543747502996403, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 339 + }, + { + "epoch": 0.13583699560527368, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 340 + }, + { + "epoch": 0.1362365161805833, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 341 + }, + { + "epoch": 0.13663603675589292, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 342 + }, + { + "epoch": 0.13703555733120257, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3889, + "step": 343 + }, + { + "epoch": 0.1374350779065122, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3995, + "step": 344 + }, + { + "epoch": 0.1378345984818218, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 345 + }, + { + "epoch": 0.13823411905713145, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 346 + }, + { + "epoch": 0.13863363963244107, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3983, + "step": 347 + }, + { + "epoch": 0.1390331602077507, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 348 + }, + { + "epoch": 0.13943268078306031, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 349 + }, + { + "epoch": 0.13983220135836996, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 350 + }, + { + "epoch": 0.14023172193367958, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 351 + }, + { + "epoch": 0.1406312425089892, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 352 + }, + { + "epoch": 0.14103076308429885, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 353 + }, + { + "epoch": 0.14143028365960847, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.391, + "step": 354 + }, + { + "epoch": 0.1418298042349181, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3937, + "step": 355 + }, + { + "epoch": 0.14222932481022774, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3858, + "step": 356 + }, + { + "epoch": 0.14262884538553736, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 357 + }, + { + "epoch": 0.14302836596084698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 358 + }, + { + "epoch": 0.14342788653615662, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 359 + }, + { + "epoch": 0.14382740711146624, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 360 + }, + { + "epoch": 0.14422692768677586, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 361 + }, + { + "epoch": 0.1446264482620855, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 362 + }, + { + "epoch": 0.14502596883739513, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 363 + }, + { + "epoch": 0.14542548941270475, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 364 + }, + { + "epoch": 0.14582500998801437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 365 + }, + { + "epoch": 0.14622453056332402, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 366 + }, + { + "epoch": 0.14662405113863364, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 367 + }, + { + "epoch": 0.14702357171394326, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 368 + }, + { + "epoch": 0.1474230922892529, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 369 + }, + { + "epoch": 0.14782261286456252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 370 + }, + { + "epoch": 0.14822213343987214, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 371 + }, + { + "epoch": 0.1486216540151818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 372 + }, + { + "epoch": 0.1490211745904914, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 373 + }, + { + "epoch": 0.14942069516580103, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3942, + "step": 374 + }, + { + "epoch": 0.14982021574111068, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 375 + }, + { + "epoch": 0.1502197363164203, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 376 + }, + { + "epoch": 0.15061925689172992, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3903, + "step": 377 + }, + { + "epoch": 0.15101877746703954, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 378 + }, + { + "epoch": 0.1514182980423492, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 379 + }, + { + "epoch": 0.1518178186176588, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 380 + }, + { + "epoch": 0.15221733919296843, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 381 + }, + { + "epoch": 0.15261685976827807, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 382 + }, + { + "epoch": 0.1530163803435877, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 383 + }, + { + "epoch": 0.1534159009188973, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 384 + }, + { + "epoch": 0.15381542149420696, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 385 + }, + { + "epoch": 0.15421494206951658, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 386 + }, + { + "epoch": 0.1546144626448262, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 387 + }, + { + "epoch": 0.15501398322013585, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 388 + }, + { + "epoch": 0.15541350379544547, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 389 + }, + { + "epoch": 0.1558130243707551, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 390 + }, + { + "epoch": 0.15621254494606474, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 391 + }, + { + "epoch": 0.15661206552137436, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 392 + }, + { + "epoch": 0.15701158609668397, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 393 + }, + { + "epoch": 0.1574111066719936, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 394 + }, + { + "epoch": 0.15781062724730324, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 395 + }, + { + "epoch": 0.15821014782261286, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 396 + }, + { + "epoch": 0.15860966839792248, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 397 + }, + { + "epoch": 0.15900918897323213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 398 + }, + { + "epoch": 0.15940870954854175, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 399 + }, + { + "epoch": 0.15980823012385137, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3929, + "step": 400 + }, + { + "epoch": 0.16020775069916102, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3835, + "step": 401 + }, + { + "epoch": 0.16060727127447064, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 402 + }, + { + "epoch": 0.16100679184978026, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 403 + }, + { + "epoch": 0.1614063124250899, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.4032, + "step": 404 + }, + { + "epoch": 0.16180583300039952, + "grad_norm": 3.984375, + "learning_rate": 0.0002, + "loss": 1.48, + "step": 405 + }, + { + "epoch": 0.16220535357570914, + "grad_norm": 2.234375, + "learning_rate": 0.0002, + "loss": 1.4185, + "step": 406 + }, + { + "epoch": 0.1626048741510188, + "grad_norm": 3.921875, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 407 + }, + { + "epoch": 0.1630043947263284, + "grad_norm": 3.578125, + "learning_rate": 0.0002, + "loss": 1.4966, + "step": 408 + }, + { + "epoch": 0.16340391530163803, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 409 + }, + { + "epoch": 0.16380343587694765, + "grad_norm": 2.671875, + "learning_rate": 0.0002, + "loss": 1.4878, + "step": 410 + }, + { + "epoch": 0.1642029564522573, + "grad_norm": 2.34375, + "learning_rate": 0.0002, + "loss": 1.491, + "step": 411 + }, + { + "epoch": 0.16460247702756692, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 412 + }, + { + "epoch": 0.16500199760287654, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4232, + "step": 413 + }, + { + "epoch": 0.16540151817818619, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.4109, + "step": 414 + }, + { + "epoch": 0.1658010387534958, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.4187, + "step": 415 + }, + { + "epoch": 0.16620055932880543, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 416 + }, + { + "epoch": 0.16660007990411507, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 417 + }, + { + "epoch": 0.1669996004794247, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 418 + }, + { + "epoch": 0.1673991210547343, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3982, + "step": 419 + }, + { + "epoch": 0.16779864163004396, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 420 + }, + { + "epoch": 0.16819816220535358, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 421 + }, + { + "epoch": 0.1685976827806632, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 422 + }, + { + "epoch": 0.16899720335597282, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3933, + "step": 423 + }, + { + "epoch": 0.16939672393128247, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 424 + }, + { + "epoch": 0.1697962445065921, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3909, + "step": 425 + }, + { + "epoch": 0.1701957650819017, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3978, + "step": 426 + }, + { + "epoch": 0.17059528565721135, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 427 + }, + { + "epoch": 0.17099480623252097, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3955, + "step": 428 + }, + { + "epoch": 0.1713943268078306, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 429 + }, + { + "epoch": 0.17179384738314024, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 430 + }, + { + "epoch": 0.17219336795844986, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3965, + "step": 431 + }, + { + "epoch": 0.17259288853375948, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 432 + }, + { + "epoch": 0.17299240910906913, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 433 + }, + { + "epoch": 0.17339192968437875, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 434 + }, + { + "epoch": 0.17379145025968837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 435 + }, + { + "epoch": 0.17419097083499802, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 436 + }, + { + "epoch": 0.17459049141030764, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 437 + }, + { + "epoch": 0.17499001198561726, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3928, + "step": 438 + }, + { + "epoch": 0.17538953256092688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 439 + }, + { + "epoch": 0.17578905313623652, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 440 + }, + { + "epoch": 0.17618857371154614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 441 + }, + { + "epoch": 0.17658809428685576, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3912, + "step": 442 + }, + { + "epoch": 0.1769876148621654, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 443 + }, + { + "epoch": 0.17738713543747503, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 444 + }, + { + "epoch": 0.17778665601278465, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 445 + }, + { + "epoch": 0.1781861765880943, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 446 + }, + { + "epoch": 0.17858569716340392, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 447 + }, + { + "epoch": 0.17898521773871354, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 448 + }, + { + "epoch": 0.17938473831402318, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 449 + }, + { + "epoch": 0.1797842588893328, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 450 + }, + { + "epoch": 0.18018377946464242, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 451 + }, + { + "epoch": 0.18058330003995204, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 452 + }, + { + "epoch": 0.1809828206152617, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 453 + }, + { + "epoch": 0.1813823411905713, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 454 + }, + { + "epoch": 0.18178186176588093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.4001, + "step": 455 + }, + { + "epoch": 0.18218138234119058, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 456 + }, + { + "epoch": 0.1825809029165002, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 457 + }, + { + "epoch": 0.18298042349180982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 458 + }, + { + "epoch": 0.18337994406711947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 459 + }, + { + "epoch": 0.18377946464242909, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 460 + }, + { + "epoch": 0.1841789852177387, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 461 + }, + { + "epoch": 0.18457850579304835, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 462 + }, + { + "epoch": 0.18497802636835797, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 463 + }, + { + "epoch": 0.1853775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 464 + }, + { + "epoch": 0.18577706751897724, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 465 + }, + { + "epoch": 0.18617658809428686, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3796, + "step": 466 + }, + { + "epoch": 0.18657610866959648, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 467 + }, + { + "epoch": 0.1869756292449061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 468 + }, + { + "epoch": 0.18737514982021575, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 469 + }, + { + "epoch": 0.18777467039552537, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 470 + }, + { + "epoch": 0.188174190970835, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 471 + }, + { + "epoch": 0.18857371154614463, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 472 + }, + { + "epoch": 0.18897323212145425, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 473 + }, + { + "epoch": 0.18937275269676387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 474 + }, + { + "epoch": 0.18977227327207352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 475 + }, + { + "epoch": 0.19017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 476 + }, + { + "epoch": 0.19057131442269276, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 477 + }, + { + "epoch": 0.1909708349980024, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 478 + }, + { + "epoch": 0.19137035557331203, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 479 + }, + { + "epoch": 0.19176987614862165, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3807, + "step": 480 + }, + { + "epoch": 0.1921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 481 + }, + { + "epoch": 0.19256891729924092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 482 + }, + { + "epoch": 0.19296843787455054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 483 + }, + { + "epoch": 0.19336795844986016, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 484 + }, + { + "epoch": 0.1937674790251698, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 485 + }, + { + "epoch": 0.19416699960047942, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 486 + }, + { + "epoch": 0.19456652017578904, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 487 + }, + { + "epoch": 0.1949660407510987, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 488 + }, + { + "epoch": 0.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 489 + }, + { + "epoch": 0.19576508190171793, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 490 + }, + { + "epoch": 0.19616460247702758, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3816, + "step": 491 + }, + { + "epoch": 0.1965641230523372, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 492 + }, + { + "epoch": 0.19696364362764682, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3824, + "step": 493 + }, + { + "epoch": 0.19736316420295646, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.387, + "step": 494 + }, + { + "epoch": 0.19776268477826608, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 495 + }, + { + "epoch": 0.1981622053535757, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 496 + }, + { + "epoch": 0.19856172592888532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 497 + }, + { + "epoch": 0.19896124650419497, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 498 + }, + { + "epoch": 0.1993607670795046, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 499 + }, + { + "epoch": 0.1997602876548142, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 500 + }, + { + "epoch": 0.20015980823012386, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 501 + }, + { + "epoch": 0.20055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 502 + }, + { + "epoch": 0.2009588493807431, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 503 + }, + { + "epoch": 0.20135836995605275, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 504 + }, + { + "epoch": 0.20175789053136237, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 505 + }, + { + "epoch": 0.20215741110667199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 506 + }, + { + "epoch": 0.20255693168198163, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 507 + }, + { + "epoch": 0.20295645225729125, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 508 + }, + { + "epoch": 0.20335597283260087, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 509 + }, + { + "epoch": 0.20375549340791052, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 510 + }, + { + "epoch": 0.20415501398322014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.384, + "step": 511 + }, + { + "epoch": 0.20455453455852976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 512 + }, + { + "epoch": 0.20495405513383938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 513 + }, + { + "epoch": 0.20535357570914903, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 514 + }, + { + "epoch": 0.20575309628445865, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 515 + }, + { + "epoch": 0.20615261685976827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 516 + }, + { + "epoch": 0.20655213743507791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.38, + "step": 517 + }, + { + "epoch": 0.20695165801038753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 518 + }, + { + "epoch": 0.20735117858569715, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 519 + }, + { + "epoch": 0.2077506991610068, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 520 + }, + { + "epoch": 0.20815021973631642, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 521 + }, + { + "epoch": 0.20854974031162604, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 522 + }, + { + "epoch": 0.2089492608869357, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 523 + }, + { + "epoch": 0.2093487814622453, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 524 + }, + { + "epoch": 0.20974830203755493, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3845, + "step": 525 + }, + { + "epoch": 0.21014782261286455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 526 + }, + { + "epoch": 0.2105473431881742, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 527 + }, + { + "epoch": 0.21094686376348382, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3958, + "step": 528 + }, + { + "epoch": 0.21134638433879344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 529 + }, + { + "epoch": 0.21174590491410308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 530 + }, + { + "epoch": 0.2121454254894127, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 531 + }, + { + "epoch": 0.21254494606472232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 532 + }, + { + "epoch": 0.21294446664003197, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 533 + }, + { + "epoch": 0.2133439872153416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 534 + }, + { + "epoch": 0.2137435077906512, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 535 + }, + { + "epoch": 0.21414302836596086, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 536 + }, + { + "epoch": 0.21454254894127048, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 537 + }, + { + "epoch": 0.2149420695165801, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 538 + }, + { + "epoch": 0.21534159009188975, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 539 + }, + { + "epoch": 0.21574111066719937, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 540 + }, + { + "epoch": 0.21614063124250898, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 541 + }, + { + "epoch": 0.2165401518178186, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 542 + }, + { + "epoch": 0.21693967239312825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3754, + "step": 543 + }, + { + "epoch": 0.21733919296843787, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 544 + }, + { + "epoch": 0.2177387135437475, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 545 + }, + { + "epoch": 0.21813823411905714, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 546 + }, + { + "epoch": 0.21853775469436676, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3887, + "step": 547 + }, + { + "epoch": 0.21893727526967638, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 548 + }, + { + "epoch": 0.21933679584498603, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 549 + }, + { + "epoch": 0.21973631642029565, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 550 + }, + { + "epoch": 0.22013583699560527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 551 + }, + { + "epoch": 0.2205353575709149, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 552 + }, + { + "epoch": 0.22093487814622453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 553 + }, + { + "epoch": 0.22133439872153415, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 554 + }, + { + "epoch": 0.2217339192968438, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3992, + "step": 555 + }, + { + "epoch": 0.22213343987215342, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 556 + }, + { + "epoch": 0.22253296044746304, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 557 + }, + { + "epoch": 0.22293248102277266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 558 + }, + { + "epoch": 0.2233320015980823, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 559 + }, + { + "epoch": 0.22373152217339193, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 560 + }, + { + "epoch": 0.22413104274870155, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 561 + }, + { + "epoch": 0.2245305633240112, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 562 + }, + { + "epoch": 0.22493008389932082, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 563 + }, + { + "epoch": 0.22532960447463043, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 564 + }, + { + "epoch": 0.22572912504994008, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 565 + }, + { + "epoch": 0.2261286456252497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 566 + }, + { + "epoch": 0.22652816620055932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 567 + }, + { + "epoch": 0.22692768677586897, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 568 + }, + { + "epoch": 0.2273272073511786, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 569 + }, + { + "epoch": 0.2277267279264882, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 570 + }, + { + "epoch": 0.22812624850179783, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 571 + }, + { + "epoch": 0.22852576907710748, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 572 + }, + { + "epoch": 0.2289252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 573 + }, + { + "epoch": 0.22932481022772672, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 574 + }, + { + "epoch": 0.22972433080303636, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 575 + }, + { + "epoch": 0.23012385137834598, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 576 + }, + { + "epoch": 0.2305233719536556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3786, + "step": 577 + }, + { + "epoch": 0.23092289252896525, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 578 + }, + { + "epoch": 0.23132241310427487, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 579 + }, + { + "epoch": 0.2317219336795845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 580 + }, + { + "epoch": 0.23212145425489414, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 581 + }, + { + "epoch": 0.23252097483020376, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3781, + "step": 582 + }, + { + "epoch": 0.23292049540551338, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 583 + }, + { + "epoch": 0.23332001598082303, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 584 + }, + { + "epoch": 0.23371953655613265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 585 + }, + { + "epoch": 0.23411905713144227, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 586 + }, + { + "epoch": 0.23451857770675189, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 587 + }, + { + "epoch": 0.23491809828206153, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 588 + }, + { + "epoch": 0.23531761885737115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 589 + }, + { + "epoch": 0.23571713943268077, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 590 + }, + { + "epoch": 0.23611666000799042, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 591 + }, + { + "epoch": 0.23651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 592 + }, + { + "epoch": 0.23691570115860966, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 593 + }, + { + "epoch": 0.2373152217339193, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 594 + }, + { + "epoch": 0.23771474230922893, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 595 + }, + { + "epoch": 0.23811426288453855, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 596 + }, + { + "epoch": 0.2385137834598482, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3914, + "step": 597 + }, + { + "epoch": 0.23891330403515781, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 598 + }, + { + "epoch": 0.23931282461046743, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 599 + }, + { + "epoch": 0.23971234518577705, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 600 + }, + { + "epoch": 0.2401118657610867, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 601 + }, + { + "epoch": 0.24051138633639632, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 602 + }, + { + "epoch": 0.24091090691170594, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 603 + }, + { + "epoch": 0.2413104274870156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 604 + }, + { + "epoch": 0.2417099480623252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 605 + }, + { + "epoch": 0.24210946863763483, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 606 + }, + { + "epoch": 0.24250898921294448, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 607 + }, + { + "epoch": 0.2429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 608 + }, + { + "epoch": 0.24330803036356372, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 609 + }, + { + "epoch": 0.24370755093887336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 610 + }, + { + "epoch": 0.24410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 611 + }, + { + "epoch": 0.2445065920894926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 612 + }, + { + "epoch": 0.24490611266480225, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 613 + }, + { + "epoch": 0.24530563324011187, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 614 + }, + { + "epoch": 0.2457051538154215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 615 + }, + { + "epoch": 0.2461046743907311, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3783, + "step": 616 + }, + { + "epoch": 0.24650419496604076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 617 + }, + { + "epoch": 0.24690371554135038, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 618 + }, + { + "epoch": 0.24730323611666, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 619 + }, + { + "epoch": 0.24770275669196964, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 620 + }, + { + "epoch": 0.24810227726727926, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 621 + }, + { + "epoch": 0.24850179784258888, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 622 + }, + { + "epoch": 0.24890131841789853, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 623 + }, + { + "epoch": 0.24930083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 624 + }, + { + "epoch": 0.24970035956851777, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 625 + }, + { + "epoch": 0.2500998801438274, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 626 + }, + { + "epoch": 0.250499400719137, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3748, + "step": 627 + }, + { + "epoch": 0.2508989212944467, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 628 + }, + { + "epoch": 0.2512984418697563, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.377, + "step": 629 + }, + { + "epoch": 0.2516979624450659, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 630 + }, + { + "epoch": 0.25209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 631 + }, + { + "epoch": 0.25249700359568517, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 632 + }, + { + "epoch": 0.2528965241709948, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 633 + }, + { + "epoch": 0.25329604474630446, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 634 + }, + { + "epoch": 0.2536955653216141, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 635 + }, + { + "epoch": 0.2540950858969237, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 636 + }, + { + "epoch": 0.2544946064722333, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 637 + }, + { + "epoch": 0.25489412704754294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 638 + }, + { + "epoch": 0.25529364762285256, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 639 + }, + { + "epoch": 0.2556931681981622, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 640 + }, + { + "epoch": 0.25609268877347185, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 641 + }, + { + "epoch": 0.2564922093487815, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 642 + }, + { + "epoch": 0.2568917299240911, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 643 + }, + { + "epoch": 0.2572912504994007, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 644 + }, + { + "epoch": 0.25769077107471033, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 645 + }, + { + "epoch": 0.25809029165001995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 646 + }, + { + "epoch": 0.25848981222532963, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 647 + }, + { + "epoch": 0.25888933280063925, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 648 + }, + { + "epoch": 0.25928885337594887, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 649 + }, + { + "epoch": 0.2596883739512585, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 650 + }, + { + "epoch": 0.2600878945265681, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 651 + }, + { + "epoch": 0.26048741510187773, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 652 + }, + { + "epoch": 0.26088693567718735, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3866, + "step": 653 + }, + { + "epoch": 0.261286456252497, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 654 + }, + { + "epoch": 0.26168597682780664, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 655 + }, + { + "epoch": 0.26208549740311626, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 656 + }, + { + "epoch": 0.2624850179784259, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 657 + }, + { + "epoch": 0.2628845385537355, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3802, + "step": 658 + }, + { + "epoch": 0.2632840591290451, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 659 + }, + { + "epoch": 0.2636835797043548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 660 + }, + { + "epoch": 0.2640831002796644, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 661 + }, + { + "epoch": 0.26448262085497404, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 662 + }, + { + "epoch": 0.26488214143028366, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 663 + }, + { + "epoch": 0.2652816620055933, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 664 + }, + { + "epoch": 0.2656811825809029, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 665 + }, + { + "epoch": 0.2660807031562126, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 666 + }, + { + "epoch": 0.2664802237315222, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 667 + }, + { + "epoch": 0.2668797443068318, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 668 + }, + { + "epoch": 0.26727926488214143, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 669 + }, + { + "epoch": 0.26767878545745105, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 670 + }, + { + "epoch": 0.26807830603276067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 671 + }, + { + "epoch": 0.2684778266080703, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 672 + }, + { + "epoch": 0.26887734718337997, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 673 + }, + { + "epoch": 0.2692768677586896, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 674 + }, + { + "epoch": 0.2696763883339992, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 675 + }, + { + "epoch": 0.2700759089093088, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 676 + }, + { + "epoch": 0.27047542948461845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 677 + }, + { + "epoch": 0.27087495005992807, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 678 + }, + { + "epoch": 0.27127447063523774, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 679 + }, + { + "epoch": 0.27167399121054736, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 680 + }, + { + "epoch": 0.272073511785857, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 681 + }, + { + "epoch": 0.2724730323611666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 682 + }, + { + "epoch": 0.2728725529364762, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 683 + }, + { + "epoch": 0.27327207351178584, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 684 + }, + { + "epoch": 0.27367159408709546, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 685 + }, + { + "epoch": 0.27407111466240514, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3747, + "step": 686 + }, + { + "epoch": 0.27447063523771476, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 687 + }, + { + "epoch": 0.2748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 688 + }, + { + "epoch": 0.275269676388334, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 689 + }, + { + "epoch": 0.2756691969636436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 690 + }, + { + "epoch": 0.27606871753895323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 691 + }, + { + "epoch": 0.2764682381142629, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 692 + }, + { + "epoch": 0.27686775868957253, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 693 + }, + { + "epoch": 0.27726727926488215, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 694 + }, + { + "epoch": 0.27766679984019177, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3842, + "step": 695 + }, + { + "epoch": 0.2780663204155014, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 696 + }, + { + "epoch": 0.278465840990811, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 697 + }, + { + "epoch": 0.27886536156612063, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 698 + }, + { + "epoch": 0.2792648821414303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 699 + }, + { + "epoch": 0.2796644027167399, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 700 + }, + { + "epoch": 0.28006392329204954, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3742, + "step": 701 + }, + { + "epoch": 0.28046344386735916, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 702 + }, + { + "epoch": 0.2808629644426688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 703 + }, + { + "epoch": 0.2812624850179784, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 704 + }, + { + "epoch": 0.2816620055932881, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 705 + }, + { + "epoch": 0.2820615261685977, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 706 + }, + { + "epoch": 0.2824610467439073, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 707 + }, + { + "epoch": 0.28286056731921694, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 708 + }, + { + "epoch": 0.28326008789452656, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 709 + }, + { + "epoch": 0.2836596084698362, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 710 + }, + { + "epoch": 0.2840591290451458, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 711 + }, + { + "epoch": 0.2844586496204555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 712 + }, + { + "epoch": 0.2848581701957651, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 713 + }, + { + "epoch": 0.2852576907710747, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 714 + }, + { + "epoch": 0.28565721134638433, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 715 + }, + { + "epoch": 0.28605673192169395, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 716 + }, + { + "epoch": 0.28645625249700357, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 717 + }, + { + "epoch": 0.28685577307231325, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 718 + }, + { + "epoch": 0.28725529364762287, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 719 + }, + { + "epoch": 0.2876548142229325, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 720 + }, + { + "epoch": 0.2880543347982421, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3826, + "step": 721 + }, + { + "epoch": 0.2884538553735517, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 722 + }, + { + "epoch": 0.28885337594886135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 723 + }, + { + "epoch": 0.289252896524171, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 724 + }, + { + "epoch": 0.28965241709948064, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 725 + }, + { + "epoch": 0.29005193767479026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 726 + }, + { + "epoch": 0.2904514582500999, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 727 + }, + { + "epoch": 0.2908509788254095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 728 + }, + { + "epoch": 0.2912504994007191, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3854, + "step": 729 + }, + { + "epoch": 0.29165001997602874, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 730 + }, + { + "epoch": 0.2920495405513384, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 731 + }, + { + "epoch": 0.29244906112664804, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 732 + }, + { + "epoch": 0.29284858170195766, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 733 + }, + { + "epoch": 0.2932481022772673, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 734 + }, + { + "epoch": 0.2936476228525769, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 735 + }, + { + "epoch": 0.2940471434278865, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 736 + }, + { + "epoch": 0.2944466640031962, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 737 + }, + { + "epoch": 0.2948461845785058, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 738 + }, + { + "epoch": 0.29524570515381543, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 739 + }, + { + "epoch": 0.29564522572912505, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 740 + }, + { + "epoch": 0.29604474630443467, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 741 + }, + { + "epoch": 0.2964442668797443, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 742 + }, + { + "epoch": 0.2968437874550539, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 743 + }, + { + "epoch": 0.2972433080303636, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 744 + }, + { + "epoch": 0.2976428286056732, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 745 + }, + { + "epoch": 0.2980423491809828, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 746 + }, + { + "epoch": 0.29844186975629244, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 747 + }, + { + "epoch": 0.29884139033160206, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 748 + }, + { + "epoch": 0.2992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 749 + }, + { + "epoch": 0.29964043148222136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.382, + "step": 750 + }, + { + "epoch": 0.300039952057531, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 751 + }, + { + "epoch": 0.3004394726328406, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 752 + }, + { + "epoch": 0.3008389932081502, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 753 + }, + { + "epoch": 0.30123851378345984, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 754 + }, + { + "epoch": 0.30163803435876946, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 755 + }, + { + "epoch": 0.3020375549340791, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 756 + }, + { + "epoch": 0.30243707550938875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 757 + }, + { + "epoch": 0.3028365960846984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 758 + }, + { + "epoch": 0.303236116660008, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 759 + }, + { + "epoch": 0.3036356372353176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3698, + "step": 760 + }, + { + "epoch": 0.30403515781062723, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 761 + }, + { + "epoch": 0.30443467838593685, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3761, + "step": 762 + }, + { + "epoch": 0.3048341989612465, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 763 + }, + { + "epoch": 0.30523371953655615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 764 + }, + { + "epoch": 0.30563324011186577, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 765 + }, + { + "epoch": 0.3060327606871754, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 766 + }, + { + "epoch": 0.306432281262485, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 767 + }, + { + "epoch": 0.3068318018377946, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 768 + }, + { + "epoch": 0.3072313224131043, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3694, + "step": 769 + }, + { + "epoch": 0.3076308429884139, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 770 + }, + { + "epoch": 0.30803036356372354, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 771 + }, + { + "epoch": 0.30842988413903316, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 772 + }, + { + "epoch": 0.3088294047143428, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 773 + }, + { + "epoch": 0.3092289252896524, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 774 + }, + { + "epoch": 0.309628445864962, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3775, + "step": 775 + }, + { + "epoch": 0.3100279664402717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 776 + }, + { + "epoch": 0.3104274870155813, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 777 + }, + { + "epoch": 0.31082700759089094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 778 + }, + { + "epoch": 0.31122652816620056, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 779 + }, + { + "epoch": 0.3116260487415102, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 780 + }, + { + "epoch": 0.3120255693168198, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 781 + }, + { + "epoch": 0.31242508989212947, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 782 + }, + { + "epoch": 0.3128246104674391, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 783 + }, + { + "epoch": 0.3132241310427487, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 784 + }, + { + "epoch": 0.31362365161805833, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 785 + }, + { + "epoch": 0.31402317219336795, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 786 + }, + { + "epoch": 0.31442269276867757, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 787 + }, + { + "epoch": 0.3148222133439872, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 788 + }, + { + "epoch": 0.31522173391929686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 789 + }, + { + "epoch": 0.3156212544946065, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 790 + }, + { + "epoch": 0.3160207750699161, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 791 + }, + { + "epoch": 0.3164202956452257, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3728, + "step": 792 + }, + { + "epoch": 0.31681981622053534, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 793 + }, + { + "epoch": 0.31721933679584496, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 794 + }, + { + "epoch": 0.31761885737115464, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 795 + }, + { + "epoch": 0.31801837794646426, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 796 + }, + { + "epoch": 0.3184178985217739, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 797 + }, + { + "epoch": 0.3188174190970835, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 798 + }, + { + "epoch": 0.3192169396723931, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 799 + }, + { + "epoch": 0.31961646024770274, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 800 + }, + { + "epoch": 0.32001598082301236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 801 + }, + { + "epoch": 0.32041550139832203, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 802 + }, + { + "epoch": 0.32081502197363165, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 803 + }, + { + "epoch": 0.3212145425489413, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 804 + }, + { + "epoch": 0.3216140631242509, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 805 + }, + { + "epoch": 0.3220135836995605, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 806 + }, + { + "epoch": 0.32241310427487013, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 807 + }, + { + "epoch": 0.3228126248501798, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 808 + }, + { + "epoch": 0.3232121454254894, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 809 + }, + { + "epoch": 0.32361166600079905, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 810 + }, + { + "epoch": 0.32401118657610867, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 811 + }, + { + "epoch": 0.3244107071514183, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 812 + }, + { + "epoch": 0.3248102277267279, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 813 + }, + { + "epoch": 0.3252097483020376, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.373, + "step": 814 + }, + { + "epoch": 0.3256092688773472, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 815 + }, + { + "epoch": 0.3260087894526568, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 816 + }, + { + "epoch": 0.32640831002796644, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 817 + }, + { + "epoch": 0.32680783060327606, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 818 + }, + { + "epoch": 0.3272073511785857, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 819 + }, + { + "epoch": 0.3276068717538953, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 820 + }, + { + "epoch": 0.328006392329205, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 821 + }, + { + "epoch": 0.3284059129045146, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 822 + }, + { + "epoch": 0.3288054334798242, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 823 + }, + { + "epoch": 0.32920495405513384, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 824 + }, + { + "epoch": 0.32960447463044346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 825 + }, + { + "epoch": 0.3300039952057531, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 826 + }, + { + "epoch": 0.33040351578106275, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 827 + }, + { + "epoch": 0.33080303635637237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 828 + }, + { + "epoch": 0.331202556931682, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 829 + }, + { + "epoch": 0.3316020775069916, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 830 + }, + { + "epoch": 0.33200159808230123, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 831 + }, + { + "epoch": 0.33240111865761085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 832 + }, + { + "epoch": 0.33280063923292047, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 833 + }, + { + "epoch": 0.33320015980823015, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 834 + }, + { + "epoch": 0.33359968038353977, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 835 + }, + { + "epoch": 0.3339992009588494, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 836 + }, + { + "epoch": 0.334398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 837 + }, + { + "epoch": 0.3347982421094686, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 838 + }, + { + "epoch": 0.33519776268477824, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 839 + }, + { + "epoch": 0.3355972832600879, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 840 + }, + { + "epoch": 0.33599680383539754, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 841 + }, + { + "epoch": 0.33639632441070716, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 842 + }, + { + "epoch": 0.3367958449860168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 843 + }, + { + "epoch": 0.3371953655613264, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 844 + }, + { + "epoch": 0.337594886136636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 845 + }, + { + "epoch": 0.33799440671194564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 846 + }, + { + "epoch": 0.3383939272872553, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 847 + }, + { + "epoch": 0.33879344786256493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 848 + }, + { + "epoch": 0.33919296843787455, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 849 + }, + { + "epoch": 0.3395924890131842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 850 + }, + { + "epoch": 0.3399920095884938, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 851 + }, + { + "epoch": 0.3403915301638034, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 852 + }, + { + "epoch": 0.3407910507391131, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 853 + }, + { + "epoch": 0.3411905713144227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 854 + }, + { + "epoch": 0.34159009188973233, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 855 + }, + { + "epoch": 0.34198961246504195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 856 + }, + { + "epoch": 0.34238913304035157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 857 + }, + { + "epoch": 0.3427886536156612, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 858 + }, + { + "epoch": 0.34318817419097086, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 859 + }, + { + "epoch": 0.3435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 860 + }, + { + "epoch": 0.3439872153415901, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 861 + }, + { + "epoch": 0.3443867359168997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 862 + }, + { + "epoch": 0.34478625649220934, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 863 + }, + { + "epoch": 0.34518577706751896, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 864 + }, + { + "epoch": 0.3455852976428286, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 865 + }, + { + "epoch": 0.34598481821813826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 866 + }, + { + "epoch": 0.3463843387934479, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3776, + "step": 867 + }, + { + "epoch": 0.3467838593687575, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 868 + }, + { + "epoch": 0.3471833799440671, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 869 + }, + { + "epoch": 0.34758290051937674, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 870 + }, + { + "epoch": 0.34798242109468636, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 871 + }, + { + "epoch": 0.34838194166999603, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 872 + }, + { + "epoch": 0.34878146224530565, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 873 + }, + { + "epoch": 0.34918098282061527, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 874 + }, + { + "epoch": 0.3495805033959249, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 875 + }, + { + "epoch": 0.3499800239712345, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 876 + }, + { + "epoch": 0.35037954454654413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 877 + }, + { + "epoch": 0.35077906512185375, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 878 + }, + { + "epoch": 0.3511785856971634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 879 + }, + { + "epoch": 0.35157810627247305, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 880 + }, + { + "epoch": 0.35197762684778267, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 881 + }, + { + "epoch": 0.3523771474230923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 882 + }, + { + "epoch": 0.3527766679984019, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 883 + }, + { + "epoch": 0.3531761885737115, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 884 + }, + { + "epoch": 0.3535757091490212, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 885 + }, + { + "epoch": 0.3539752297243308, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 886 + }, + { + "epoch": 0.35437475029964044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 887 + }, + { + "epoch": 0.35477427087495006, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3751, + "step": 888 + }, + { + "epoch": 0.3551737914502597, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 889 + }, + { + "epoch": 0.3555733120255693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 890 + }, + { + "epoch": 0.3559728326008789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 891 + }, + { + "epoch": 0.3563723531761886, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 892 + }, + { + "epoch": 0.3567718737514982, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 893 + }, + { + "epoch": 0.35717139432680783, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 894 + }, + { + "epoch": 0.35757091490211745, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 895 + }, + { + "epoch": 0.3579704354774271, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3758, + "step": 896 + }, + { + "epoch": 0.3583699560527367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 897 + }, + { + "epoch": 0.35876947662804637, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 898 + }, + { + "epoch": 0.359168997203356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 899 + }, + { + "epoch": 0.3595685177786656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 900 + }, + { + "epoch": 0.35996803835397523, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 901 + }, + { + "epoch": 0.36036755892928485, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3774, + "step": 902 + }, + { + "epoch": 0.36076707950459447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 903 + }, + { + "epoch": 0.3611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 904 + }, + { + "epoch": 0.36156612065521376, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 905 + }, + { + "epoch": 0.3619656412305234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 906 + }, + { + "epoch": 0.362365161805833, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 907 + }, + { + "epoch": 0.3627646823811426, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 908 + }, + { + "epoch": 0.36316420295645224, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 909 + }, + { + "epoch": 0.36356372353176186, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 910 + }, + { + "epoch": 0.36396324410707154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 911 + }, + { + "epoch": 0.36436276468238116, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 912 + }, + { + "epoch": 0.3647622852576908, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 913 + }, + { + "epoch": 0.3651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 914 + }, + { + "epoch": 0.36556132640831, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 915 + }, + { + "epoch": 0.36596084698361964, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 916 + }, + { + "epoch": 0.3663603675589293, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 917 + }, + { + "epoch": 0.36675988813423893, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 918 + }, + { + "epoch": 0.36715940870954855, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 919 + }, + { + "epoch": 0.36755892928485817, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 920 + }, + { + "epoch": 0.3679584498601678, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 921 + }, + { + "epoch": 0.3683579704354774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 922 + }, + { + "epoch": 0.36875749101078703, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 923 + }, + { + "epoch": 0.3691570115860967, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 924 + }, + { + "epoch": 0.3695565321614063, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 925 + }, + { + "epoch": 0.36995605273671595, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 926 + }, + { + "epoch": 0.37035557331202557, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 927 + }, + { + "epoch": 0.3707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 928 + }, + { + "epoch": 0.3711546144626448, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 929 + }, + { + "epoch": 0.3715541350379545, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 930 + }, + { + "epoch": 0.3719536556132641, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 931 + }, + { + "epoch": 0.3723531761885737, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 932 + }, + { + "epoch": 0.37275269676388334, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 933 + }, + { + "epoch": 0.37315221733919296, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 934 + }, + { + "epoch": 0.3735517379145026, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 935 + }, + { + "epoch": 0.3739512584898122, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 936 + }, + { + "epoch": 0.3743507790651219, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 937 + }, + { + "epoch": 0.3747502996404315, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 938 + }, + { + "epoch": 0.3751498202157411, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 939 + }, + { + "epoch": 0.37554934079105073, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 940 + }, + { + "epoch": 0.37594886136636035, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 941 + }, + { + "epoch": 0.37634838194167, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 942 + }, + { + "epoch": 0.37674790251697965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 943 + }, + { + "epoch": 0.37714742309228927, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 944 + }, + { + "epoch": 0.3775469436675989, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 945 + }, + { + "epoch": 0.3779464642429085, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 946 + }, + { + "epoch": 0.37834598481821813, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 947 + }, + { + "epoch": 0.37874550539352775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 948 + }, + { + "epoch": 0.37914502596883737, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 949 + }, + { + "epoch": 0.37954454654414704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 950 + }, + { + "epoch": 0.37994406711945666, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 951 + }, + { + "epoch": 0.3803435876947663, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 952 + }, + { + "epoch": 0.3807431082700759, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 953 + }, + { + "epoch": 0.3811426288453855, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 954 + }, + { + "epoch": 0.38154214942069514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 955 + }, + { + "epoch": 0.3819416699960048, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 956 + }, + { + "epoch": 0.38234119057131444, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 957 + }, + { + "epoch": 0.38274071114662406, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 958 + }, + { + "epoch": 0.3831402317219337, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 959 + }, + { + "epoch": 0.3835397522972433, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 960 + }, + { + "epoch": 0.3839392728725529, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 961 + }, + { + "epoch": 0.3843387934478626, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3661, + "step": 962 + }, + { + "epoch": 0.3847383140231722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 963 + }, + { + "epoch": 0.38513783459848183, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 964 + }, + { + "epoch": 0.38553735517379145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 965 + }, + { + "epoch": 0.38593687574910107, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 966 + }, + { + "epoch": 0.3863363963244107, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 967 + }, + { + "epoch": 0.3867359168997203, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 968 + }, + { + "epoch": 0.38713543747503, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 969 + }, + { + "epoch": 0.3875349580503396, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 970 + }, + { + "epoch": 0.3879344786256492, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 971 + }, + { + "epoch": 0.38833399920095885, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 972 + }, + { + "epoch": 0.38873351977626847, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 973 + }, + { + "epoch": 0.3891330403515781, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 974 + }, + { + "epoch": 0.38953256092688776, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 975 + }, + { + "epoch": 0.3899320815021974, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 976 + }, + { + "epoch": 0.390331602077507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 977 + }, + { + "epoch": 0.3907311226528166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 978 + }, + { + "epoch": 0.39113064322812624, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 979 + }, + { + "epoch": 0.39153016380343586, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 980 + }, + { + "epoch": 0.3919296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 981 + }, + { + "epoch": 0.39232920495405516, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 982 + }, + { + "epoch": 0.3927287255293648, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 983 + }, + { + "epoch": 0.3931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 984 + }, + { + "epoch": 0.393527766679984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 985 + }, + { + "epoch": 0.39392728725529363, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 986 + }, + { + "epoch": 0.39432680783060325, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 987 + }, + { + "epoch": 0.39472632840591293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 988 + }, + { + "epoch": 0.39512584898122255, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 989 + }, + { + "epoch": 0.39552536955653217, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 990 + }, + { + "epoch": 0.3959248901318418, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 991 + }, + { + "epoch": 0.3963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 992 + }, + { + "epoch": 0.39672393128246103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 993 + }, + { + "epoch": 0.39712345185777065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 994 + }, + { + "epoch": 0.3975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 995 + }, + { + "epoch": 0.39792249300838994, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 996 + }, + { + "epoch": 0.39832201358369956, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 997 + }, + { + "epoch": 0.3987215341590092, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 998 + }, + { + "epoch": 0.3991210547343188, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 999 + }, + { + "epoch": 0.3995205753096284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1000 + }, + { + "epoch": 0.3999200958849381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1001 + }, + { + "epoch": 0.4003196164602477, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1002 + }, + { + "epoch": 0.40071913703555734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1003 + }, + { + "epoch": 0.40111865761086696, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1004 + }, + { + "epoch": 0.4015181781861766, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 1005 + }, + { + "epoch": 0.4019176987614862, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1006 + }, + { + "epoch": 0.4023172193367959, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1007 + }, + { + "epoch": 0.4027167399121055, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1008 + }, + { + "epoch": 0.4031162604874151, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1009 + }, + { + "epoch": 0.40351578106272473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1010 + }, + { + "epoch": 0.40391530163803435, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1011 + }, + { + "epoch": 0.40431482221334397, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1012 + }, + { + "epoch": 0.4047143427886536, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 1013 + }, + { + "epoch": 0.40511386336396327, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1014 + }, + { + "epoch": 0.4055133839392729, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1015 + }, + { + "epoch": 0.4059129045145825, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1016 + }, + { + "epoch": 0.4063124250898921, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1017 + }, + { + "epoch": 0.40671194566520175, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 1018 + }, + { + "epoch": 0.40711146624051137, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1019 + }, + { + "epoch": 0.40751098681582104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1020 + }, + { + "epoch": 0.40791050739113066, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1021 + }, + { + "epoch": 0.4083100279664403, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1022 + }, + { + "epoch": 0.4087095485417499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1023 + }, + { + "epoch": 0.4091090691170595, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 1024 + }, + { + "epoch": 0.40950858969236914, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1025 + }, + { + "epoch": 0.40990811026767876, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1026 + }, + { + "epoch": 0.41030763084298844, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1027 + }, + { + "epoch": 0.41070715141829806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1028 + }, + { + "epoch": 0.4111066719936077, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1029 + }, + { + "epoch": 0.4115061925689173, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1030 + }, + { + "epoch": 0.4119057131442269, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1031 + }, + { + "epoch": 0.41230523371953653, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1032 + }, + { + "epoch": 0.4127047542948462, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1033 + }, + { + "epoch": 0.41310427487015583, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1034 + }, + { + "epoch": 0.41350379544546545, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1035 + }, + { + "epoch": 0.41390331602077507, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1036 + }, + { + "epoch": 0.4143028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1037 + }, + { + "epoch": 0.4147023571713943, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1038 + }, + { + "epoch": 0.41510187774670393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1039 + }, + { + "epoch": 0.4155013983220136, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1040 + }, + { + "epoch": 0.4159009188973232, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1041 + }, + { + "epoch": 0.41630043947263284, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1042 + }, + { + "epoch": 0.41669996004794246, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1043 + }, + { + "epoch": 0.4170994806232521, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1044 + }, + { + "epoch": 0.4174990011985617, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1045 + }, + { + "epoch": 0.4178985217738714, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 1046 + }, + { + "epoch": 0.418298042349181, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1047 + }, + { + "epoch": 0.4186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1048 + }, + { + "epoch": 0.41909708349980024, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3711, + "step": 1049 + }, + { + "epoch": 0.41949660407510986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1050 + }, + { + "epoch": 0.4198961246504195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1051 + }, + { + "epoch": 0.4202956452257291, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1052 + }, + { + "epoch": 0.4206951658010388, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1053 + }, + { + "epoch": 0.4210946863763484, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1054 + }, + { + "epoch": 0.421494206951658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1055 + }, + { + "epoch": 0.42189372752696763, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1056 + }, + { + "epoch": 0.42229324810227725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 1057 + }, + { + "epoch": 0.42269276867758687, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1058 + }, + { + "epoch": 0.42309228925289655, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 1059 + }, + { + "epoch": 0.42349180982820617, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1060 + }, + { + "epoch": 0.4238913304035158, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1061 + }, + { + "epoch": 0.4242908509788254, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1062 + }, + { + "epoch": 0.424690371554135, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1063 + }, + { + "epoch": 0.42508989212944465, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1064 + }, + { + "epoch": 0.4254894127047543, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 1065 + }, + { + "epoch": 0.42588893328006394, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1066 + }, + { + "epoch": 0.42628845385537356, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1067 + }, + { + "epoch": 0.4266879744306832, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 1068 + }, + { + "epoch": 0.4270874950059928, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1069 + }, + { + "epoch": 0.4274870155813024, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1070 + }, + { + "epoch": 0.42788653615661204, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1071 + }, + { + "epoch": 0.4282860567319217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1072 + }, + { + "epoch": 0.42868557730723134, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1073 + }, + { + "epoch": 0.42908509788254096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1074 + }, + { + "epoch": 0.4294846184578506, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1075 + }, + { + "epoch": 0.4298841390331602, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1076 + }, + { + "epoch": 0.4302836596084698, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1077 + }, + { + "epoch": 0.4306831801837795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1078 + }, + { + "epoch": 0.4310827007590891, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1079 + }, + { + "epoch": 0.43148222133439873, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1080 + }, + { + "epoch": 0.43188174190970835, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1081 + }, + { + "epoch": 0.43228126248501797, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1082 + }, + { + "epoch": 0.4326807830603276, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 1083 + }, + { + "epoch": 0.4330803036356372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1084 + }, + { + "epoch": 0.4334798242109469, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1085 + }, + { + "epoch": 0.4338793447862565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1086 + }, + { + "epoch": 0.4342788653615661, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 1087 + }, + { + "epoch": 0.43467838593687574, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1088 + }, + { + "epoch": 0.43507790651218536, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1089 + }, + { + "epoch": 0.435477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1090 + }, + { + "epoch": 0.43587694766280466, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1091 + }, + { + "epoch": 0.4362764682381143, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1092 + }, + { + "epoch": 0.4366759888134239, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1093 + }, + { + "epoch": 0.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1094 + }, + { + "epoch": 0.43747502996404314, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1095 + }, + { + "epoch": 0.43787455053935276, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1096 + }, + { + "epoch": 0.4382740711146624, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1097 + }, + { + "epoch": 0.43867359168997205, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1098 + }, + { + "epoch": 0.4390731122652817, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1099 + }, + { + "epoch": 0.4394726328405913, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1100 + }, + { + "epoch": 0.4398721534159009, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1101 + }, + { + "epoch": 0.44027167399121053, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1102 + }, + { + "epoch": 0.44067119456652015, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 1103 + }, + { + "epoch": 0.4410707151418298, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1104 + }, + { + "epoch": 0.44147023571713945, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1105 + }, + { + "epoch": 0.44186975629244907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1106 + }, + { + "epoch": 0.4422692768677587, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1107 + }, + { + "epoch": 0.4426687974430683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1108 + }, + { + "epoch": 0.4430683180183779, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1109 + }, + { + "epoch": 0.4434678385936876, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 1110 + }, + { + "epoch": 0.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1111 + }, + { + "epoch": 0.44426687974430684, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1112 + }, + { + "epoch": 0.44466640031961646, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1113 + }, + { + "epoch": 0.4450659208949261, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1114 + }, + { + "epoch": 0.4454654414702357, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1115 + }, + { + "epoch": 0.4458649620455453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1116 + }, + { + "epoch": 0.446264482620855, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 1117 + }, + { + "epoch": 0.4466640031961646, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1118 + }, + { + "epoch": 0.44706352377147424, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1119 + }, + { + "epoch": 0.44746304434678386, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1120 + }, + { + "epoch": 0.4478625649220935, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1121 + }, + { + "epoch": 0.4482620854974031, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1122 + }, + { + "epoch": 0.44866160607271277, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1123 + }, + { + "epoch": 0.4490611266480224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1124 + }, + { + "epoch": 0.449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 1125 + }, + { + "epoch": 0.44986016779864163, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1126 + }, + { + "epoch": 0.45025968837395125, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1127 + }, + { + "epoch": 0.45065920894926087, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1128 + }, + { + "epoch": 0.4510587295245705, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1129 + }, + { + "epoch": 0.45145825009988017, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1130 + }, + { + "epoch": 0.4518577706751898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1131 + }, + { + "epoch": 0.4522572912504994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1132 + }, + { + "epoch": 0.452656811825809, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 1133 + }, + { + "epoch": 0.45305633240111864, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1134 + }, + { + "epoch": 0.45345585297642826, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1135 + }, + { + "epoch": 0.45385537355173794, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1136 + }, + { + "epoch": 0.45425489412704756, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1137 + }, + { + "epoch": 0.4546544147023572, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1138 + }, + { + "epoch": 0.4550539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 1139 + }, + { + "epoch": 0.4554534558529764, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1140 + }, + { + "epoch": 0.45585297642828604, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 1141 + }, + { + "epoch": 0.45625249700359566, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1142 + }, + { + "epoch": 0.45665201757890533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1143 + }, + { + "epoch": 0.45705153815421495, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1144 + }, + { + "epoch": 0.4574510587295246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1145 + }, + { + "epoch": 0.4578505793048342, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1146 + }, + { + "epoch": 0.4582500998801438, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1147 + }, + { + "epoch": 0.45864962045545343, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1148 + }, + { + "epoch": 0.4590491410307631, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1149 + }, + { + "epoch": 0.45944866160607273, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1150 + }, + { + "epoch": 0.45984818218138235, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3724, + "step": 1151 + }, + { + "epoch": 0.46024770275669197, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1152 + }, + { + "epoch": 0.4606472233320016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1153 + }, + { + "epoch": 0.4610467439073112, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1154 + }, + { + "epoch": 0.4614462644826209, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1155 + }, + { + "epoch": 0.4618457850579305, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1156 + }, + { + "epoch": 0.4622453056332401, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1157 + }, + { + "epoch": 0.46264482620854974, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1158 + }, + { + "epoch": 0.46304434678385936, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1159 + }, + { + "epoch": 0.463443867359169, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 1160 + }, + { + "epoch": 0.4638433879344786, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1161 + }, + { + "epoch": 0.4642429085097883, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1162 + }, + { + "epoch": 0.4646424290850979, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1163 + }, + { + "epoch": 0.4650419496604075, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 1164 + }, + { + "epoch": 0.46544147023571714, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1165 + }, + { + "epoch": 0.46584099081102676, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1166 + }, + { + "epoch": 0.4662405113863364, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1167 + }, + { + "epoch": 0.46664003196164605, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1168 + }, + { + "epoch": 0.46703955253695567, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1169 + }, + { + "epoch": 0.4674390731122653, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1170 + }, + { + "epoch": 0.4678385936875749, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1171 + }, + { + "epoch": 0.46823811426288453, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1172 + }, + { + "epoch": 0.46863763483819415, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1173 + }, + { + "epoch": 0.46903715541350377, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1174 + }, + { + "epoch": 0.46943667598881345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1175 + }, + { + "epoch": 0.46983619656412307, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1176 + }, + { + "epoch": 0.4702357171394327, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1177 + }, + { + "epoch": 0.4706352377147423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 1178 + }, + { + "epoch": 0.4710347582900519, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1179 + }, + { + "epoch": 0.47143427886536154, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1180 + }, + { + "epoch": 0.4718337994406712, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 1181 + }, + { + "epoch": 0.47223332001598084, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1182 + }, + { + "epoch": 0.47263284059129046, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1183 + }, + { + "epoch": 0.4730323611666001, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1184 + }, + { + "epoch": 0.4734318817419097, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1185 + }, + { + "epoch": 0.4738314023172193, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1186 + }, + { + "epoch": 0.47423092289252894, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1187 + }, + { + "epoch": 0.4746304434678386, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1188 + }, + { + "epoch": 0.47502996404314823, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1189 + }, + { + "epoch": 0.47542948461845785, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 1190 + }, + { + "epoch": 0.4758290051937675, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1191 + }, + { + "epoch": 0.4762285257690771, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1192 + }, + { + "epoch": 0.4766280463443867, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1193 + }, + { + "epoch": 0.4770275669196964, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1194 + }, + { + "epoch": 0.477427087495006, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1195 + }, + { + "epoch": 0.47782660807031563, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1196 + }, + { + "epoch": 0.47822612864562525, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1197 + }, + { + "epoch": 0.47862564922093487, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1198 + }, + { + "epoch": 0.4790251697962445, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1199 + }, + { + "epoch": 0.4794246903715541, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1200 + }, + { + "epoch": 0.4798242109468638, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1201 + }, + { + "epoch": 0.4802237315221734, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1202 + }, + { + "epoch": 0.480623252097483, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1203 + }, + { + "epoch": 0.48102277267279264, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1204 + }, + { + "epoch": 0.48142229324810226, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1205 + }, + { + "epoch": 0.4818218138234119, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1206 + }, + { + "epoch": 0.48222133439872156, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1207 + }, + { + "epoch": 0.4826208549740312, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1208 + }, + { + "epoch": 0.4830203755493408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1209 + }, + { + "epoch": 0.4834198961246504, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1210 + }, + { + "epoch": 0.48381941669996004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1211 + }, + { + "epoch": 0.48421893727526966, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1212 + }, + { + "epoch": 0.48461845785057933, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1213 + }, + { + "epoch": 0.48501797842588895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1214 + }, + { + "epoch": 0.48541749900119857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1215 + }, + { + "epoch": 0.4858170195765082, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1216 + }, + { + "epoch": 0.4862165401518178, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1217 + }, + { + "epoch": 0.48661606072712743, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 1218 + }, + { + "epoch": 0.48701558130243705, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1219 + }, + { + "epoch": 0.4874151018777467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1220 + }, + { + "epoch": 0.48781462245305635, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1221 + }, + { + "epoch": 0.48821414302836597, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1222 + }, + { + "epoch": 0.4886136636036756, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1223 + }, + { + "epoch": 0.4890131841789852, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1224 + }, + { + "epoch": 0.4894127047542948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1225 + }, + { + "epoch": 0.4898122253296045, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1226 + }, + { + "epoch": 0.4902117459049141, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1227 + }, + { + "epoch": 0.49061126648022374, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1228 + }, + { + "epoch": 0.49101078705553336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1229 + }, + { + "epoch": 0.491410307630843, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 1230 + }, + { + "epoch": 0.4918098282061526, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1231 + }, + { + "epoch": 0.4922093487814622, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1232 + }, + { + "epoch": 0.4926088693567719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1233 + }, + { + "epoch": 0.4930083899320815, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1234 + }, + { + "epoch": 0.49340791050739113, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1235 + }, + { + "epoch": 0.49380743108270075, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1236 + }, + { + "epoch": 0.4942069516580104, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 1237 + }, + { + "epoch": 0.49460647223332, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1238 + }, + { + "epoch": 0.49500599280862967, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1239 + }, + { + "epoch": 0.4954055133839393, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1240 + }, + { + "epoch": 0.4958050339592489, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1241 + }, + { + "epoch": 0.49620455453455853, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1242 + }, + { + "epoch": 0.49660407510986815, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1243 + }, + { + "epoch": 0.49700359568517777, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1244 + }, + { + "epoch": 0.4974031162604874, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1245 + }, + { + "epoch": 0.49780263683579706, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1246 + }, + { + "epoch": 0.4982021574111067, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1247 + }, + { + "epoch": 0.4986016779864163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1248 + }, + { + "epoch": 0.4990011985617259, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1249 + }, + { + "epoch": 0.49940071913703554, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1250 + }, + { + "epoch": 0.49980023971234516, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1251 + }, + { + "epoch": 0.5001997602876548, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1252 + }, + { + "epoch": 0.5005992808629645, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1253 + }, + { + "epoch": 0.500998801438274, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1254 + }, + { + "epoch": 0.5013983220135837, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1255 + }, + { + "epoch": 0.5017978425888934, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1256 + }, + { + "epoch": 0.5021973631642029, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1257 + }, + { + "epoch": 0.5025968837395126, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1258 + }, + { + "epoch": 0.5029964043148222, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1259 + }, + { + "epoch": 0.5033959248901319, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1260 + }, + { + "epoch": 0.5037954454654414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1261 + }, + { + "epoch": 0.5041949660407511, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1262 + }, + { + "epoch": 0.5045944866160608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 1263 + }, + { + "epoch": 0.5049940071913703, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1264 + }, + { + "epoch": 0.50539352776668, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1265 + }, + { + "epoch": 0.5057930483419896, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1266 + }, + { + "epoch": 0.5061925689172992, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1267 + }, + { + "epoch": 0.5065920894926089, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3678, + "step": 1268 + }, + { + "epoch": 0.5069916100679185, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.365, + "step": 1269 + }, + { + "epoch": 0.5073911306432282, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1270 + }, + { + "epoch": 0.5077906512185377, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1271 + }, + { + "epoch": 0.5081901717938474, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1272 + }, + { + "epoch": 0.508589692369157, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1273 + }, + { + "epoch": 0.5089892129444666, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1274 + }, + { + "epoch": 0.5093887335197763, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1275 + }, + { + "epoch": 0.5097882540950859, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1276 + }, + { + "epoch": 0.5101877746703956, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1277 + }, + { + "epoch": 0.5105872952457051, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 1278 + }, + { + "epoch": 0.5109868158210148, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1279 + }, + { + "epoch": 0.5113863363963244, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1280 + }, + { + "epoch": 0.511785856971634, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1281 + }, + { + "epoch": 0.5121853775469437, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1282 + }, + { + "epoch": 0.5125848981222533, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1283 + }, + { + "epoch": 0.512984418697563, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1284 + }, + { + "epoch": 0.5133839392728725, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1285 + }, + { + "epoch": 0.5137834598481822, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.368, + "step": 1286 + }, + { + "epoch": 0.5141829804234919, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1287 + }, + { + "epoch": 0.5145825009988014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1288 + }, + { + "epoch": 0.5149820215741111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1289 + }, + { + "epoch": 0.5153815421494207, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1290 + }, + { + "epoch": 0.5157810627247303, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1291 + }, + { + "epoch": 0.5161805833000399, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 1292 + }, + { + "epoch": 0.5165801038753496, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1293 + }, + { + "epoch": 0.5169796244506593, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1294 + }, + { + "epoch": 0.5173791450259688, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1295 + }, + { + "epoch": 0.5177786656012785, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 1296 + }, + { + "epoch": 0.5181781861765881, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1297 + }, + { + "epoch": 0.5185777067518977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1298 + }, + { + "epoch": 0.5189772273272073, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1299 + }, + { + "epoch": 0.519376747902517, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1300 + }, + { + "epoch": 0.5197762684778267, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1301 + }, + { + "epoch": 0.5201757890531362, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1302 + }, + { + "epoch": 0.5205753096284459, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1303 + }, + { + "epoch": 0.5209748302037555, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1304 + }, + { + "epoch": 0.5213743507790651, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1305 + }, + { + "epoch": 0.5217738713543747, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1306 + }, + { + "epoch": 0.5221733919296844, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1307 + }, + { + "epoch": 0.522572912504994, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1308 + }, + { + "epoch": 0.5229724330803036, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1309 + }, + { + "epoch": 0.5233719536556133, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1310 + }, + { + "epoch": 0.5237714742309229, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1311 + }, + { + "epoch": 0.5241709948062325, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1312 + }, + { + "epoch": 0.5245705153815422, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1313 + }, + { + "epoch": 0.5249700359568518, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1314 + }, + { + "epoch": 0.5253695565321614, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1315 + }, + { + "epoch": 0.525769077107471, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1316 + }, + { + "epoch": 0.5261685976827807, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1317 + }, + { + "epoch": 0.5265681182580902, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1318 + }, + { + "epoch": 0.5269676388333999, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 1319 + }, + { + "epoch": 0.5273671594087096, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1320 + }, + { + "epoch": 0.5277666799840192, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1321 + }, + { + "epoch": 0.5281662005593288, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1322 + }, + { + "epoch": 0.5285657211346384, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1323 + }, + { + "epoch": 0.5289652417099481, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1324 + }, + { + "epoch": 0.5293647622852576, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1325 + }, + { + "epoch": 0.5297642828605673, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1326 + }, + { + "epoch": 0.530163803435877, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1327 + }, + { + "epoch": 0.5305633240111866, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 1328 + }, + { + "epoch": 0.5309628445864962, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1329 + }, + { + "epoch": 0.5313623651618058, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1330 + }, + { + "epoch": 0.5317618857371155, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1331 + }, + { + "epoch": 0.5321614063124251, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1332 + }, + { + "epoch": 0.5325609268877347, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1333 + }, + { + "epoch": 0.5329604474630444, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1334 + }, + { + "epoch": 0.533359968038354, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1335 + }, + { + "epoch": 0.5337594886136636, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1336 + }, + { + "epoch": 0.5341590091889732, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1337 + }, + { + "epoch": 0.5345585297642829, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1338 + }, + { + "epoch": 0.5349580503395925, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1339 + }, + { + "epoch": 0.5353575709149021, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1340 + }, + { + "epoch": 0.5357570914902118, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1341 + }, + { + "epoch": 0.5361566120655213, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1342 + }, + { + "epoch": 0.536556132640831, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1343 + }, + { + "epoch": 0.5369556532161406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1344 + }, + { + "epoch": 0.5373551737914503, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1345 + }, + { + "epoch": 0.5377546943667599, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1346 + }, + { + "epoch": 0.5381542149420695, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1347 + }, + { + "epoch": 0.5385537355173792, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1348 + }, + { + "epoch": 0.5389532560926887, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 1349 + }, + { + "epoch": 0.5393527766679984, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1350 + }, + { + "epoch": 0.539752297243308, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1351 + }, + { + "epoch": 0.5401518178186177, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1352 + }, + { + "epoch": 0.5405513383939273, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1353 + }, + { + "epoch": 0.5409508589692369, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 1354 + }, + { + "epoch": 0.5413503795445466, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1355 + }, + { + "epoch": 0.5417499001198561, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1356 + }, + { + "epoch": 0.5421494206951658, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1357 + }, + { + "epoch": 0.5425489412704755, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1358 + }, + { + "epoch": 0.542948461845785, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1359 + }, + { + "epoch": 0.5433479824210947, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1360 + }, + { + "epoch": 0.5437475029964043, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1361 + }, + { + "epoch": 0.544147023571714, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1362 + }, + { + "epoch": 0.5445465441470235, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1363 + }, + { + "epoch": 0.5449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1364 + }, + { + "epoch": 0.5453455852976429, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1365 + }, + { + "epoch": 0.5457451058729524, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1366 + }, + { + "epoch": 0.5461446264482621, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1367 + }, + { + "epoch": 0.5465441470235717, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 1368 + }, + { + "epoch": 0.5469436675988814, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1369 + }, + { + "epoch": 0.5473431881741909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1370 + }, + { + "epoch": 0.5477427087495006, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1371 + }, + { + "epoch": 0.5481422293248103, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 1372 + }, + { + "epoch": 0.5485417499001198, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1373 + }, + { + "epoch": 0.5489412704754295, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1374 + }, + { + "epoch": 0.5493407910507391, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1375 + }, + { + "epoch": 0.5497403116260487, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1376 + }, + { + "epoch": 0.5501398322013584, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1377 + }, + { + "epoch": 0.550539352776668, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1378 + }, + { + "epoch": 0.5509388733519777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1379 + }, + { + "epoch": 0.5513383939272872, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1380 + }, + { + "epoch": 0.5517379145025969, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1381 + }, + { + "epoch": 0.5521374350779065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1382 + }, + { + "epoch": 0.5525369556532161, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1383 + }, + { + "epoch": 0.5529364762285258, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1384 + }, + { + "epoch": 0.5533359968038354, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1385 + }, + { + "epoch": 0.5537355173791451, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1386 + }, + { + "epoch": 0.5541350379544546, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1387 + }, + { + "epoch": 0.5545345585297643, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1388 + }, + { + "epoch": 0.5549340791050739, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1389 + }, + { + "epoch": 0.5553335996803835, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1390 + }, + { + "epoch": 0.5557331202556932, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1391 + }, + { + "epoch": 0.5561326408310028, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1392 + }, + { + "epoch": 0.5565321614063125, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1393 + }, + { + "epoch": 0.556931681981622, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 1394 + }, + { + "epoch": 0.5573312025569317, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1395 + }, + { + "epoch": 0.5577307231322413, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1396 + }, + { + "epoch": 0.5581302437075509, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1397 + }, + { + "epoch": 0.5585297642828606, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1398 + }, + { + "epoch": 0.5589292848581702, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1399 + }, + { + "epoch": 0.5593288054334798, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1400 + }, + { + "epoch": 0.5597283260087894, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1401 + }, + { + "epoch": 0.5601278465840991, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1402 + }, + { + "epoch": 0.5605273671594088, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1403 + }, + { + "epoch": 0.5609268877347183, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1404 + }, + { + "epoch": 0.561326408310028, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1405 + }, + { + "epoch": 0.5617259288853376, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1406 + }, + { + "epoch": 0.5621254494606472, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 1407 + }, + { + "epoch": 0.5625249700359568, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1408 + }, + { + "epoch": 0.5629244906112665, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1409 + }, + { + "epoch": 0.5633240111865762, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1410 + }, + { + "epoch": 0.5637235317618857, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1411 + }, + { + "epoch": 0.5641230523371954, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1412 + }, + { + "epoch": 0.564522572912505, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1413 + }, + { + "epoch": 0.5649220934878146, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1414 + }, + { + "epoch": 0.5653216140631242, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1415 + }, + { + "epoch": 0.5657211346384339, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1416 + }, + { + "epoch": 0.5661206552137436, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1417 + }, + { + "epoch": 0.5665201757890531, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1418 + }, + { + "epoch": 0.5669196963643628, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1419 + }, + { + "epoch": 0.5673192169396724, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1420 + }, + { + "epoch": 0.567718737514982, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1421 + }, + { + "epoch": 0.5681182580902916, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1422 + }, + { + "epoch": 0.5685177786656013, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1423 + }, + { + "epoch": 0.568917299240911, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1424 + }, + { + "epoch": 0.5693168198162205, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1425 + }, + { + "epoch": 0.5697163403915302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1426 + }, + { + "epoch": 0.5701158609668397, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1427 + }, + { + "epoch": 0.5705153815421494, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1428 + }, + { + "epoch": 0.5709149021174591, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1429 + }, + { + "epoch": 0.5713144226927687, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1430 + }, + { + "epoch": 0.5717139432680783, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1431 + }, + { + "epoch": 0.5721134638433879, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1432 + }, + { + "epoch": 0.5725129844186976, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1433 + }, + { + "epoch": 0.5729125049940071, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1434 + }, + { + "epoch": 0.5733120255693168, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1435 + }, + { + "epoch": 0.5737115461446265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1436 + }, + { + "epoch": 0.5741110667199361, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1437 + }, + { + "epoch": 0.5745105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1438 + }, + { + "epoch": 0.5749101078705553, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1439 + }, + { + "epoch": 0.575309628445865, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1440 + }, + { + "epoch": 0.5757091490211745, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1441 + }, + { + "epoch": 0.5761086695964842, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1442 + }, + { + "epoch": 0.5765081901717939, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1443 + }, + { + "epoch": 0.5769077107471035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1444 + }, + { + "epoch": 0.5773072313224131, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1445 + }, + { + "epoch": 0.5777067518977227, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1446 + }, + { + "epoch": 0.5781062724730324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1447 + }, + { + "epoch": 0.578505793048342, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1448 + }, + { + "epoch": 0.5789053136236516, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1449 + }, + { + "epoch": 0.5793048341989613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 1450 + }, + { + "epoch": 0.5797043547742708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1451 + }, + { + "epoch": 0.5801038753495805, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1452 + }, + { + "epoch": 0.5805033959248901, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1453 + }, + { + "epoch": 0.5809029165001998, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1454 + }, + { + "epoch": 0.5813024370755094, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1455 + }, + { + "epoch": 0.581701957650819, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1456 + }, + { + "epoch": 0.5821014782261287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1457 + }, + { + "epoch": 0.5825009988014382, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1458 + }, + { + "epoch": 0.5829005193767479, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1459 + }, + { + "epoch": 0.5833000399520575, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1460 + }, + { + "epoch": 0.5836995605273672, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1461 + }, + { + "epoch": 0.5840990811026768, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1462 + }, + { + "epoch": 0.5844986016779864, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1463 + }, + { + "epoch": 0.5848981222532961, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1464 + }, + { + "epoch": 0.5852976428286056, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1465 + }, + { + "epoch": 0.5856971634039153, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1466 + }, + { + "epoch": 0.5860966839792249, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1467 + }, + { + "epoch": 0.5864962045545346, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1468 + }, + { + "epoch": 0.5868957251298442, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1469 + }, + { + "epoch": 0.5872952457051538, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1470 + }, + { + "epoch": 0.5876947662804635, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1471 + }, + { + "epoch": 0.588094286855773, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1472 + }, + { + "epoch": 0.5884938074310827, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1473 + }, + { + "epoch": 0.5888933280063924, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1474 + }, + { + "epoch": 0.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1475 + }, + { + "epoch": 0.5896923691570116, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1476 + }, + { + "epoch": 0.5900918897323212, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1477 + }, + { + "epoch": 0.5904914103076309, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1478 + }, + { + "epoch": 0.5908909308829404, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1479 + }, + { + "epoch": 0.5912904514582501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1480 + }, + { + "epoch": 0.5916899720335598, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1481 + }, + { + "epoch": 0.5920894926088693, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1482 + }, + { + "epoch": 0.592489013184179, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1483 + }, + { + "epoch": 0.5928885337594886, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1484 + }, + { + "epoch": 0.5932880543347983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1485 + }, + { + "epoch": 0.5936875749101078, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1486 + }, + { + "epoch": 0.5940870954854175, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1487 + }, + { + "epoch": 0.5944866160607272, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1488 + }, + { + "epoch": 0.5948861366360367, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1489 + }, + { + "epoch": 0.5952856572113464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1490 + }, + { + "epoch": 0.595685177786656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1491 + }, + { + "epoch": 0.5960846983619656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1492 + }, + { + "epoch": 0.5964842189372753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1493 + }, + { + "epoch": 0.5968837395125849, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1494 + }, + { + "epoch": 0.5972832600878946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1495 + }, + { + "epoch": 0.5976827806632041, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1496 + }, + { + "epoch": 0.5980823012385138, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1497 + }, + { + "epoch": 0.5984818218138234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 1498 + }, + { + "epoch": 0.598881342389133, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1499 + }, + { + "epoch": 0.5992808629644427, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1500 + }, + { + "epoch": 0.5996803835397523, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1501 + }, + { + "epoch": 0.600079904115062, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1502 + }, + { + "epoch": 0.6004794246903715, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1503 + }, + { + "epoch": 0.6008789452656812, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1504 + }, + { + "epoch": 0.6012784658409908, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1505 + }, + { + "epoch": 0.6016779864163004, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1506 + }, + { + "epoch": 0.6020775069916101, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1507 + }, + { + "epoch": 0.6024770275669197, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1508 + }, + { + "epoch": 0.6028765481422294, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1509 + }, + { + "epoch": 0.6032760687175389, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1510 + }, + { + "epoch": 0.6036755892928486, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1511 + }, + { + "epoch": 0.6040751098681582, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1512 + }, + { + "epoch": 0.6044746304434678, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3669, + "step": 1513 + }, + { + "epoch": 0.6048741510187775, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1514 + }, + { + "epoch": 0.6052736715940871, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1515 + }, + { + "epoch": 0.6056731921693967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1516 + }, + { + "epoch": 0.6060727127447063, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1517 + }, + { + "epoch": 0.606472233320016, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1518 + }, + { + "epoch": 0.6068717538953257, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1519 + }, + { + "epoch": 0.6072712744706352, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1520 + }, + { + "epoch": 0.6076707950459449, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1521 + }, + { + "epoch": 0.6080703156212545, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1522 + }, + { + "epoch": 0.6084698361965641, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1523 + }, + { + "epoch": 0.6088693567718737, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1524 + }, + { + "epoch": 0.6092688773471834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1525 + }, + { + "epoch": 0.609668397922493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1526 + }, + { + "epoch": 0.6100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1527 + }, + { + "epoch": 0.6104674390731123, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1528 + }, + { + "epoch": 0.6108669596484219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1529 + }, + { + "epoch": 0.6112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1530 + }, + { + "epoch": 0.6116660007990411, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1531 + }, + { + "epoch": 0.6120655213743508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1532 + }, + { + "epoch": 0.6124650419496604, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1533 + }, + { + "epoch": 0.61286456252497, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1534 + }, + { + "epoch": 0.6132640831002797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1535 + }, + { + "epoch": 0.6136636036755893, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1536 + }, + { + "epoch": 0.6140631242508989, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1537 + }, + { + "epoch": 0.6144626448262086, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1538 + }, + { + "epoch": 0.6148621654015182, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1539 + }, + { + "epoch": 0.6152616859768278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1540 + }, + { + "epoch": 0.6156612065521374, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1541 + }, + { + "epoch": 0.6160607271274471, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1542 + }, + { + "epoch": 0.6164602477027566, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1543 + }, + { + "epoch": 0.6168597682780663, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1544 + }, + { + "epoch": 0.617259288853376, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1545 + }, + { + "epoch": 0.6176588094286856, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1546 + }, + { + "epoch": 0.6180583300039952, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1547 + }, + { + "epoch": 0.6184578505793048, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1548 + }, + { + "epoch": 0.6188573711546145, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1549 + }, + { + "epoch": 0.619256891729924, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1550 + }, + { + "epoch": 0.6196564123052337, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1551 + }, + { + "epoch": 0.6200559328805434, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1552 + }, + { + "epoch": 0.620455453455853, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1553 + }, + { + "epoch": 0.6208549740311626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 1554 + }, + { + "epoch": 0.6212544946064722, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1555 + }, + { + "epoch": 0.6216540151817819, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1556 + }, + { + "epoch": 0.6220535357570914, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1557 + }, + { + "epoch": 0.6224530563324011, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1558 + }, + { + "epoch": 0.6228525769077108, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1559 + }, + { + "epoch": 0.6232520974830204, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1560 + }, + { + "epoch": 0.62365161805833, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1561 + }, + { + "epoch": 0.6240511386336396, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1562 + }, + { + "epoch": 0.6244506592089493, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1563 + }, + { + "epoch": 0.6248501797842589, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1564 + }, + { + "epoch": 0.6252497003595685, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1565 + }, + { + "epoch": 0.6256492209348782, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1566 + }, + { + "epoch": 0.6260487415101877, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1567 + }, + { + "epoch": 0.6264482620854974, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1568 + }, + { + "epoch": 0.626847782660807, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1569 + }, + { + "epoch": 0.6272473032361167, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1570 + }, + { + "epoch": 0.6276468238114263, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1571 + }, + { + "epoch": 0.6280463443867359, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1572 + }, + { + "epoch": 0.6284458649620456, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1573 + }, + { + "epoch": 0.6288453855373551, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1574 + }, + { + "epoch": 0.6292449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1575 + }, + { + "epoch": 0.6296444266879744, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1576 + }, + { + "epoch": 0.630043947263284, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1577 + }, + { + "epoch": 0.6304434678385937, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1578 + }, + { + "epoch": 0.6308429884139033, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1579 + }, + { + "epoch": 0.631242508989213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1580 + }, + { + "epoch": 0.6316420295645225, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1581 + }, + { + "epoch": 0.6320415501398322, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1582 + }, + { + "epoch": 0.6324410707151419, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1583 + }, + { + "epoch": 0.6328405912904514, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1584 + }, + { + "epoch": 0.6332401118657611, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1585 + }, + { + "epoch": 0.6336396324410707, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1586 + }, + { + "epoch": 0.6340391530163804, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1587 + }, + { + "epoch": 0.6344386735916899, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1588 + }, + { + "epoch": 0.6348381941669996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1589 + }, + { + "epoch": 0.6352377147423093, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1590 + }, + { + "epoch": 0.6356372353176188, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1591 + }, + { + "epoch": 0.6360367558929285, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1592 + }, + { + "epoch": 0.6364362764682381, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1593 + }, + { + "epoch": 0.6368357970435478, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1594 + }, + { + "epoch": 0.6372353176188573, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1595 + }, + { + "epoch": 0.637634838194167, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1596 + }, + { + "epoch": 0.6380343587694767, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1597 + }, + { + "epoch": 0.6384338793447862, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1598 + }, + { + "epoch": 0.6388333999200959, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1599 + }, + { + "epoch": 0.6392329204954055, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1600 + }, + { + "epoch": 0.6396324410707152, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1601 + }, + { + "epoch": 0.6400319616460247, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1602 + }, + { + "epoch": 0.6404314822213344, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1603 + }, + { + "epoch": 0.6408310027966441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1604 + }, + { + "epoch": 0.6412305233719536, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1605 + }, + { + "epoch": 0.6416300439472633, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1606 + }, + { + "epoch": 0.6420295645225729, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1607 + }, + { + "epoch": 0.6424290850978825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1608 + }, + { + "epoch": 0.6428286056731922, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1609 + }, + { + "epoch": 0.6432281262485018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1610 + }, + { + "epoch": 0.6436276468238115, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1611 + }, + { + "epoch": 0.644027167399121, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1612 + }, + { + "epoch": 0.6444266879744307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1613 + }, + { + "epoch": 0.6448262085497403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1614 + }, + { + "epoch": 0.6452257291250499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1615 + }, + { + "epoch": 0.6456252497003596, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 1616 + }, + { + "epoch": 0.6460247702756692, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1617 + }, + { + "epoch": 0.6464242908509789, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 1618 + }, + { + "epoch": 0.6468238114262884, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1619 + }, + { + "epoch": 0.6472233320015981, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1620 + }, + { + "epoch": 0.6476228525769077, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1621 + }, + { + "epoch": 0.6480223731522173, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1622 + }, + { + "epoch": 0.648421893727527, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1623 + }, + { + "epoch": 0.6488214143028366, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1624 + }, + { + "epoch": 0.6492209348781462, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1625 + }, + { + "epoch": 0.6496204554534558, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1626 + }, + { + "epoch": 0.6500199760287655, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1627 + }, + { + "epoch": 0.6504194966040752, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1628 + }, + { + "epoch": 0.6508190171793847, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1629 + }, + { + "epoch": 0.6512185377546944, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1630 + }, + { + "epoch": 0.651618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1631 + }, + { + "epoch": 0.6520175789053136, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1632 + }, + { + "epoch": 0.6524170994806232, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1633 + }, + { + "epoch": 0.6528166200559329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1634 + }, + { + "epoch": 0.6532161406312426, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1635 + }, + { + "epoch": 0.6536156612065521, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 1636 + }, + { + "epoch": 0.6540151817818618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1637 + }, + { + "epoch": 0.6544147023571714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1638 + }, + { + "epoch": 0.654814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1639 + }, + { + "epoch": 0.6552137435077906, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1640 + }, + { + "epoch": 0.6556132640831003, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1641 + }, + { + "epoch": 0.65601278465841, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1642 + }, + { + "epoch": 0.6564123052337195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1643 + }, + { + "epoch": 0.6568118258090292, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1644 + }, + { + "epoch": 0.6572113463843388, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1645 + }, + { + "epoch": 0.6576108669596484, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1646 + }, + { + "epoch": 0.658010387534958, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1647 + }, + { + "epoch": 0.6584099081102677, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1648 + }, + { + "epoch": 0.6588094286855773, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1649 + }, + { + "epoch": 0.6592089492608869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 1650 + }, + { + "epoch": 0.6596084698361966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 1651 + }, + { + "epoch": 0.6600079904115062, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1652 + }, + { + "epoch": 0.6604075109868158, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1653 + }, + { + "epoch": 0.6608070315621255, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1654 + }, + { + "epoch": 0.6612065521374351, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1655 + }, + { + "epoch": 0.6616060727127447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1656 + }, + { + "epoch": 0.6620055932880543, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1657 + }, + { + "epoch": 0.662405113863364, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1658 + }, + { + "epoch": 0.6628046344386735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1659 + }, + { + "epoch": 0.6632041550139832, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1660 + }, + { + "epoch": 0.6636036755892929, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1661 + }, + { + "epoch": 0.6640031961646025, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1662 + }, + { + "epoch": 0.6644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1663 + }, + { + "epoch": 0.6648022373152217, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1664 + }, + { + "epoch": 0.6652017578905314, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1665 + }, + { + "epoch": 0.6656012784658409, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1666 + }, + { + "epoch": 0.6660007990411506, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1667 + }, + { + "epoch": 0.6664003196164603, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1668 + }, + { + "epoch": 0.6667998401917699, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1669 + }, + { + "epoch": 0.6671993607670795, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1670 + }, + { + "epoch": 0.6675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1671 + }, + { + "epoch": 0.6679984019176988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1672 + }, + { + "epoch": 0.6683979224930084, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1673 + }, + { + "epoch": 0.668797443068318, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1674 + }, + { + "epoch": 0.6691969636436277, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1675 + }, + { + "epoch": 0.6695964842189372, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1676 + }, + { + "epoch": 0.6699960047942469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 1677 + }, + { + "epoch": 0.6703955253695565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1678 + }, + { + "epoch": 0.6707950459448662, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1679 + }, + { + "epoch": 0.6711945665201758, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1680 + }, + { + "epoch": 0.6715940870954854, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1681 + }, + { + "epoch": 0.6719936076707951, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1682 + }, + { + "epoch": 0.6723931282461046, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1683 + }, + { + "epoch": 0.6727926488214143, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1684 + }, + { + "epoch": 0.6731921693967239, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1685 + }, + { + "epoch": 0.6735916899720336, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1686 + }, + { + "epoch": 0.6739912105473432, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1687 + }, + { + "epoch": 0.6743907311226528, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1688 + }, + { + "epoch": 0.6747902516979625, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1689 + }, + { + "epoch": 0.675189772273272, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1690 + }, + { + "epoch": 0.6755892928485817, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1691 + }, + { + "epoch": 0.6759888134238913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1692 + }, + { + "epoch": 0.676388333999201, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1693 + }, + { + "epoch": 0.6767878545745106, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1694 + }, + { + "epoch": 0.6771873751498202, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1695 + }, + { + "epoch": 0.6775868957251299, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1696 + }, + { + "epoch": 0.6779864163004394, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1697 + }, + { + "epoch": 0.6783859368757491, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1698 + }, + { + "epoch": 0.6787854574510588, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1699 + }, + { + "epoch": 0.6791849780263683, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 1700 + }, + { + "epoch": 0.679584498601678, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1701 + }, + { + "epoch": 0.6799840191769876, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1702 + }, + { + "epoch": 0.6803835397522973, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1703 + }, + { + "epoch": 0.6807830603276068, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1704 + }, + { + "epoch": 0.6811825809029165, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1705 + }, + { + "epoch": 0.6815821014782262, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1706 + }, + { + "epoch": 0.6819816220535357, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1707 + }, + { + "epoch": 0.6823811426288454, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1708 + }, + { + "epoch": 0.682780663204155, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1709 + }, + { + "epoch": 0.6831801837794647, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1710 + }, + { + "epoch": 0.6835797043547742, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1711 + }, + { + "epoch": 0.6839792249300839, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1712 + }, + { + "epoch": 0.6843787455053936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1713 + }, + { + "epoch": 0.6847782660807031, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1714 + }, + { + "epoch": 0.6851777866560128, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1715 + }, + { + "epoch": 0.6855773072313224, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1716 + }, + { + "epoch": 0.685976827806632, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1717 + }, + { + "epoch": 0.6863763483819417, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1718 + }, + { + "epoch": 0.6867758689572513, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1719 + }, + { + "epoch": 0.687175389532561, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1720 + }, + { + "epoch": 0.6875749101078705, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1721 + }, + { + "epoch": 0.6879744306831802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1722 + }, + { + "epoch": 0.6883739512584898, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1723 + }, + { + "epoch": 0.6887734718337994, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1724 + }, + { + "epoch": 0.6891729924091091, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1725 + }, + { + "epoch": 0.6895725129844187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1726 + }, + { + "epoch": 0.6899720335597284, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1727 + }, + { + "epoch": 0.6903715541350379, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1728 + }, + { + "epoch": 0.6907710747103476, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1729 + }, + { + "epoch": 0.6911705952856572, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1730 + }, + { + "epoch": 0.6915701158609668, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1731 + }, + { + "epoch": 0.6919696364362765, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1732 + }, + { + "epoch": 0.6923691570115861, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1733 + }, + { + "epoch": 0.6927686775868958, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1734 + }, + { + "epoch": 0.6931681981622053, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1735 + }, + { + "epoch": 0.693567718737515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1736 + }, + { + "epoch": 0.6939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1737 + }, + { + "epoch": 0.6943667598881342, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1738 + }, + { + "epoch": 0.6947662804634439, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 1739 + }, + { + "epoch": 0.6951658010387535, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1740 + }, + { + "epoch": 0.6955653216140631, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1741 + }, + { + "epoch": 0.6959648421893727, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 1742 + }, + { + "epoch": 0.6963643627646824, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1743 + }, + { + "epoch": 0.6967638833399921, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1744 + }, + { + "epoch": 0.6971634039153016, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1745 + }, + { + "epoch": 0.6975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1746 + }, + { + "epoch": 0.6979624450659209, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1747 + }, + { + "epoch": 0.6983619656412305, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1748 + }, + { + "epoch": 0.6987614862165401, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1749 + }, + { + "epoch": 0.6991610067918498, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1750 + }, + { + "epoch": 0.6995605273671595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 1751 + }, + { + "epoch": 0.699960047942469, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1752 + }, + { + "epoch": 0.7003595685177787, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1753 + }, + { + "epoch": 0.7007590890930883, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1754 + }, + { + "epoch": 0.7011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1755 + }, + { + "epoch": 0.7015581302437075, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1756 + }, + { + "epoch": 0.7019576508190172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1757 + }, + { + "epoch": 0.7023571713943269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1758 + }, + { + "epoch": 0.7027566919696364, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1759 + }, + { + "epoch": 0.7031562125449461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1760 + }, + { + "epoch": 0.7035557331202557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1761 + }, + { + "epoch": 0.7039552536955653, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1762 + }, + { + "epoch": 0.7043547742708749, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1763 + }, + { + "epoch": 0.7047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1764 + }, + { + "epoch": 0.7051538154214942, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1765 + }, + { + "epoch": 0.7055533359968038, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1766 + }, + { + "epoch": 0.7059528565721135, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1767 + }, + { + "epoch": 0.706352377147423, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1768 + }, + { + "epoch": 0.7067518977227327, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1769 + }, + { + "epoch": 0.7071514182980424, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 1770 + }, + { + "epoch": 0.707550938873352, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1771 + }, + { + "epoch": 0.7079504594486616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1772 + }, + { + "epoch": 0.7083499800239712, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1773 + }, + { + "epoch": 0.7087495005992809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1774 + }, + { + "epoch": 0.7091490211745904, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1775 + }, + { + "epoch": 0.7095485417499001, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 1776 + }, + { + "epoch": 0.7099480623252098, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1777 + }, + { + "epoch": 0.7103475829005194, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1778 + }, + { + "epoch": 0.710747103475829, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1779 + }, + { + "epoch": 0.7111466240511386, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1780 + }, + { + "epoch": 0.7115461446264483, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1781 + }, + { + "epoch": 0.7119456652017578, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1782 + }, + { + "epoch": 0.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1783 + }, + { + "epoch": 0.7127447063523772, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1784 + }, + { + "epoch": 0.7131442269276868, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1785 + }, + { + "epoch": 0.7135437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1786 + }, + { + "epoch": 0.713943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1787 + }, + { + "epoch": 0.7143427886536157, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1788 + }, + { + "epoch": 0.7147423092289253, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1789 + }, + { + "epoch": 0.7151418298042349, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1790 + }, + { + "epoch": 0.7155413503795446, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1791 + }, + { + "epoch": 0.7159408709548541, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1792 + }, + { + "epoch": 0.7163403915301638, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1793 + }, + { + "epoch": 0.7167399121054734, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1794 + }, + { + "epoch": 0.7171394326807831, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1795 + }, + { + "epoch": 0.7175389532560927, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1796 + }, + { + "epoch": 0.7179384738314023, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1797 + }, + { + "epoch": 0.718337994406712, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1798 + }, + { + "epoch": 0.7187375149820215, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1799 + }, + { + "epoch": 0.7191370355573312, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 1800 + }, + { + "epoch": 0.7195365561326408, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1801 + }, + { + "epoch": 0.7199360767079505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1802 + }, + { + "epoch": 0.7203355972832601, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 1803 + }, + { + "epoch": 0.7207351178585697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1804 + }, + { + "epoch": 0.7211346384338794, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1805 + }, + { + "epoch": 0.7215341590091889, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1806 + }, + { + "epoch": 0.7219336795844986, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1807 + }, + { + "epoch": 0.7223332001598082, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1808 + }, + { + "epoch": 0.7227327207351179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1809 + }, + { + "epoch": 0.7231322413104275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1810 + }, + { + "epoch": 0.7235317618857371, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1811 + }, + { + "epoch": 0.7239312824610468, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1812 + }, + { + "epoch": 0.7243308030363563, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1813 + }, + { + "epoch": 0.724730323611666, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1814 + }, + { + "epoch": 0.7251298441869757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1815 + }, + { + "epoch": 0.7255293647622852, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1816 + }, + { + "epoch": 0.7259288853375949, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1817 + }, + { + "epoch": 0.7263284059129045, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1818 + }, + { + "epoch": 0.7267279264882142, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1819 + }, + { + "epoch": 0.7271274470635237, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1820 + }, + { + "epoch": 0.7275269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1821 + }, + { + "epoch": 0.7279264882141431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1822 + }, + { + "epoch": 0.7283260087894526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1823 + }, + { + "epoch": 0.7287255293647623, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1824 + }, + { + "epoch": 0.7291250499400719, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 1825 + }, + { + "epoch": 0.7295245705153816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1826 + }, + { + "epoch": 0.7299240910906911, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1827 + }, + { + "epoch": 0.7303236116660008, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1828 + }, + { + "epoch": 0.7307231322413105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 1829 + }, + { + "epoch": 0.73112265281662, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1830 + }, + { + "epoch": 0.7315221733919297, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1831 + }, + { + "epoch": 0.7319216939672393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1832 + }, + { + "epoch": 0.732321214542549, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1833 + }, + { + "epoch": 0.7327207351178586, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1834 + }, + { + "epoch": 0.7331202556931682, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1835 + }, + { + "epoch": 0.7335197762684779, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1836 + }, + { + "epoch": 0.7339192968437874, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1837 + }, + { + "epoch": 0.7343188174190971, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1838 + }, + { + "epoch": 0.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1839 + }, + { + "epoch": 0.7351178585697163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1840 + }, + { + "epoch": 0.735517379145026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1841 + }, + { + "epoch": 0.7359168997203356, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1842 + }, + { + "epoch": 0.7363164202956453, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 1843 + }, + { + "epoch": 0.7367159408709548, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1844 + }, + { + "epoch": 0.7371154614462645, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1845 + }, + { + "epoch": 0.7375149820215741, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1846 + }, + { + "epoch": 0.7379145025968837, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1847 + }, + { + "epoch": 0.7383140231721934, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1848 + }, + { + "epoch": 0.738713543747503, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 1849 + }, + { + "epoch": 0.7391130643228127, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1850 + }, + { + "epoch": 0.7395125848981222, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1851 + }, + { + "epoch": 0.7399121054734319, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1852 + }, + { + "epoch": 0.7403116260487415, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1853 + }, + { + "epoch": 0.7407111466240511, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 1854 + }, + { + "epoch": 0.7411106671993608, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1855 + }, + { + "epoch": 0.7415101877746704, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1856 + }, + { + "epoch": 0.74190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1857 + }, + { + "epoch": 0.7423092289252896, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1858 + }, + { + "epoch": 0.7427087495005993, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1859 + }, + { + "epoch": 0.743108270075909, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1860 + }, + { + "epoch": 0.7435077906512185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1861 + }, + { + "epoch": 0.7439073112265282, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1862 + }, + { + "epoch": 0.7443068318018378, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 1863 + }, + { + "epoch": 0.7447063523771474, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1864 + }, + { + "epoch": 0.745105872952457, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1865 + }, + { + "epoch": 0.7455053935277667, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1866 + }, + { + "epoch": 0.7459049141030764, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1867 + }, + { + "epoch": 0.7463044346783859, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1868 + }, + { + "epoch": 0.7467039552536956, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1869 + }, + { + "epoch": 0.7471034758290052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 1870 + }, + { + "epoch": 0.7475029964043148, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1871 + }, + { + "epoch": 0.7479025169796244, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1872 + }, + { + "epoch": 0.7483020375549341, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1873 + }, + { + "epoch": 0.7487015581302437, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1874 + }, + { + "epoch": 0.7491010787055533, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1875 + }, + { + "epoch": 0.749500599280863, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1876 + }, + { + "epoch": 0.7499001198561726, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1877 + }, + { + "epoch": 0.7502996404314822, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1878 + }, + { + "epoch": 0.7506991610067919, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1879 + }, + { + "epoch": 0.7510986815821015, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1880 + }, + { + "epoch": 0.7514982021574111, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 1881 + }, + { + "epoch": 0.7518977227327207, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1882 + }, + { + "epoch": 0.7522972433080304, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1883 + }, + { + "epoch": 0.75269676388334, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1884 + }, + { + "epoch": 0.7530962844586496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1885 + }, + { + "epoch": 0.7534958050339593, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1886 + }, + { + "epoch": 0.7538953256092689, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1887 + }, + { + "epoch": 0.7542948461845785, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1888 + }, + { + "epoch": 0.7546943667598881, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1889 + }, + { + "epoch": 0.7550938873351978, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 1890 + }, + { + "epoch": 0.7554934079105073, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1891 + }, + { + "epoch": 0.755892928485817, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1892 + }, + { + "epoch": 0.7562924490611267, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1893 + }, + { + "epoch": 0.7566919696364363, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1894 + }, + { + "epoch": 0.7570914902117459, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1895 + }, + { + "epoch": 0.7574910107870555, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1896 + }, + { + "epoch": 0.7578905313623652, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1897 + }, + { + "epoch": 0.7582900519376747, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 1898 + }, + { + "epoch": 0.7586895725129844, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1899 + }, + { + "epoch": 0.7590890930882941, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1900 + }, + { + "epoch": 0.7594886136636037, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1901 + }, + { + "epoch": 0.7598881342389133, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1902 + }, + { + "epoch": 0.7602876548142229, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1903 + }, + { + "epoch": 0.7606871753895326, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1904 + }, + { + "epoch": 0.7610866959648422, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 1905 + }, + { + "epoch": 0.7614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1906 + }, + { + "epoch": 0.7618857371154615, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1907 + }, + { + "epoch": 0.762285257690771, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1908 + }, + { + "epoch": 0.7626847782660807, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1909 + }, + { + "epoch": 0.7630842988413903, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1910 + }, + { + "epoch": 0.7634838194167, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1911 + }, + { + "epoch": 0.7638833399920096, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1912 + }, + { + "epoch": 0.7642828605673192, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1913 + }, + { + "epoch": 0.7646823811426289, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1914 + }, + { + "epoch": 0.7650819017179384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1915 + }, + { + "epoch": 0.7654814222932481, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1916 + }, + { + "epoch": 0.7658809428685577, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1917 + }, + { + "epoch": 0.7662804634438674, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1918 + }, + { + "epoch": 0.766679984019177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1919 + }, + { + "epoch": 0.7670795045944866, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1920 + }, + { + "epoch": 0.7674790251697963, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1921 + }, + { + "epoch": 0.7678785457451058, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1922 + }, + { + "epoch": 0.7682780663204155, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1923 + }, + { + "epoch": 0.7686775868957252, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1924 + }, + { + "epoch": 0.7690771074710347, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1925 + }, + { + "epoch": 0.7694766280463444, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 1926 + }, + { + "epoch": 0.769876148621654, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1927 + }, + { + "epoch": 0.7702756691969637, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1928 + }, + { + "epoch": 0.7706751897722732, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1929 + }, + { + "epoch": 0.7710747103475829, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1930 + }, + { + "epoch": 0.7714742309228926, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1931 + }, + { + "epoch": 0.7718737514982021, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1932 + }, + { + "epoch": 0.7722732720735118, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1933 + }, + { + "epoch": 0.7726727926488214, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1934 + }, + { + "epoch": 0.7730723132241311, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1935 + }, + { + "epoch": 0.7734718337994406, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1936 + }, + { + "epoch": 0.7738713543747503, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1937 + }, + { + "epoch": 0.77427087495006, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1938 + }, + { + "epoch": 0.7746703955253695, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 1939 + }, + { + "epoch": 0.7750699161006792, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1940 + }, + { + "epoch": 0.7754694366759888, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1941 + }, + { + "epoch": 0.7758689572512985, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 1942 + }, + { + "epoch": 0.776268477826608, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 1943 + }, + { + "epoch": 0.7766679984019177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1944 + }, + { + "epoch": 0.7770675189772274, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 1945 + }, + { + "epoch": 0.7774670395525369, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1946 + }, + { + "epoch": 0.7778665601278466, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1947 + }, + { + "epoch": 0.7782660807031562, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1948 + }, + { + "epoch": 0.7786656012784658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1949 + }, + { + "epoch": 0.7790651218537755, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 1950 + }, + { + "epoch": 0.7794646424290851, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1951 + }, + { + "epoch": 0.7798641630043948, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1952 + }, + { + "epoch": 0.7802636835797043, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 1953 + }, + { + "epoch": 0.780663204155014, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1954 + }, + { + "epoch": 0.7810627247303236, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1955 + }, + { + "epoch": 0.7814622453056332, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1956 + }, + { + "epoch": 0.7818617658809429, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1957 + }, + { + "epoch": 0.7822612864562525, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1958 + }, + { + "epoch": 0.7826608070315622, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1959 + }, + { + "epoch": 0.7830603276068717, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1960 + }, + { + "epoch": 0.7834598481821814, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1961 + }, + { + "epoch": 0.783859368757491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1962 + }, + { + "epoch": 0.7842588893328006, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1963 + }, + { + "epoch": 0.7846584099081103, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1964 + }, + { + "epoch": 0.7850579304834199, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1965 + }, + { + "epoch": 0.7854574510587295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1966 + }, + { + "epoch": 0.7858569716340391, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 1967 + }, + { + "epoch": 0.7862564922093488, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1968 + }, + { + "epoch": 0.7866560127846585, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 1969 + }, + { + "epoch": 0.787055533359968, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1970 + }, + { + "epoch": 0.7874550539352777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 1971 + }, + { + "epoch": 0.7878545745105873, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1972 + }, + { + "epoch": 0.7882540950858969, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1973 + }, + { + "epoch": 0.7886536156612065, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1974 + }, + { + "epoch": 0.7890531362365162, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1975 + }, + { + "epoch": 0.7894526568118259, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1976 + }, + { + "epoch": 0.7898521773871354, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1977 + }, + { + "epoch": 0.7902516979624451, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1978 + }, + { + "epoch": 0.7906512185377547, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 1979 + }, + { + "epoch": 0.7910507391130643, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1980 + }, + { + "epoch": 0.7914502596883739, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1981 + }, + { + "epoch": 0.7918497802636836, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 1982 + }, + { + "epoch": 0.7922493008389933, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 1983 + }, + { + "epoch": 0.7926488214143028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1984 + }, + { + "epoch": 0.7930483419896125, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1985 + }, + { + "epoch": 0.7934478625649221, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1986 + }, + { + "epoch": 0.7938473831402317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1987 + }, + { + "epoch": 0.7942469037155413, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1988 + }, + { + "epoch": 0.794646424290851, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 1989 + }, + { + "epoch": 0.7950459448661606, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1990 + }, + { + "epoch": 0.7954454654414702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 1991 + }, + { + "epoch": 0.7958449860167799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1992 + }, + { + "epoch": 0.7962445065920895, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1993 + }, + { + "epoch": 0.7966440271673991, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1994 + }, + { + "epoch": 0.7970435477427088, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1995 + }, + { + "epoch": 0.7974430683180184, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 1996 + }, + { + "epoch": 0.797842588893328, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1997 + }, + { + "epoch": 0.7982421094686376, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1998 + }, + { + "epoch": 0.7986416300439473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1999 + }, + { + "epoch": 0.7990411506192568, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2000 + }, + { + "epoch": 0.7994406711945665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2001 + }, + { + "epoch": 0.7998401917698762, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2002 + }, + { + "epoch": 0.8002397123451858, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2003 + }, + { + "epoch": 0.8006392329204954, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2004 + }, + { + "epoch": 0.801038753495805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 2005 + }, + { + "epoch": 0.8014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2006 + }, + { + "epoch": 0.8018377946464242, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2007 + }, + { + "epoch": 0.8022373152217339, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2008 + }, + { + "epoch": 0.8026368357970436, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2009 + }, + { + "epoch": 0.8030363563723532, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 2010 + }, + { + "epoch": 0.8034358769476628, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2011 + }, + { + "epoch": 0.8038353975229724, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2012 + }, + { + "epoch": 0.8042349180982821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2013 + }, + { + "epoch": 0.8046344386735917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 2014 + }, + { + "epoch": 0.8050339592489013, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 2015 + }, + { + "epoch": 0.805433479824211, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2016 + }, + { + "epoch": 0.8058330003995205, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2017 + }, + { + "epoch": 0.8062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2018 + }, + { + "epoch": 0.8066320415501398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2019 + }, + { + "epoch": 0.8070315621254495, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2020 + }, + { + "epoch": 0.8074310827007591, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2021 + }, + { + "epoch": 0.8078306032760687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2022 + }, + { + "epoch": 0.8082301238513784, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2023 + }, + { + "epoch": 0.8086296444266879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2024 + }, + { + "epoch": 0.8090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2025 + }, + { + "epoch": 0.8094286855773072, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 2026 + }, + { + "epoch": 0.8098282061526169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2027 + }, + { + "epoch": 0.8102277267279265, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2028 + }, + { + "epoch": 0.8106272473032361, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2029 + }, + { + "epoch": 0.8110267678785458, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2030 + }, + { + "epoch": 0.8114262884538553, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2031 + }, + { + "epoch": 0.811825809029165, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 2032 + }, + { + "epoch": 0.8122253296044746, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2033 + }, + { + "epoch": 0.8126248501797843, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2034 + }, + { + "epoch": 0.8130243707550939, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 2035 + }, + { + "epoch": 0.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2036 + }, + { + "epoch": 0.8138234119057132, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 2037 + }, + { + "epoch": 0.8142229324810227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2038 + }, + { + "epoch": 0.8146224530563324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2039 + }, + { + "epoch": 0.8150219736316421, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2040 + }, + { + "epoch": 0.8154214942069516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2041 + }, + { + "epoch": 0.8158210147822613, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2042 + }, + { + "epoch": 0.8162205353575709, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2043 + }, + { + "epoch": 0.8166200559328806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2044 + }, + { + "epoch": 0.8170195765081901, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2045 + }, + { + "epoch": 0.8174190970834998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2046 + }, + { + "epoch": 0.8178186176588095, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2047 + }, + { + "epoch": 0.818218138234119, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2048 + }, + { + "epoch": 0.8186176588094287, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 2049 + }, + { + "epoch": 0.8190171793847383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2050 + }, + { + "epoch": 0.819416699960048, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2051 + }, + { + "epoch": 0.8198162205353575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2052 + }, + { + "epoch": 0.8202157411106672, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2053 + }, + { + "epoch": 0.8206152616859769, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2054 + }, + { + "epoch": 0.8210147822612864, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2055 + }, + { + "epoch": 0.8214143028365961, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2056 + }, + { + "epoch": 0.8218138234119057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2057 + }, + { + "epoch": 0.8222133439872154, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2058 + }, + { + "epoch": 0.8226128645625249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2059 + }, + { + "epoch": 0.8230123851378346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2060 + }, + { + "epoch": 0.8234119057131443, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2061 + }, + { + "epoch": 0.8238114262884538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2062 + }, + { + "epoch": 0.8242109468637635, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2063 + }, + { + "epoch": 0.8246104674390731, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2064 + }, + { + "epoch": 0.8250099880143827, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2065 + }, + { + "epoch": 0.8254095085896924, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2066 + }, + { + "epoch": 0.825809029165002, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2067 + }, + { + "epoch": 0.8262085497403117, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2068 + }, + { + "epoch": 0.8266080703156212, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2069 + }, + { + "epoch": 0.8270075908909309, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2070 + }, + { + "epoch": 0.8274071114662405, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2071 + }, + { + "epoch": 0.8278066320415501, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2072 + }, + { + "epoch": 0.8282061526168598, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 2073 + }, + { + "epoch": 0.8286056731921694, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 2074 + }, + { + "epoch": 0.829005193767479, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2075 + }, + { + "epoch": 0.8294047143427886, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2076 + }, + { + "epoch": 0.8298042349180983, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2077 + }, + { + "epoch": 0.8302037554934079, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2078 + }, + { + "epoch": 0.8306032760687175, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2079 + }, + { + "epoch": 0.8310027966440272, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2080 + }, + { + "epoch": 0.8314023172193368, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2081 + }, + { + "epoch": 0.8318018377946464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2082 + }, + { + "epoch": 0.832201358369956, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2083 + }, + { + "epoch": 0.8326008789452657, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2084 + }, + { + "epoch": 0.8330003995205754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2085 + }, + { + "epoch": 0.8333999200958849, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2086 + }, + { + "epoch": 0.8337994406711946, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2087 + }, + { + "epoch": 0.8341989612465042, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2088 + }, + { + "epoch": 0.8345984818218138, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2089 + }, + { + "epoch": 0.8349980023971234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 2090 + }, + { + "epoch": 0.8353975229724331, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 2091 + }, + { + "epoch": 0.8357970435477428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2092 + }, + { + "epoch": 0.8361965641230523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2093 + }, + { + "epoch": 0.836596084698362, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2094 + }, + { + "epoch": 0.8369956052736716, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2095 + }, + { + "epoch": 0.8373951258489812, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2096 + }, + { + "epoch": 0.8377946464242908, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2097 + }, + { + "epoch": 0.8381941669996005, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2098 + }, + { + "epoch": 0.8385936875749102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2099 + }, + { + "epoch": 0.8389932081502197, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2100 + }, + { + "epoch": 0.8393927287255294, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 2101 + }, + { + "epoch": 0.839792249300839, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2102 + }, + { + "epoch": 0.8401917698761486, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2103 + }, + { + "epoch": 0.8405912904514582, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 2104 + }, + { + "epoch": 0.8409908110267679, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2105 + }, + { + "epoch": 0.8413903316020775, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2106 + }, + { + "epoch": 0.8417898521773871, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2107 + }, + { + "epoch": 0.8421893727526968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2108 + }, + { + "epoch": 0.8425888933280063, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2109 + }, + { + "epoch": 0.842988413903316, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2110 + }, + { + "epoch": 0.8433879344786257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2111 + }, + { + "epoch": 0.8437874550539353, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2112 + }, + { + "epoch": 0.8441869756292449, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2113 + }, + { + "epoch": 0.8445864962045545, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2114 + }, + { + "epoch": 0.8449860167798642, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2115 + }, + { + "epoch": 0.8453855373551737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2116 + }, + { + "epoch": 0.8457850579304834, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2117 + }, + { + "epoch": 0.8461845785057931, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2118 + }, + { + "epoch": 0.8465840990811027, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2119 + }, + { + "epoch": 0.8469836196564123, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2120 + }, + { + "epoch": 0.8473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2121 + }, + { + "epoch": 0.8477826608070316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2122 + }, + { + "epoch": 0.8481821813823411, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 2123 + }, + { + "epoch": 0.8485817019576508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2124 + }, + { + "epoch": 0.8489812225329605, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 2125 + }, + { + "epoch": 0.84938074310827, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2126 + }, + { + "epoch": 0.8497802636835797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2127 + }, + { + "epoch": 0.8501797842588893, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2128 + }, + { + "epoch": 0.850579304834199, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2129 + }, + { + "epoch": 0.8509788254095086, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2130 + }, + { + "epoch": 0.8513783459848182, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2131 + }, + { + "epoch": 0.8517778665601279, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2132 + }, + { + "epoch": 0.8521773871354374, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 0.8525769077107471, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2134 + }, + { + "epoch": 0.8529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2135 + }, + { + "epoch": 0.8533759488613664, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2136 + }, + { + "epoch": 0.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2137 + }, + { + "epoch": 0.8541749900119856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2138 + }, + { + "epoch": 0.8545745105872953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2139 + }, + { + "epoch": 0.8549740311626048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2140 + }, + { + "epoch": 0.8553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2141 + }, + { + "epoch": 0.8557730723132241, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2142 + }, + { + "epoch": 0.8561725928885338, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 2143 + }, + { + "epoch": 0.8565721134638434, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2144 + }, + { + "epoch": 0.856971634039153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2145 + }, + { + "epoch": 0.8573711546144627, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2146 + }, + { + "epoch": 0.8577706751897722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2147 + }, + { + "epoch": 0.8581701957650819, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 2148 + }, + { + "epoch": 0.8585697163403915, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2149 + }, + { + "epoch": 0.8589692369157012, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2150 + }, + { + "epoch": 0.8593687574910108, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2151 + }, + { + "epoch": 0.8597682780663204, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2152 + }, + { + "epoch": 0.8601677986416301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2153 + }, + { + "epoch": 0.8605673192169396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 2154 + }, + { + "epoch": 0.8609668397922493, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2155 + }, + { + "epoch": 0.861366360367559, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2156 + }, + { + "epoch": 0.8617658809428685, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2157 + }, + { + "epoch": 0.8621654015181782, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2158 + }, + { + "epoch": 0.8625649220934878, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2159 + }, + { + "epoch": 0.8629644426687975, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2160 + }, + { + "epoch": 0.863363963244107, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2161 + }, + { + "epoch": 0.8637634838194167, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2162 + }, + { + "epoch": 0.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2163 + }, + { + "epoch": 0.8645625249700359, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2164 + }, + { + "epoch": 0.8649620455453456, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2165 + }, + { + "epoch": 0.8653615661206552, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2166 + }, + { + "epoch": 0.8657610866959649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2167 + }, + { + "epoch": 0.8661606072712744, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2168 + }, + { + "epoch": 0.8665601278465841, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2169 + }, + { + "epoch": 0.8669596484218938, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2170 + }, + { + "epoch": 0.8673591689972033, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2171 + }, + { + "epoch": 0.867758689572513, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2172 + }, + { + "epoch": 0.8681582101478226, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2173 + }, + { + "epoch": 0.8685577307231322, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2174 + }, + { + "epoch": 0.8689572512984419, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2175 + }, + { + "epoch": 0.8693567718737515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2176 + }, + { + "epoch": 0.8697562924490612, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2177 + }, + { + "epoch": 0.8701558130243707, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2178 + }, + { + "epoch": 0.8705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 2179 + }, + { + "epoch": 0.87095485417499, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2180 + }, + { + "epoch": 0.8713543747502996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2181 + }, + { + "epoch": 0.8717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2182 + }, + { + "epoch": 0.8721534159009189, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2183 + }, + { + "epoch": 0.8725529364762286, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2184 + }, + { + "epoch": 0.8729524570515381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2185 + }, + { + "epoch": 0.8733519776268478, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2186 + }, + { + "epoch": 0.8737514982021574, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2187 + }, + { + "epoch": 0.874151018777467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2188 + }, + { + "epoch": 0.8745505393527767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2189 + }, + { + "epoch": 0.8749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2190 + }, + { + "epoch": 0.875349580503396, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2191 + }, + { + "epoch": 0.8757491010787055, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2192 + }, + { + "epoch": 0.8761486216540152, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2193 + }, + { + "epoch": 0.8765481422293248, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2194 + }, + { + "epoch": 0.8769476628046344, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2195 + }, + { + "epoch": 0.8773471833799441, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2196 + }, + { + "epoch": 0.8777467039552537, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2197 + }, + { + "epoch": 0.8781462245305633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2198 + }, + { + "epoch": 0.8785457451058729, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 2199 + }, + { + "epoch": 0.8789452656811826, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2200 + }, + { + "epoch": 0.8793447862564923, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2201 + }, + { + "epoch": 0.8797443068318018, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2202 + }, + { + "epoch": 0.8801438274071115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2203 + }, + { + "epoch": 0.8805433479824211, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2204 + }, + { + "epoch": 0.8809428685577307, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2205 + }, + { + "epoch": 0.8813423891330403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2206 + }, + { + "epoch": 0.88174190970835, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2207 + }, + { + "epoch": 0.8821414302836597, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2208 + }, + { + "epoch": 0.8825409508589692, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2209 + }, + { + "epoch": 0.8829404714342789, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2210 + }, + { + "epoch": 0.8833399920095885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2211 + }, + { + "epoch": 0.8837395125848981, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 2212 + }, + { + "epoch": 0.8841390331602077, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2213 + }, + { + "epoch": 0.8845385537355174, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2214 + }, + { + "epoch": 0.884938074310827, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2215 + }, + { + "epoch": 0.8853375948861366, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2216 + }, + { + "epoch": 0.8857371154614463, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2217 + }, + { + "epoch": 0.8861366360367559, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2218 + }, + { + "epoch": 0.8865361566120655, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2219 + }, + { + "epoch": 0.8869356771873752, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2220 + }, + { + "epoch": 0.8873351977626848, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2221 + }, + { + "epoch": 0.8877347183379944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2222 + }, + { + "epoch": 0.888134238913304, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2223 + }, + { + "epoch": 0.8885337594886137, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2224 + }, + { + "epoch": 0.8889332800639232, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2225 + }, + { + "epoch": 0.8893328006392329, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2226 + }, + { + "epoch": 0.8897323212145426, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2227 + }, + { + "epoch": 0.8901318417898522, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2228 + }, + { + "epoch": 0.8905313623651618, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2229 + }, + { + "epoch": 0.8909308829404714, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2230 + }, + { + "epoch": 0.8913304035157811, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2231 + }, + { + "epoch": 0.8917299240910906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2232 + }, + { + "epoch": 0.8921294446664003, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2233 + }, + { + "epoch": 0.89252896524171, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2234 + }, + { + "epoch": 0.8929284858170196, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2235 + }, + { + "epoch": 0.8933280063923292, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2236 + }, + { + "epoch": 0.8937275269676388, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2237 + }, + { + "epoch": 0.8941270475429485, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2238 + }, + { + "epoch": 0.894526568118258, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2239 + }, + { + "epoch": 0.8949260886935677, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2240 + }, + { + "epoch": 0.8953256092688774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2241 + }, + { + "epoch": 0.895725129844187, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2242 + }, + { + "epoch": 0.8961246504194966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2243 + }, + { + "epoch": 0.8965241709948062, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2244 + }, + { + "epoch": 0.8969236915701159, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2245 + }, + { + "epoch": 0.8973232121454255, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2246 + }, + { + "epoch": 0.8977227327207351, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2247 + }, + { + "epoch": 0.8981222532960448, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2248 + }, + { + "epoch": 0.8985217738713543, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2249 + }, + { + "epoch": 0.898921294446664, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2250 + }, + { + "epoch": 0.8993208150219736, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2251 + }, + { + "epoch": 0.8997203355972833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2252 + }, + { + "epoch": 0.9001198561725929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2253 + }, + { + "epoch": 0.9005193767479025, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2254 + }, + { + "epoch": 0.9009188973232122, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2255 + }, + { + "epoch": 0.9013184178985217, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2256 + }, + { + "epoch": 0.9017179384738314, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2257 + }, + { + "epoch": 0.902117459049141, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2258 + }, + { + "epoch": 0.9025169796244507, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2259 + }, + { + "epoch": 0.9029165001997603, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2260 + }, + { + "epoch": 0.9033160207750699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2261 + }, + { + "epoch": 0.9037155413503796, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2262 + }, + { + "epoch": 0.9041150619256891, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2263 + }, + { + "epoch": 0.9045145825009988, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2264 + }, + { + "epoch": 0.9049141030763085, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2265 + }, + { + "epoch": 0.905313623651618, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2266 + }, + { + "epoch": 0.9057131442269277, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2267 + }, + { + "epoch": 0.9061126648022373, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2268 + }, + { + "epoch": 0.906512185377547, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 2269 + }, + { + "epoch": 0.9069117059528565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2270 + }, + { + "epoch": 0.9073112265281662, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2271 + }, + { + "epoch": 0.9077107471034759, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2272 + }, + { + "epoch": 0.9081102676787854, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2273 + }, + { + "epoch": 0.9085097882540951, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2274 + }, + { + "epoch": 0.9089093088294047, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2275 + }, + { + "epoch": 0.9093088294047144, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2276 + }, + { + "epoch": 0.9097083499800239, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2277 + }, + { + "epoch": 0.9101078705553336, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2278 + }, + { + "epoch": 0.9105073911306433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2279 + }, + { + "epoch": 0.9109069117059528, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2280 + }, + { + "epoch": 0.9113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2281 + }, + { + "epoch": 0.9117059528565721, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2282 + }, + { + "epoch": 0.9121054734318818, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2283 + }, + { + "epoch": 0.9125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2284 + }, + { + "epoch": 0.912904514582501, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2285 + }, + { + "epoch": 0.9133040351578107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2286 + }, + { + "epoch": 0.9137035557331202, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2287 + }, + { + "epoch": 0.9141030763084299, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2288 + }, + { + "epoch": 0.9145025968837395, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2289 + }, + { + "epoch": 0.9149021174590491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2290 + }, + { + "epoch": 0.9153016380343588, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 2291 + }, + { + "epoch": 0.9157011586096684, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2292 + }, + { + "epoch": 0.9161006791849781, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2293 + }, + { + "epoch": 0.9165001997602876, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2294 + }, + { + "epoch": 0.9168997203355973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2295 + }, + { + "epoch": 0.9172992409109069, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2296 + }, + { + "epoch": 0.9176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2297 + }, + { + "epoch": 0.9180982820615262, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2298 + }, + { + "epoch": 0.9184978026368358, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2299 + }, + { + "epoch": 0.9188973232121455, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2300 + }, + { + "epoch": 0.919296843787455, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2301 + }, + { + "epoch": 0.9196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2302 + }, + { + "epoch": 0.9200958849380743, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 2303 + }, + { + "epoch": 0.9204954055133839, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2304 + }, + { + "epoch": 0.9208949260886936, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2305 + }, + { + "epoch": 0.9212944466640032, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2306 + }, + { + "epoch": 0.9216939672393129, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2307 + }, + { + "epoch": 0.9220934878146224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2308 + }, + { + "epoch": 0.9224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2309 + }, + { + "epoch": 0.9228925289652418, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2310 + }, + { + "epoch": 0.9232920495405513, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2311 + }, + { + "epoch": 0.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2312 + }, + { + "epoch": 0.9240910906911706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2313 + }, + { + "epoch": 0.9244906112664802, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2314 + }, + { + "epoch": 0.9248901318417898, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2315 + }, + { + "epoch": 0.9252896524170995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 2316 + }, + { + "epoch": 0.9256891729924092, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2317 + }, + { + "epoch": 0.9260886935677187, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2318 + }, + { + "epoch": 0.9264882141430284, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2319 + }, + { + "epoch": 0.926887734718338, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2320 + }, + { + "epoch": 0.9272872552936476, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2321 + }, + { + "epoch": 0.9276867758689572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2322 + }, + { + "epoch": 0.9280862964442669, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2323 + }, + { + "epoch": 0.9284858170195766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2324 + }, + { + "epoch": 0.9288853375948861, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2325 + }, + { + "epoch": 0.9292848581701958, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2326 + }, + { + "epoch": 0.9296843787455054, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2327 + }, + { + "epoch": 0.930083899320815, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2328 + }, + { + "epoch": 0.9304834198961246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 2329 + }, + { + "epoch": 0.9308829404714343, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2330 + }, + { + "epoch": 0.931282461046744, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2331 + }, + { + "epoch": 0.9316819816220535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 2332 + }, + { + "epoch": 0.9320815021973632, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2333 + }, + { + "epoch": 0.9324810227726728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2334 + }, + { + "epoch": 0.9328805433479824, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2335 + }, + { + "epoch": 0.9332800639232921, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2336 + }, + { + "epoch": 0.9336795844986017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2337 + }, + { + "epoch": 0.9340791050739113, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2338 + }, + { + "epoch": 0.9344786256492209, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2339 + }, + { + "epoch": 0.9348781462245306, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2340 + }, + { + "epoch": 0.9352776667998401, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2341 + }, + { + "epoch": 0.9356771873751498, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2342 + }, + { + "epoch": 0.9360767079504595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2343 + }, + { + "epoch": 0.9364762285257691, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 2344 + }, + { + "epoch": 0.9368757491010787, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2345 + }, + { + "epoch": 0.9372752696763883, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2346 + }, + { + "epoch": 0.937674790251698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2347 + }, + { + "epoch": 0.9380743108270075, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2348 + }, + { + "epoch": 0.9384738314023172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2349 + }, + { + "epoch": 0.9388733519776269, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2350 + }, + { + "epoch": 0.9392728725529365, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2351 + }, + { + "epoch": 0.9396723931282461, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2352 + }, + { + "epoch": 0.9400719137035557, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2353 + }, + { + "epoch": 0.9404714342788654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2354 + }, + { + "epoch": 0.9408709548541749, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2355 + }, + { + "epoch": 0.9412704754294846, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2356 + }, + { + "epoch": 0.9416699960047943, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2357 + }, + { + "epoch": 0.9420695165801038, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2358 + }, + { + "epoch": 0.9424690371554135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 2359 + }, + { + "epoch": 0.9428685577307231, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2360 + }, + { + "epoch": 0.9432680783060328, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 2361 + }, + { + "epoch": 0.9436675988813424, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 2362 + }, + { + "epoch": 0.944067119456652, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 2363 + }, + { + "epoch": 0.9444666400319617, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2364 + }, + { + "epoch": 0.9448661606072712, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 2365 + }, + { + "epoch": 0.9452656811825809, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2366 + }, + { + "epoch": 0.9456652017578905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2367 + }, + { + "epoch": 0.9460647223332002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2368 + }, + { + "epoch": 0.9464642429085098, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2369 + }, + { + "epoch": 0.9468637634838194, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2370 + }, + { + "epoch": 0.9472632840591291, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2371 + }, + { + "epoch": 0.9476628046344386, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2372 + }, + { + "epoch": 0.9480623252097483, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2373 + }, + { + "epoch": 0.9484618457850579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2374 + }, + { + "epoch": 0.9488613663603676, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2375 + }, + { + "epoch": 0.9492608869356772, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2376 + }, + { + "epoch": 0.9496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2377 + }, + { + "epoch": 0.9500599280862965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2378 + }, + { + "epoch": 0.950459448661606, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2379 + }, + { + "epoch": 0.9508589692369157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2380 + }, + { + "epoch": 0.9512584898122254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2381 + }, + { + "epoch": 0.951658010387535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 2382 + }, + { + "epoch": 0.9520575309628446, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2383 + }, + { + "epoch": 0.9524570515381542, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2384 + }, + { + "epoch": 0.9528565721134639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2385 + }, + { + "epoch": 0.9532560926887734, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2386 + }, + { + "epoch": 0.9536556132640831, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2387 + }, + { + "epoch": 0.9540551338393928, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2388 + }, + { + "epoch": 0.9544546544147023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2389 + }, + { + "epoch": 0.954854174990012, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 2390 + }, + { + "epoch": 0.9552536955653216, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2391 + }, + { + "epoch": 0.9556532161406313, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 2392 + }, + { + "epoch": 0.9560527367159408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2393 + }, + { + "epoch": 0.9564522572912505, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2394 + }, + { + "epoch": 0.9568517778665602, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2395 + }, + { + "epoch": 0.9572512984418697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2396 + }, + { + "epoch": 0.9576508190171794, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2397 + }, + { + "epoch": 0.958050339592489, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2398 + }, + { + "epoch": 0.9584498601677987, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2399 + }, + { + "epoch": 0.9588493807431082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2400 + }, + { + "epoch": 0.9592489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2401 + }, + { + "epoch": 0.9596484218937276, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2402 + }, + { + "epoch": 0.9600479424690371, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2403 + }, + { + "epoch": 0.9604474630443468, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2404 + }, + { + "epoch": 0.9608469836196564, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2405 + }, + { + "epoch": 0.961246504194966, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2406 + }, + { + "epoch": 0.9616460247702757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2407 + }, + { + "epoch": 0.9620455453455853, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 2408 + }, + { + "epoch": 0.962445065920895, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2409 + }, + { + "epoch": 0.9628445864962045, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2410 + }, + { + "epoch": 0.9632441070715142, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2411 + }, + { + "epoch": 0.9636436276468238, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2412 + }, + { + "epoch": 0.9640431482221334, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2413 + }, + { + "epoch": 0.9644426687974431, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2414 + }, + { + "epoch": 0.9648421893727527, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2415 + }, + { + "epoch": 0.9652417099480624, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2416 + }, + { + "epoch": 0.9656412305233719, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2417 + }, + { + "epoch": 0.9660407510986816, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2418 + }, + { + "epoch": 0.9664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2419 + }, + { + "epoch": 0.9668397922493008, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2420 + }, + { + "epoch": 0.9672393128246105, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2421 + }, + { + "epoch": 0.9676388333999201, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2422 + }, + { + "epoch": 0.9680383539752297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2423 + }, + { + "epoch": 0.9684378745505393, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2424 + }, + { + "epoch": 0.968837395125849, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 2425 + }, + { + "epoch": 0.9692369157011587, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2426 + }, + { + "epoch": 0.9696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 2427 + }, + { + "epoch": 0.9700359568517779, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2428 + }, + { + "epoch": 0.9704354774270875, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2429 + }, + { + "epoch": 0.9708349980023971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2430 + }, + { + "epoch": 0.9712345185777067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2431 + }, + { + "epoch": 0.9716340391530164, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2432 + }, + { + "epoch": 0.9720335597283261, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2433 + }, + { + "epoch": 0.9724330803036356, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2434 + }, + { + "epoch": 0.9728326008789453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2435 + }, + { + "epoch": 0.9732321214542549, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2436 + }, + { + "epoch": 0.9736316420295645, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2437 + }, + { + "epoch": 0.9740311626048741, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2438 + }, + { + "epoch": 0.9744306831801838, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2439 + }, + { + "epoch": 0.9748302037554935, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2440 + }, + { + "epoch": 0.975229724330803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2441 + }, + { + "epoch": 0.9756292449061127, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2442 + }, + { + "epoch": 0.9760287654814223, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 2443 + }, + { + "epoch": 0.9764282860567319, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2444 + }, + { + "epoch": 0.9768278066320415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2445 + }, + { + "epoch": 0.9772273272073512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2446 + }, + { + "epoch": 0.9776268477826608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2447 + }, + { + "epoch": 0.9780263683579704, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2448 + }, + { + "epoch": 0.9784258889332801, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2449 + }, + { + "epoch": 0.9788254095085897, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2450 + }, + { + "epoch": 0.9792249300838993, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2451 + }, + { + "epoch": 0.979624450659209, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2452 + }, + { + "epoch": 0.9800239712345186, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2453 + }, + { + "epoch": 0.9804234918098282, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2454 + }, + { + "epoch": 0.9808230123851378, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2455 + }, + { + "epoch": 0.9812225329604475, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2456 + }, + { + "epoch": 0.981622053535757, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2457 + }, + { + "epoch": 0.9820215741110667, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2458 + }, + { + "epoch": 0.9824210946863764, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2459 + }, + { + "epoch": 0.982820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2460 + }, + { + "epoch": 0.9832201358369956, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2461 + }, + { + "epoch": 0.9836196564123052, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2462 + }, + { + "epoch": 0.9840191769876149, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2463 + }, + { + "epoch": 0.9844186975629244, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2464 + }, + { + "epoch": 0.9848182181382341, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2465 + }, + { + "epoch": 0.9852177387135438, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2466 + }, + { + "epoch": 0.9856172592888534, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2467 + }, + { + "epoch": 0.986016779864163, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2468 + }, + { + "epoch": 0.9864163004394726, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2469 + }, + { + "epoch": 0.9868158210147823, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2470 + }, + { + "epoch": 0.9872153415900919, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2471 + }, + { + "epoch": 0.9876148621654015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2472 + }, + { + "epoch": 0.9880143827407112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2473 + }, + { + "epoch": 0.9884139033160207, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2474 + }, + { + "epoch": 0.9888134238913304, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2475 + }, + { + "epoch": 0.98921294446664, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2476 + }, + { + "epoch": 0.9896124650419497, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2477 + }, + { + "epoch": 0.9900119856172593, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2478 + }, + { + "epoch": 0.9904115061925689, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2479 + }, + { + "epoch": 0.9908110267678786, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2480 + }, + { + "epoch": 0.9912105473431881, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2481 + }, + { + "epoch": 0.9916100679184978, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2482 + }, + { + "epoch": 0.9920095884938074, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2483 + }, + { + "epoch": 0.9924091090691171, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2484 + }, + { + "epoch": 0.9928086296444267, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2485 + }, + { + "epoch": 0.9932081502197363, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2486 + }, + { + "epoch": 0.993607670795046, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2487 + }, + { + "epoch": 0.9940071913703555, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 2488 + }, + { + "epoch": 0.9944067119456652, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2489 + }, + { + "epoch": 0.9948062325209748, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2490 + }, + { + "epoch": 0.9952057530962845, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2491 + }, + { + "epoch": 0.9956052736715941, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2492 + }, + { + "epoch": 0.9960047942469037, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2493 + }, + { + "epoch": 0.9964043148222134, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2494 + }, + { + "epoch": 0.9968038353975229, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2495 + }, + { + "epoch": 0.9972033559728326, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2496 + }, + { + "epoch": 0.9976028765481423, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2497 + }, + { + "epoch": 0.9980023971234518, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2498 + }, + { + "epoch": 0.9984019176987615, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2499 + }, + { + "epoch": 0.9988014382740711, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2500 + }, + { + "epoch": 0.9992009588493808, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2501 + }, + { + "epoch": 0.9996004794246903, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2502 + }, + { + "epoch": 1.0, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2503 + }, + { + "epoch": 1.0003995205753096, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2504 + }, + { + "epoch": 1.0007990411506194, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2505 + }, + { + "epoch": 1.001198561725929, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2506 + }, + { + "epoch": 1.0015980823012385, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2507 + }, + { + "epoch": 1.001997602876548, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2508 + }, + { + "epoch": 1.0023971234518578, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2509 + }, + { + "epoch": 1.0027966440271674, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2510 + }, + { + "epoch": 1.003196164602477, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2511 + }, + { + "epoch": 1.0035956851777867, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 2512 + }, + { + "epoch": 1.0039952057530963, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2513 + }, + { + "epoch": 1.0043947263284059, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2514 + }, + { + "epoch": 1.0047942469037154, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2515 + }, + { + "epoch": 1.0051937674790252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2516 + }, + { + "epoch": 1.0055932880543348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2517 + }, + { + "epoch": 1.0059928086296444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2518 + }, + { + "epoch": 1.0063923292049541, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2519 + }, + { + "epoch": 1.0067918497802637, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2520 + }, + { + "epoch": 1.0071913703555733, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2521 + }, + { + "epoch": 1.0075908909308828, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2522 + }, + { + "epoch": 1.0079904115061926, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2523 + }, + { + "epoch": 1.0083899320815022, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2524 + }, + { + "epoch": 1.0087894526568117, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 2525 + }, + { + "epoch": 1.0091889732321215, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2526 + }, + { + "epoch": 1.009588493807431, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2527 + }, + { + "epoch": 1.0099880143827407, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2528 + }, + { + "epoch": 1.0103875349580504, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2529 + }, + { + "epoch": 1.01078705553336, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2530 + }, + { + "epoch": 1.0111865761086696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2531 + }, + { + "epoch": 1.0115860966839791, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2532 + }, + { + "epoch": 1.011985617259289, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2533 + }, + { + "epoch": 1.0123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2534 + }, + { + "epoch": 1.012784658409908, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2535 + }, + { + "epoch": 1.0131841789852178, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 2536 + }, + { + "epoch": 1.0135836995605274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2537 + }, + { + "epoch": 1.013983220135837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2538 + }, + { + "epoch": 1.0143827407111465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2539 + }, + { + "epoch": 1.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2540 + }, + { + "epoch": 1.0151817818617659, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 2541 + }, + { + "epoch": 1.0155813024370755, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2542 + }, + { + "epoch": 1.0159808230123852, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2543 + }, + { + "epoch": 1.0163803435876948, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 2544 + }, + { + "epoch": 1.0167798641630044, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2545 + }, + { + "epoch": 1.017179384738314, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2546 + }, + { + "epoch": 1.0175789053136237, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2547 + }, + { + "epoch": 1.0179784258889333, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2548 + }, + { + "epoch": 1.0183779464642428, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2549 + }, + { + "epoch": 1.0187774670395526, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2550 + }, + { + "epoch": 1.0191769876148622, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2551 + }, + { + "epoch": 1.0195765081901718, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 2552 + }, + { + "epoch": 1.0199760287654813, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2553 + }, + { + "epoch": 1.020375549340791, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2554 + }, + { + "epoch": 1.0207750699161007, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2555 + }, + { + "epoch": 1.0211745904914102, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2556 + }, + { + "epoch": 1.02157411106672, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 2557 + }, + { + "epoch": 1.0219736316420296, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2558 + }, + { + "epoch": 1.0223731522173392, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2559 + }, + { + "epoch": 1.0227726727926487, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2560 + }, + { + "epoch": 1.0231721933679585, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2561 + }, + { + "epoch": 1.023571713943268, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 2562 + }, + { + "epoch": 1.0239712345185776, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2563 + }, + { + "epoch": 1.0243707550938874, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2564 + }, + { + "epoch": 1.024770275669197, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2565 + }, + { + "epoch": 1.0251697962445065, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2566 + }, + { + "epoch": 1.0255693168198161, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2567 + }, + { + "epoch": 1.025968837395126, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2568 + }, + { + "epoch": 1.0263683579704355, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2569 + }, + { + "epoch": 1.026767878545745, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2570 + }, + { + "epoch": 1.0271673991210548, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2571 + }, + { + "epoch": 1.0275669196963644, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2572 + }, + { + "epoch": 1.027966440271674, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2573 + }, + { + "epoch": 1.0283659608469837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2574 + }, + { + "epoch": 1.0287654814222933, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2575 + }, + { + "epoch": 1.0291650019976029, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 2576 + }, + { + "epoch": 1.0295645225729124, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2577 + }, + { + "epoch": 1.0299640431482222, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2578 + }, + { + "epoch": 1.0303635637235318, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 2579 + }, + { + "epoch": 1.0307630842988413, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2580 + }, + { + "epoch": 1.0311626048741511, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2581 + }, + { + "epoch": 1.0315621254494607, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2582 + }, + { + "epoch": 1.0319616460247703, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2583 + }, + { + "epoch": 1.0323611666000798, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2584 + }, + { + "epoch": 1.0327606871753896, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2585 + }, + { + "epoch": 1.0331602077506992, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2586 + }, + { + "epoch": 1.0335597283260087, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2587 + }, + { + "epoch": 1.0339592489013185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2588 + }, + { + "epoch": 1.034358769476628, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2589 + }, + { + "epoch": 1.0347582900519376, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2590 + }, + { + "epoch": 1.0351578106272472, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2591 + }, + { + "epoch": 1.035557331202557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 2592 + }, + { + "epoch": 1.0359568517778666, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2593 + }, + { + "epoch": 1.0363563723531761, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2594 + }, + { + "epoch": 1.036755892928486, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2595 + }, + { + "epoch": 1.0371554135037955, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2596 + }, + { + "epoch": 1.037554934079105, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 2597 + }, + { + "epoch": 1.0379544546544146, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2598 + }, + { + "epoch": 1.0383539752297244, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2599 + }, + { + "epoch": 1.038753495805034, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2600 + }, + { + "epoch": 1.0391530163803435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2601 + }, + { + "epoch": 1.0395525369556533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2602 + }, + { + "epoch": 1.0399520575309629, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2603 + }, + { + "epoch": 1.0403515781062724, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 2604 + }, + { + "epoch": 1.040751098681582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2605 + }, + { + "epoch": 1.0411506192568918, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2606 + }, + { + "epoch": 1.0415501398322013, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2607 + }, + { + "epoch": 1.041949660407511, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2608 + }, + { + "epoch": 1.0423491809828207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 2609 + }, + { + "epoch": 1.0427487015581303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2610 + }, + { + "epoch": 1.0431482221334398, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2611 + }, + { + "epoch": 1.0435477427087494, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2612 + }, + { + "epoch": 1.0439472632840592, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2613 + }, + { + "epoch": 1.0443467838593687, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2614 + }, + { + "epoch": 1.0447463044346783, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2615 + }, + { + "epoch": 1.045145825009988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2616 + }, + { + "epoch": 1.0455453455852977, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2617 + }, + { + "epoch": 1.0459448661606072, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2618 + }, + { + "epoch": 1.046344386735917, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2619 + }, + { + "epoch": 1.0467439073112266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2620 + }, + { + "epoch": 1.0471434278865361, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2621 + }, + { + "epoch": 1.0475429484618457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2622 + }, + { + "epoch": 1.0479424690371555, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2623 + }, + { + "epoch": 1.048341989612465, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2624 + }, + { + "epoch": 1.0487415101877746, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2625 + }, + { + "epoch": 1.0491410307630844, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 2626 + }, + { + "epoch": 1.049540551338394, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2627 + }, + { + "epoch": 1.0499400719137035, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2628 + }, + { + "epoch": 1.050339592489013, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2629 + }, + { + "epoch": 1.0507391130643229, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2630 + }, + { + "epoch": 1.0511386336396324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2631 + }, + { + "epoch": 1.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2632 + }, + { + "epoch": 1.0519376747902518, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2633 + }, + { + "epoch": 1.0523371953655614, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2634 + }, + { + "epoch": 1.052736715940871, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2635 + }, + { + "epoch": 1.0531362365161805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2636 + }, + { + "epoch": 1.0535357570914903, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2637 + }, + { + "epoch": 1.0539352776667998, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2638 + }, + { + "epoch": 1.0543347982421094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2639 + }, + { + "epoch": 1.0547343188174192, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2640 + }, + { + "epoch": 1.0551338393927288, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2641 + }, + { + "epoch": 1.0555333599680383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2642 + }, + { + "epoch": 1.0559328805433479, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2643 + }, + { + "epoch": 1.0563324011186577, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2644 + }, + { + "epoch": 1.0567319216939672, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2645 + }, + { + "epoch": 1.0571314422692768, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2646 + }, + { + "epoch": 1.0575309628445866, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2647 + }, + { + "epoch": 1.0579304834198962, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2648 + }, + { + "epoch": 1.0583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2649 + }, + { + "epoch": 1.0587295245705153, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2650 + }, + { + "epoch": 1.059129045145825, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 2651 + }, + { + "epoch": 1.0595285657211346, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2652 + }, + { + "epoch": 1.0599280862964442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2653 + }, + { + "epoch": 1.060327606871754, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2654 + }, + { + "epoch": 1.0607271274470635, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2655 + }, + { + "epoch": 1.061126648022373, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2656 + }, + { + "epoch": 1.0615261685976827, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2657 + }, + { + "epoch": 1.0619256891729925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2658 + }, + { + "epoch": 1.062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2659 + }, + { + "epoch": 1.0627247303236116, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2660 + }, + { + "epoch": 1.0631242508989214, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2661 + }, + { + "epoch": 1.063523771474231, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2662 + }, + { + "epoch": 1.0639232920495405, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2663 + }, + { + "epoch": 1.0643228126248503, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2664 + }, + { + "epoch": 1.0647223332001599, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2665 + }, + { + "epoch": 1.0651218537754694, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2666 + }, + { + "epoch": 1.065521374350779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2667 + }, + { + "epoch": 1.0659208949260888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2668 + }, + { + "epoch": 1.0663204155013983, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 2669 + }, + { + "epoch": 1.066719936076708, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2670 + }, + { + "epoch": 1.0671194566520175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2671 + }, + { + "epoch": 1.0675189772273272, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2672 + }, + { + "epoch": 1.0679184978026368, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2673 + }, + { + "epoch": 1.0683180183779464, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2674 + }, + { + "epoch": 1.0687175389532562, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2675 + }, + { + "epoch": 1.0691170595285657, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2676 + }, + { + "epoch": 1.0695165801038753, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2677 + }, + { + "epoch": 1.069916100679185, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 2678 + }, + { + "epoch": 1.0703156212544946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2679 + }, + { + "epoch": 1.0707151418298042, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2680 + }, + { + "epoch": 1.0711146624051138, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2681 + }, + { + "epoch": 1.0715141829804236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2682 + }, + { + "epoch": 1.0719137035557331, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2683 + }, + { + "epoch": 1.0723132241310427, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2684 + }, + { + "epoch": 1.0727127447063525, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2685 + }, + { + "epoch": 1.073112265281662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2686 + }, + { + "epoch": 1.0735117858569716, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2687 + }, + { + "epoch": 1.0739113064322812, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2688 + }, + { + "epoch": 1.074310827007591, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2689 + }, + { + "epoch": 1.0747103475829005, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2690 + }, + { + "epoch": 1.07510986815821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2691 + }, + { + "epoch": 1.0755093887335199, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2692 + }, + { + "epoch": 1.0759089093088294, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2693 + }, + { + "epoch": 1.076308429884139, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2694 + }, + { + "epoch": 1.0767079504594486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2695 + }, + { + "epoch": 1.0771074710347583, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2696 + }, + { + "epoch": 1.077506991610068, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2697 + }, + { + "epoch": 1.0779065121853775, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2698 + }, + { + "epoch": 1.0783060327606873, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 1.0787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2700 + }, + { + "epoch": 1.0791050739113064, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2701 + }, + { + "epoch": 1.079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 2702 + }, + { + "epoch": 1.0799041150619257, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2703 + }, + { + "epoch": 1.0803036356372353, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2704 + }, + { + "epoch": 1.0807031562125449, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2705 + }, + { + "epoch": 1.0811026767878547, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2706 + }, + { + "epoch": 1.0815021973631642, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2707 + }, + { + "epoch": 1.0819017179384738, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2708 + }, + { + "epoch": 1.0823012385137836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2709 + }, + { + "epoch": 1.0827007590890931, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2710 + }, + { + "epoch": 1.0831002796644027, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 2711 + }, + { + "epoch": 1.0834998002397123, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2712 + }, + { + "epoch": 1.083899320815022, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2713 + }, + { + "epoch": 1.0842988413903316, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2714 + }, + { + "epoch": 1.0846983619656412, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2715 + }, + { + "epoch": 1.0850978825409507, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2716 + }, + { + "epoch": 1.0854974031162605, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2717 + }, + { + "epoch": 1.08589692369157, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2718 + }, + { + "epoch": 1.0862964442668797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2719 + }, + { + "epoch": 1.0866959648421894, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2720 + }, + { + "epoch": 1.087095485417499, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2721 + }, + { + "epoch": 1.0874950059928086, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2722 + }, + { + "epoch": 1.0878945265681184, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2723 + }, + { + "epoch": 1.088294047143428, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2724 + }, + { + "epoch": 1.0886935677187375, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2725 + }, + { + "epoch": 1.089093088294047, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2726 + }, + { + "epoch": 1.0894926088693568, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2727 + }, + { + "epoch": 1.0898921294446664, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2728 + }, + { + "epoch": 1.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2729 + }, + { + "epoch": 1.0906911705952858, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2730 + }, + { + "epoch": 1.0910906911705953, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2731 + }, + { + "epoch": 1.0914902117459049, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2732 + }, + { + "epoch": 1.0918897323212144, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2733 + }, + { + "epoch": 1.0922892528965242, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2734 + }, + { + "epoch": 1.0926887734718338, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2735 + }, + { + "epoch": 1.0930882940471434, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2736 + }, + { + "epoch": 1.0934878146224531, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2737 + }, + { + "epoch": 1.0938873351977627, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2738 + }, + { + "epoch": 1.0942868557730723, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2739 + }, + { + "epoch": 1.0946863763483818, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2740 + }, + { + "epoch": 1.0950858969236916, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2741 + }, + { + "epoch": 1.0954854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2742 + }, + { + "epoch": 1.0958849380743108, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2743 + }, + { + "epoch": 1.0962844586496205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2744 + }, + { + "epoch": 1.09668397922493, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2745 + }, + { + "epoch": 1.0970834998002397, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2746 + }, + { + "epoch": 1.0974830203755492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2747 + }, + { + "epoch": 1.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2748 + }, + { + "epoch": 1.0982820615261686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2749 + }, + { + "epoch": 1.0986815821014781, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2750 + }, + { + "epoch": 1.099081102676788, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2751 + }, + { + "epoch": 1.0994806232520975, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2752 + }, + { + "epoch": 1.099880143827407, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2753 + }, + { + "epoch": 1.1002796644027169, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2754 + }, + { + "epoch": 1.1006791849780264, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2755 + }, + { + "epoch": 1.101078705553336, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 2756 + }, + { + "epoch": 1.1014782261286455, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2757 + }, + { + "epoch": 1.1018777467039553, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2758 + }, + { + "epoch": 1.102277267279265, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2759 + }, + { + "epoch": 1.1026767878545745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2760 + }, + { + "epoch": 1.103076308429884, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2761 + }, + { + "epoch": 1.1034758290051938, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2762 + }, + { + "epoch": 1.1038753495805034, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 2763 + }, + { + "epoch": 1.104274870155813, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2764 + }, + { + "epoch": 1.1046743907311227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2765 + }, + { + "epoch": 1.1050739113064323, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2766 + }, + { + "epoch": 1.1054734318817419, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2767 + }, + { + "epoch": 1.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2768 + }, + { + "epoch": 1.1062724730323612, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2769 + }, + { + "epoch": 1.1066719936076708, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2770 + }, + { + "epoch": 1.1070715141829803, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2771 + }, + { + "epoch": 1.1074710347582901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2772 + }, + { + "epoch": 1.1078705553335997, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2773 + }, + { + "epoch": 1.1082700759089092, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 2774 + }, + { + "epoch": 1.108669596484219, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2775 + }, + { + "epoch": 1.1090691170595286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2776 + }, + { + "epoch": 1.1094686376348382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2777 + }, + { + "epoch": 1.1098681582101477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2778 + }, + { + "epoch": 1.1102676787854575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2779 + }, + { + "epoch": 1.110667199360767, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2780 + }, + { + "epoch": 1.1110667199360766, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2781 + }, + { + "epoch": 1.1114662405113864, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2782 + }, + { + "epoch": 1.111865761086696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2783 + }, + { + "epoch": 1.1122652816620056, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2784 + }, + { + "epoch": 1.1126648022373151, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2785 + }, + { + "epoch": 1.113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2786 + }, + { + "epoch": 1.1134638433879345, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2787 + }, + { + "epoch": 1.113863363963244, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2788 + }, + { + "epoch": 1.1142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2789 + }, + { + "epoch": 1.1146624051138634, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2790 + }, + { + "epoch": 1.115061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2791 + }, + { + "epoch": 1.1154614462644825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2792 + }, + { + "epoch": 1.1158609668397923, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2793 + }, + { + "epoch": 1.1162604874151019, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2794 + }, + { + "epoch": 1.1166600079904114, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2795 + }, + { + "epoch": 1.1170595285657212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2796 + }, + { + "epoch": 1.1174590491410308, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2797 + }, + { + "epoch": 1.1178585697163403, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2798 + }, + { + "epoch": 1.1182580902916501, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2799 + }, + { + "epoch": 1.1186576108669597, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2800 + }, + { + "epoch": 1.1190571314422693, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2801 + }, + { + "epoch": 1.1194566520175788, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2802 + }, + { + "epoch": 1.1198561725928886, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2803 + }, + { + "epoch": 1.1202556931681982, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2804 + }, + { + "epoch": 1.1206552137435077, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2805 + }, + { + "epoch": 1.1210547343188173, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2806 + }, + { + "epoch": 1.121454254894127, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2807 + }, + { + "epoch": 1.1218537754694367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2808 + }, + { + "epoch": 1.1222532960447462, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2809 + }, + { + "epoch": 1.122652816620056, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2810 + }, + { + "epoch": 1.1230523371953656, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2811 + }, + { + "epoch": 1.1234518577706751, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2812 + }, + { + "epoch": 1.123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2813 + }, + { + "epoch": 1.1242508989212945, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2814 + }, + { + "epoch": 1.124650419496604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2815 + }, + { + "epoch": 1.1250499400719136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2816 + }, + { + "epoch": 1.1254494606472234, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2817 + }, + { + "epoch": 1.125848981222533, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2818 + }, + { + "epoch": 1.1262485017978425, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2819 + }, + { + "epoch": 1.1266480223731523, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2820 + }, + { + "epoch": 1.1270475429484619, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2821 + }, + { + "epoch": 1.1274470635237714, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2822 + }, + { + "epoch": 1.127846584099081, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 2823 + }, + { + "epoch": 1.1282461046743908, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 2824 + }, + { + "epoch": 1.1286456252497004, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2825 + }, + { + "epoch": 1.12904514582501, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2826 + }, + { + "epoch": 1.1294446664003197, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2827 + }, + { + "epoch": 1.1298441869756293, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2828 + }, + { + "epoch": 1.1302437075509388, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2829 + }, + { + "epoch": 1.1306432281262486, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2830 + }, + { + "epoch": 1.1310427487015582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2831 + }, + { + "epoch": 1.1314422692768678, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2832 + }, + { + "epoch": 1.1318417898521773, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2833 + }, + { + "epoch": 1.132241310427487, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2834 + }, + { + "epoch": 1.1326408310027967, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2835 + }, + { + "epoch": 1.1330403515781062, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2836 + }, + { + "epoch": 1.1334398721534158, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2837 + }, + { + "epoch": 1.1338393927287256, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2838 + }, + { + "epoch": 1.1342389133040351, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2839 + }, + { + "epoch": 1.1346384338793447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2840 + }, + { + "epoch": 1.1350379544546545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2841 + }, + { + "epoch": 1.135437475029964, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2842 + }, + { + "epoch": 1.1358369956052736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2843 + }, + { + "epoch": 1.1362365161805834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2844 + }, + { + "epoch": 1.136636036755893, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2845 + }, + { + "epoch": 1.1370355573312025, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2846 + }, + { + "epoch": 1.137435077906512, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2847 + }, + { + "epoch": 1.137834598481822, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2848 + }, + { + "epoch": 1.1382341190571315, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2849 + }, + { + "epoch": 1.138633639632441, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2850 + }, + { + "epoch": 1.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2851 + }, + { + "epoch": 1.1394326807830604, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2852 + }, + { + "epoch": 1.13983220135837, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2853 + }, + { + "epoch": 1.1402317219336795, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2854 + }, + { + "epoch": 1.1406312425089893, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2855 + }, + { + "epoch": 1.1410307630842988, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2856 + }, + { + "epoch": 1.1414302836596084, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2857 + }, + { + "epoch": 1.1418298042349182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2858 + }, + { + "epoch": 1.1422293248102278, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2859 + }, + { + "epoch": 1.1426288453855373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2860 + }, + { + "epoch": 1.143028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2861 + }, + { + "epoch": 1.1434278865361567, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2862 + }, + { + "epoch": 1.1438274071114662, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2863 + }, + { + "epoch": 1.1442269276867758, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2864 + }, + { + "epoch": 1.1446264482620856, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 2865 + }, + { + "epoch": 1.1450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2866 + }, + { + "epoch": 1.1454254894127047, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2867 + }, + { + "epoch": 1.1458250099880143, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 2868 + }, + { + "epoch": 1.146224530563324, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2869 + }, + { + "epoch": 1.1466240511386336, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2870 + }, + { + "epoch": 1.1470235717139432, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2871 + }, + { + "epoch": 1.147423092289253, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 2872 + }, + { + "epoch": 1.1478226128645626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2873 + }, + { + "epoch": 1.1482221334398721, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2874 + }, + { + "epoch": 1.148621654015182, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2875 + }, + { + "epoch": 1.1490211745904915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2876 + }, + { + "epoch": 1.149420695165801, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2877 + }, + { + "epoch": 1.1498202157411106, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 2878 + }, + { + "epoch": 1.1502197363164204, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2879 + }, + { + "epoch": 1.15061925689173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2880 + }, + { + "epoch": 1.1510187774670395, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2881 + }, + { + "epoch": 1.151418298042349, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2882 + }, + { + "epoch": 1.1518178186176589, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2883 + }, + { + "epoch": 1.1522173391929684, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2884 + }, + { + "epoch": 1.152616859768278, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2885 + }, + { + "epoch": 1.1530163803435878, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2886 + }, + { + "epoch": 1.1534159009188973, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2887 + }, + { + "epoch": 1.153815421494207, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 2888 + }, + { + "epoch": 1.1542149420695167, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2889 + }, + { + "epoch": 1.1546144626448263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2890 + }, + { + "epoch": 1.1550139832201358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 2891 + }, + { + "epoch": 1.1554135037954454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2892 + }, + { + "epoch": 1.1558130243707552, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2893 + }, + { + "epoch": 1.1562125449460647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2894 + }, + { + "epoch": 1.1566120655213743, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2895 + }, + { + "epoch": 1.1570115860966839, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2896 + }, + { + "epoch": 1.1574111066719937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2897 + }, + { + "epoch": 1.1578106272473032, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 2898 + }, + { + "epoch": 1.1582101478226128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2899 + }, + { + "epoch": 1.1586096683979226, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2900 + }, + { + "epoch": 1.1590091889732321, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2901 + }, + { + "epoch": 1.1594087095485417, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2902 + }, + { + "epoch": 1.1598082301238515, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2903 + }, + { + "epoch": 1.160207750699161, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2904 + }, + { + "epoch": 1.1606072712744706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2905 + }, + { + "epoch": 1.1610067918497802, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2906 + }, + { + "epoch": 1.16140631242509, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 2907 + }, + { + "epoch": 1.1618058330003995, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2908 + }, + { + "epoch": 1.162205353575709, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2909 + }, + { + "epoch": 1.1626048741510189, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2910 + }, + { + "epoch": 1.1630043947263284, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2911 + }, + { + "epoch": 1.163403915301638, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2912 + }, + { + "epoch": 1.1638034358769476, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2913 + }, + { + "epoch": 1.1642029564522574, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2914 + }, + { + "epoch": 1.164602477027567, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2915 + }, + { + "epoch": 1.1650019976028765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2916 + }, + { + "epoch": 1.1654015181781863, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2917 + }, + { + "epoch": 1.1658010387534958, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2918 + }, + { + "epoch": 1.1662005593288054, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2919 + }, + { + "epoch": 1.1666000799041152, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2920 + }, + { + "epoch": 1.1669996004794247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2921 + }, + { + "epoch": 1.1673991210547343, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2922 + }, + { + "epoch": 1.1677986416300439, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2923 + }, + { + "epoch": 1.1681981622053537, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2924 + }, + { + "epoch": 1.1685976827806632, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2925 + }, + { + "epoch": 1.1689972033559728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2926 + }, + { + "epoch": 1.1693967239312824, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2927 + }, + { + "epoch": 1.1697962445065921, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2928 + }, + { + "epoch": 1.1701957650819017, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2929 + }, + { + "epoch": 1.1705952856572113, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2930 + }, + { + "epoch": 1.170994806232521, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2931 + }, + { + "epoch": 1.1713943268078306, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2932 + }, + { + "epoch": 1.1717938473831402, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2933 + }, + { + "epoch": 1.17219336795845, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 2934 + }, + { + "epoch": 1.1725928885337595, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2935 + }, + { + "epoch": 1.172992409109069, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2936 + }, + { + "epoch": 1.1733919296843787, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2937 + }, + { + "epoch": 1.1737914502596885, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2938 + }, + { + "epoch": 1.174190970834998, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2939 + }, + { + "epoch": 1.1745904914103076, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2940 + }, + { + "epoch": 1.1749900119856171, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2941 + }, + { + "epoch": 1.175389532560927, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2942 + }, + { + "epoch": 1.1757890531362365, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2943 + }, + { + "epoch": 1.176188573711546, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2944 + }, + { + "epoch": 1.1765880942868558, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2945 + }, + { + "epoch": 1.1769876148621654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2946 + }, + { + "epoch": 1.177387135437475, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2947 + }, + { + "epoch": 1.1777866560127848, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 2948 + }, + { + "epoch": 1.1781861765880943, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2949 + }, + { + "epoch": 1.178585697163404, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2950 + }, + { + "epoch": 1.1789852177387135, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2951 + }, + { + "epoch": 1.1793847383140232, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2952 + }, + { + "epoch": 1.1797842588893328, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2953 + }, + { + "epoch": 1.1801837794646424, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2954 + }, + { + "epoch": 1.180583300039952, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2955 + }, + { + "epoch": 1.1809828206152617, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2956 + }, + { + "epoch": 1.1813823411905713, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2957 + }, + { + "epoch": 1.1817818617658808, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2958 + }, + { + "epoch": 1.1821813823411906, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2959 + }, + { + "epoch": 1.1825809029165002, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2960 + }, + { + "epoch": 1.1829804234918098, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2961 + }, + { + "epoch": 1.1833799440671195, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 2962 + }, + { + "epoch": 1.1837794646424291, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2963 + }, + { + "epoch": 1.1841789852177387, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 2964 + }, + { + "epoch": 1.1845785057930485, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2965 + }, + { + "epoch": 1.184978026368358, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2966 + }, + { + "epoch": 1.1853775469436676, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2967 + }, + { + "epoch": 1.1857770675189772, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2968 + }, + { + "epoch": 1.186176588094287, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2969 + }, + { + "epoch": 1.1865761086695965, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 2970 + }, + { + "epoch": 1.186975629244906, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2971 + }, + { + "epoch": 1.1873751498202156, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2972 + }, + { + "epoch": 1.1877746703955254, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 2973 + }, + { + "epoch": 1.188174190970835, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2974 + }, + { + "epoch": 1.1885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2975 + }, + { + "epoch": 1.1889732321214543, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2976 + }, + { + "epoch": 1.189372752696764, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2977 + }, + { + "epoch": 1.1897722732720735, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2978 + }, + { + "epoch": 1.1901717938473833, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2979 + }, + { + "epoch": 1.1905713144226928, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2980 + }, + { + "epoch": 1.1909708349980024, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2981 + }, + { + "epoch": 1.191370355573312, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2982 + }, + { + "epoch": 1.1917698761486217, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2983 + }, + { + "epoch": 1.1921693967239313, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2984 + }, + { + "epoch": 1.1925689172992409, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 2985 + }, + { + "epoch": 1.1929684378745504, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2986 + }, + { + "epoch": 1.1933679584498602, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2987 + }, + { + "epoch": 1.1937674790251698, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2988 + }, + { + "epoch": 1.1941669996004793, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2989 + }, + { + "epoch": 1.1945665201757891, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2990 + }, + { + "epoch": 1.1949660407510987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2991 + }, + { + "epoch": 1.1953655613264083, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2992 + }, + { + "epoch": 1.195765081901718, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2993 + }, + { + "epoch": 1.1961646024770276, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2994 + }, + { + "epoch": 1.1965641230523372, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2995 + }, + { + "epoch": 1.1969636436276467, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2996 + }, + { + "epoch": 1.1973631642029565, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2997 + }, + { + "epoch": 1.197762684778266, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2998 + }, + { + "epoch": 1.1981622053535756, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2999 + }, + { + "epoch": 1.1985617259288852, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3000 + }, + { + "epoch": 1.198961246504195, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 3001 + }, + { + "epoch": 1.1993607670795046, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3002 + }, + { + "epoch": 1.1997602876548141, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3003 + }, + { + "epoch": 1.200159808230124, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 3004 + }, + { + "epoch": 1.2005593288054335, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3005 + }, + { + "epoch": 1.200958849380743, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3006 + }, + { + "epoch": 1.2013583699560528, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3007 + }, + { + "epoch": 1.2017578905313624, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3008 + }, + { + "epoch": 1.202157411106672, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3009 + }, + { + "epoch": 1.2025569316819817, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 3010 + }, + { + "epoch": 1.2029564522572913, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 3011 + }, + { + "epoch": 1.2033559728326009, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3012 + }, + { + "epoch": 1.2037554934079104, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3013 + }, + { + "epoch": 1.2041550139832202, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3014 + }, + { + "epoch": 1.2045545345585298, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3015 + }, + { + "epoch": 1.2049540551338394, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3016 + }, + { + "epoch": 1.205353575709149, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3017 + }, + { + "epoch": 1.2057530962844587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3018 + }, + { + "epoch": 1.2061526168597683, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3019 + }, + { + "epoch": 1.2065521374350778, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3020 + }, + { + "epoch": 1.2069516580103876, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3021 + }, + { + "epoch": 1.2073511785856972, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3022 + }, + { + "epoch": 1.2077506991610067, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3023 + }, + { + "epoch": 1.2081502197363165, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3024 + }, + { + "epoch": 1.208549740311626, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3025 + }, + { + "epoch": 1.2089492608869357, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3026 + }, + { + "epoch": 1.2093487814622452, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3027 + }, + { + "epoch": 1.209748302037555, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3028 + }, + { + "epoch": 1.2101478226128646, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 3029 + }, + { + "epoch": 1.2105473431881741, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3030 + }, + { + "epoch": 1.2109468637634837, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3031 + }, + { + "epoch": 1.2113463843387935, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3032 + }, + { + "epoch": 1.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3033 + }, + { + "epoch": 1.2121454254894126, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3034 + }, + { + "epoch": 1.2125449460647224, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3035 + }, + { + "epoch": 1.212944466640032, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3036 + }, + { + "epoch": 1.2133439872153415, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3037 + }, + { + "epoch": 1.2137435077906513, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3038 + }, + { + "epoch": 1.2141430283659609, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 3039 + }, + { + "epoch": 1.2145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3040 + }, + { + "epoch": 1.21494206951658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3041 + }, + { + "epoch": 1.2153415900918898, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3042 + }, + { + "epoch": 1.2157411106671994, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3043 + }, + { + "epoch": 1.216140631242509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3044 + }, + { + "epoch": 1.2165401518178185, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3045 + }, + { + "epoch": 1.2169396723931283, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3046 + }, + { + "epoch": 1.2173391929684378, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3047 + }, + { + "epoch": 1.2177387135437474, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3048 + }, + { + "epoch": 1.2181382341190572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3049 + }, + { + "epoch": 1.2185377546943668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3050 + }, + { + "epoch": 1.2189372752696763, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 3051 + }, + { + "epoch": 1.219336795844986, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3052 + }, + { + "epoch": 1.2197363164202957, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3053 + }, + { + "epoch": 1.2201358369956052, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3054 + }, + { + "epoch": 1.220535357570915, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 3055 + }, + { + "epoch": 1.2209348781462246, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3056 + }, + { + "epoch": 1.2213343987215342, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3057 + }, + { + "epoch": 1.2217339192968437, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3058 + }, + { + "epoch": 1.2221334398721535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3059 + }, + { + "epoch": 1.222532960447463, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3060 + }, + { + "epoch": 1.2229324810227726, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3061 + }, + { + "epoch": 1.2233320015980822, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3062 + }, + { + "epoch": 1.223731522173392, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3063 + }, + { + "epoch": 1.2241310427487015, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3064 + }, + { + "epoch": 1.2245305633240111, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3065 + }, + { + "epoch": 1.224930083899321, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3066 + }, + { + "epoch": 1.2253296044746305, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3067 + }, + { + "epoch": 1.22572912504994, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3068 + }, + { + "epoch": 1.2261286456252498, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3069 + }, + { + "epoch": 1.2265281662005594, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3070 + }, + { + "epoch": 1.226927686775869, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 3071 + }, + { + "epoch": 1.2273272073511785, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3072 + }, + { + "epoch": 1.2277267279264883, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 3073 + }, + { + "epoch": 1.2281262485017979, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3074 + }, + { + "epoch": 1.2285257690771074, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3075 + }, + { + "epoch": 1.228925289652417, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 3076 + }, + { + "epoch": 1.2293248102277268, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3077 + }, + { + "epoch": 1.2297243308030363, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3078 + }, + { + "epoch": 1.230123851378346, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3079 + }, + { + "epoch": 1.2305233719536557, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3080 + }, + { + "epoch": 1.2309228925289653, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3081 + }, + { + "epoch": 1.2313224131042748, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3082 + }, + { + "epoch": 1.2317219336795846, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3083 + }, + { + "epoch": 1.2321214542548942, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3084 + }, + { + "epoch": 1.2325209748302037, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3085 + }, + { + "epoch": 1.2329204954055133, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3086 + }, + { + "epoch": 1.233320015980823, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3087 + }, + { + "epoch": 1.2337195365561326, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3088 + }, + { + "epoch": 1.2341190571314422, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3089 + }, + { + "epoch": 1.2345185777067518, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3090 + }, + { + "epoch": 1.2349180982820616, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3091 + }, + { + "epoch": 1.2353176188573711, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3092 + }, + { + "epoch": 1.2357171394326807, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3093 + }, + { + "epoch": 1.2361166600079905, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3094 + }, + { + "epoch": 1.2365161805833, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3095 + }, + { + "epoch": 1.2369157011586096, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3096 + }, + { + "epoch": 1.2373152217339194, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3097 + }, + { + "epoch": 1.237714742309229, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3098 + }, + { + "epoch": 1.2381142628845385, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3099 + }, + { + "epoch": 1.2385137834598483, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 3100 + }, + { + "epoch": 1.2389133040351579, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3101 + }, + { + "epoch": 1.2393128246104674, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3102 + }, + { + "epoch": 1.239712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3103 + }, + { + "epoch": 1.2401118657610868, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3104 + }, + { + "epoch": 1.2405113863363963, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3105 + }, + { + "epoch": 1.240910906911706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3106 + }, + { + "epoch": 1.2413104274870155, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3107 + }, + { + "epoch": 1.2417099480623253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3108 + }, + { + "epoch": 1.2421094686376348, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3109 + }, + { + "epoch": 1.2425089892129444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3110 + }, + { + "epoch": 1.2429085097882542, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3111 + }, + { + "epoch": 1.2433080303635637, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3112 + }, + { + "epoch": 1.2437075509388733, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3113 + }, + { + "epoch": 1.244107071514183, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3114 + }, + { + "epoch": 1.2445065920894927, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3115 + }, + { + "epoch": 1.2449061126648022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3116 + }, + { + "epoch": 1.2453056332401118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3117 + }, + { + "epoch": 1.2457051538154216, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3118 + }, + { + "epoch": 1.2461046743907311, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3119 + }, + { + "epoch": 1.2465041949660407, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3120 + }, + { + "epoch": 1.2469037155413503, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3121 + }, + { + "epoch": 1.24730323611666, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3122 + }, + { + "epoch": 1.2477027566919696, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3123 + }, + { + "epoch": 1.2481022772672792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3124 + }, + { + "epoch": 1.248501797842589, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3125 + }, + { + "epoch": 1.2489013184178985, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3126 + }, + { + "epoch": 1.249300838993208, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3127 + }, + { + "epoch": 1.2497003595685179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3128 + }, + { + "epoch": 1.2500998801438274, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 3129 + }, + { + "epoch": 1.250499400719137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3130 + }, + { + "epoch": 1.2508989212944468, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3131 + }, + { + "epoch": 1.2512984418697564, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3132 + }, + { + "epoch": 1.251697962445066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3133 + }, + { + "epoch": 1.2520974830203755, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3134 + }, + { + "epoch": 1.252497003595685, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3135 + }, + { + "epoch": 1.2528965241709948, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3136 + }, + { + "epoch": 1.2532960447463044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3137 + }, + { + "epoch": 1.253695565321614, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3138 + }, + { + "epoch": 1.2540950858969238, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3139 + }, + { + "epoch": 1.2544946064722333, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3140 + }, + { + "epoch": 1.2548941270475429, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3141 + }, + { + "epoch": 1.2552936476228527, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3142 + }, + { + "epoch": 1.2556931681981622, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3143 + }, + { + "epoch": 1.2560926887734718, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3144 + }, + { + "epoch": 1.2564922093487816, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3145 + }, + { + "epoch": 1.2568917299240912, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3146 + }, + { + "epoch": 1.2572912504994007, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3147 + }, + { + "epoch": 1.2576907710747103, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3148 + }, + { + "epoch": 1.2580902916500198, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3149 + }, + { + "epoch": 1.2584898122253296, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3150 + }, + { + "epoch": 1.2588893328006392, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 3151 + }, + { + "epoch": 1.2592888533759488, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3152 + }, + { + "epoch": 1.2596883739512585, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3153 + }, + { + "epoch": 1.260087894526568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3154 + }, + { + "epoch": 1.2604874151018777, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3155 + }, + { + "epoch": 1.2608869356771875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3156 + }, + { + "epoch": 1.261286456252497, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3157 + }, + { + "epoch": 1.2616859768278066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3158 + }, + { + "epoch": 1.2620854974031164, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3159 + }, + { + "epoch": 1.262485017978426, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3160 + }, + { + "epoch": 1.2628845385537355, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3161 + }, + { + "epoch": 1.263284059129045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3162 + }, + { + "epoch": 1.2636835797043549, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3163 + }, + { + "epoch": 1.2640831002796644, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3164 + }, + { + "epoch": 1.264482620854974, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3165 + }, + { + "epoch": 1.2648821414302835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3166 + }, + { + "epoch": 1.2652816620055933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3167 + }, + { + "epoch": 1.265681182580903, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3168 + }, + { + "epoch": 1.2660807031562125, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3169 + }, + { + "epoch": 1.2664802237315222, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3170 + }, + { + "epoch": 1.2668797443068318, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 3171 + }, + { + "epoch": 1.2672792648821414, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3172 + }, + { + "epoch": 1.2676787854574512, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3173 + }, + { + "epoch": 1.2680783060327607, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3174 + }, + { + "epoch": 1.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3175 + }, + { + "epoch": 1.26887734718338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3176 + }, + { + "epoch": 1.2692768677586896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3177 + }, + { + "epoch": 1.2696763883339992, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3178 + }, + { + "epoch": 1.2700759089093088, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3179 + }, + { + "epoch": 1.2704754294846183, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3180 + }, + { + "epoch": 1.2708749500599281, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3181 + }, + { + "epoch": 1.2712744706352377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3182 + }, + { + "epoch": 1.2716739912105472, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3183 + }, + { + "epoch": 1.272073511785857, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3184 + }, + { + "epoch": 1.2724730323611666, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3185 + }, + { + "epoch": 1.2728725529364762, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3186 + }, + { + "epoch": 1.273272073511786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3187 + }, + { + "epoch": 1.2736715940870955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3188 + }, + { + "epoch": 1.274071114662405, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3189 + }, + { + "epoch": 1.2744706352377149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3190 + }, + { + "epoch": 1.2748701558130244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3191 + }, + { + "epoch": 1.275269676388334, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3192 + }, + { + "epoch": 1.2756691969636436, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3193 + }, + { + "epoch": 1.2760687175389531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3194 + }, + { + "epoch": 1.276468238114263, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3195 + }, + { + "epoch": 1.2768677586895725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3196 + }, + { + "epoch": 1.277267279264882, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3197 + }, + { + "epoch": 1.2776667998401918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3198 + }, + { + "epoch": 1.2780663204155014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3199 + }, + { + "epoch": 1.278465840990811, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3200 + }, + { + "epoch": 1.2788653615661207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3201 + }, + { + "epoch": 1.2792648821414303, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3202 + }, + { + "epoch": 1.2796644027167399, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3203 + }, + { + "epoch": 1.2800639232920497, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 3204 + }, + { + "epoch": 1.2804634438673592, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3205 + }, + { + "epoch": 1.2808629644426688, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 3206 + }, + { + "epoch": 1.2812624850179783, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3207 + }, + { + "epoch": 1.2816620055932881, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 3208 + }, + { + "epoch": 1.2820615261685977, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3209 + }, + { + "epoch": 1.2824610467439073, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3210 + }, + { + "epoch": 1.2828605673192168, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3211 + }, + { + "epoch": 1.2832600878945266, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3212 + }, + { + "epoch": 1.2836596084698362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3213 + }, + { + "epoch": 1.2840591290451457, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3214 + }, + { + "epoch": 1.2844586496204555, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3215 + }, + { + "epoch": 1.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3216 + }, + { + "epoch": 1.2852576907710747, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3217 + }, + { + "epoch": 1.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3218 + }, + { + "epoch": 1.286056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3219 + }, + { + "epoch": 1.2864562524970036, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 3220 + }, + { + "epoch": 1.2868557730723134, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3221 + }, + { + "epoch": 1.287255293647623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3222 + }, + { + "epoch": 1.2876548142229325, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3223 + }, + { + "epoch": 1.288054334798242, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3224 + }, + { + "epoch": 1.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3225 + }, + { + "epoch": 1.2888533759488614, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3226 + }, + { + "epoch": 1.289252896524171, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3227 + }, + { + "epoch": 1.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3228 + }, + { + "epoch": 1.2900519376747903, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3229 + }, + { + "epoch": 1.2904514582500999, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3230 + }, + { + "epoch": 1.2908509788254094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3231 + }, + { + "epoch": 1.2912504994007192, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3232 + }, + { + "epoch": 1.2916500199760288, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3233 + }, + { + "epoch": 1.2920495405513384, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3234 + }, + { + "epoch": 1.2924490611266481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3235 + }, + { + "epoch": 1.2928485817019577, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3236 + }, + { + "epoch": 1.2932481022772673, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3237 + }, + { + "epoch": 1.2936476228525768, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3238 + }, + { + "epoch": 1.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3239 + }, + { + "epoch": 1.2944466640031962, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 3240 + }, + { + "epoch": 1.2948461845785058, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3241 + }, + { + "epoch": 1.2952457051538153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3242 + }, + { + "epoch": 1.295645225729125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3243 + }, + { + "epoch": 1.2960447463044347, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3244 + }, + { + "epoch": 1.2964442668797442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 3245 + }, + { + "epoch": 1.296843787455054, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3246 + }, + { + "epoch": 1.2972433080303636, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3247 + }, + { + "epoch": 1.2976428286056731, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3248 + }, + { + "epoch": 1.298042349180983, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3249 + }, + { + "epoch": 1.2984418697562925, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3250 + }, + { + "epoch": 1.298841390331602, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3251 + }, + { + "epoch": 1.2992409109069116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3252 + }, + { + "epoch": 1.2996404314822214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3253 + }, + { + "epoch": 1.300039952057531, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3254 + }, + { + "epoch": 1.3004394726328405, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3255 + }, + { + "epoch": 1.30083899320815, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3256 + }, + { + "epoch": 1.30123851378346, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3257 + }, + { + "epoch": 1.3016380343587695, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3258 + }, + { + "epoch": 1.302037554934079, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3259 + }, + { + "epoch": 1.3024370755093888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3260 + }, + { + "epoch": 1.3028365960846984, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3261 + }, + { + "epoch": 1.303236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3262 + }, + { + "epoch": 1.3036356372353177, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3263 + }, + { + "epoch": 1.3040351578106273, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3264 + }, + { + "epoch": 1.3044346783859369, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3265 + }, + { + "epoch": 1.3048341989612466, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3266 + }, + { + "epoch": 1.3052337195365562, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3267 + }, + { + "epoch": 1.3056332401118658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3268 + }, + { + "epoch": 1.3060327606871753, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3269 + }, + { + "epoch": 1.306432281262485, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3270 + }, + { + "epoch": 1.3068318018377947, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3271 + }, + { + "epoch": 1.3072313224131042, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3272 + }, + { + "epoch": 1.3076308429884138, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3273 + }, + { + "epoch": 1.3080303635637236, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3274 + }, + { + "epoch": 1.3084298841390332, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3275 + }, + { + "epoch": 1.3088294047143427, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3276 + }, + { + "epoch": 1.3092289252896525, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3277 + }, + { + "epoch": 1.309628445864962, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3278 + }, + { + "epoch": 1.3100279664402716, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3279 + }, + { + "epoch": 1.3104274870155814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3280 + }, + { + "epoch": 1.310827007590891, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3281 + }, + { + "epoch": 1.3112265281662006, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 3282 + }, + { + "epoch": 1.3116260487415101, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3283 + }, + { + "epoch": 1.3120255693168197, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3284 + }, + { + "epoch": 1.3124250898921295, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3285 + }, + { + "epoch": 1.312824610467439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3286 + }, + { + "epoch": 1.3132241310427486, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3287 + }, + { + "epoch": 1.3136236516180584, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3288 + }, + { + "epoch": 1.314023172193368, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3289 + }, + { + "epoch": 1.3144226927686775, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3290 + }, + { + "epoch": 1.3148222133439873, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3291 + }, + { + "epoch": 1.3152217339192969, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3292 + }, + { + "epoch": 1.3156212544946064, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3293 + }, + { + "epoch": 1.3160207750699162, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3294 + }, + { + "epoch": 1.3164202956452258, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3295 + }, + { + "epoch": 1.3168198162205353, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3296 + }, + { + "epoch": 1.317219336795845, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3297 + }, + { + "epoch": 1.3176188573711547, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3298 + }, + { + "epoch": 1.3180183779464643, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3299 + }, + { + "epoch": 1.3184178985217738, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3300 + }, + { + "epoch": 1.3188174190970834, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3301 + }, + { + "epoch": 1.3192169396723932, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3302 + }, + { + "epoch": 1.3196164602477027, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3303 + }, + { + "epoch": 1.3200159808230123, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3304 + }, + { + "epoch": 1.320415501398322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3305 + }, + { + "epoch": 1.3208150219736317, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3306 + }, + { + "epoch": 1.3212145425489412, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3307 + }, + { + "epoch": 1.321614063124251, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3308 + }, + { + "epoch": 1.3220135836995606, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3309 + }, + { + "epoch": 1.3224131042748701, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3310 + }, + { + "epoch": 1.32281262485018, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3311 + }, + { + "epoch": 1.3232121454254895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3312 + }, + { + "epoch": 1.323611666000799, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3313 + }, + { + "epoch": 1.3240111865761086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3314 + }, + { + "epoch": 1.3244107071514182, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3315 + }, + { + "epoch": 1.324810227726728, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3316 + }, + { + "epoch": 1.3252097483020375, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3317 + }, + { + "epoch": 1.325609268877347, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3318 + }, + { + "epoch": 1.3260087894526569, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3319 + }, + { + "epoch": 1.3264083100279664, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3320 + }, + { + "epoch": 1.326807830603276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3321 + }, + { + "epoch": 1.3272073511785858, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3322 + }, + { + "epoch": 1.3276068717538954, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3323 + }, + { + "epoch": 1.328006392329205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3324 + }, + { + "epoch": 1.3284059129045147, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3325 + }, + { + "epoch": 1.3288054334798243, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3326 + }, + { + "epoch": 1.3292049540551338, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3327 + }, + { + "epoch": 1.3296044746304434, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3328 + }, + { + "epoch": 1.330003995205753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3329 + }, + { + "epoch": 1.3304035157810628, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3330 + }, + { + "epoch": 1.3308030363563723, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3331 + }, + { + "epoch": 1.3312025569316819, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3332 + }, + { + "epoch": 1.3316020775069917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3333 + }, + { + "epoch": 1.3320015980823012, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3334 + }, + { + "epoch": 1.3324011186576108, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3335 + }, + { + "epoch": 1.3328006392329206, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3336 + }, + { + "epoch": 1.3332001598082301, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3337 + }, + { + "epoch": 1.3335996803835397, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3338 + }, + { + "epoch": 1.3339992009588495, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3339 + }, + { + "epoch": 1.334398721534159, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3340 + }, + { + "epoch": 1.3347982421094686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3341 + }, + { + "epoch": 1.3351977626847782, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3342 + }, + { + "epoch": 1.335597283260088, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 3343 + }, + { + "epoch": 1.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3344 + }, + { + "epoch": 1.336396324410707, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3345 + }, + { + "epoch": 1.3367958449860167, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 3346 + }, + { + "epoch": 1.3371953655613265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3347 + }, + { + "epoch": 1.337594886136636, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3348 + }, + { + "epoch": 1.3379944067119456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3349 + }, + { + "epoch": 1.3383939272872554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3350 + }, + { + "epoch": 1.338793447862565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3351 + }, + { + "epoch": 1.3391929684378745, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3352 + }, + { + "epoch": 1.3395924890131843, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3353 + }, + { + "epoch": 1.3399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3354 + }, + { + "epoch": 1.3403915301638034, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 3355 + }, + { + "epoch": 1.3407910507391132, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3356 + }, + { + "epoch": 1.3411905713144228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3357 + }, + { + "epoch": 1.3415900918897323, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3358 + }, + { + "epoch": 1.341989612465042, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3359 + }, + { + "epoch": 1.3423891330403515, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3360 + }, + { + "epoch": 1.3427886536156612, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3361 + }, + { + "epoch": 1.3431881741909708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3362 + }, + { + "epoch": 1.3435876947662804, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3363 + }, + { + "epoch": 1.3439872153415902, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3364 + }, + { + "epoch": 1.3443867359168997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3365 + }, + { + "epoch": 1.3447862564922093, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3366 + }, + { + "epoch": 1.345185777067519, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3367 + }, + { + "epoch": 1.3455852976428286, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3368 + }, + { + "epoch": 1.3459848182181382, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3369 + }, + { + "epoch": 1.346384338793448, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3370 + }, + { + "epoch": 1.3467838593687576, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 3371 + }, + { + "epoch": 1.3471833799440671, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3372 + }, + { + "epoch": 1.3475829005193767, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3373 + }, + { + "epoch": 1.3479824210946862, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3374 + }, + { + "epoch": 1.348381941669996, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 3375 + }, + { + "epoch": 1.3487814622453056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3376 + }, + { + "epoch": 1.3491809828206152, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3377 + }, + { + "epoch": 1.349580503395925, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 3378 + }, + { + "epoch": 1.3499800239712345, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 3379 + }, + { + "epoch": 1.350379544546544, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3380 + }, + { + "epoch": 1.3507790651218539, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3381 + }, + { + "epoch": 1.3511785856971634, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3382 + }, + { + "epoch": 1.351578106272473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3383 + }, + { + "epoch": 1.3519776268477828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3384 + }, + { + "epoch": 1.3523771474230923, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 3385 + }, + { + "epoch": 1.352776667998402, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3386 + }, + { + "epoch": 1.3531761885737115, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3387 + }, + { + "epoch": 1.3535757091490213, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3388 + }, + { + "epoch": 1.3539752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3389 + }, + { + "epoch": 1.3543747502996404, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3390 + }, + { + "epoch": 1.35477427087495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3391 + }, + { + "epoch": 1.3551737914502597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 3392 + }, + { + "epoch": 1.3555733120255693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3393 + }, + { + "epoch": 1.3559728326008789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3394 + }, + { + "epoch": 1.3563723531761886, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3395 + }, + { + "epoch": 1.3567718737514982, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3396 + }, + { + "epoch": 1.3571713943268078, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3397 + }, + { + "epoch": 1.3575709149021176, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3398 + }, + { + "epoch": 1.3579704354774271, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 3399 + }, + { + "epoch": 1.3583699560527367, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3400 + }, + { + "epoch": 1.3587694766280465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3401 + }, + { + "epoch": 1.359168997203356, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3402 + }, + { + "epoch": 1.3595685177786656, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3403 + }, + { + "epoch": 1.3599680383539752, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3404 + }, + { + "epoch": 1.3603675589292847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3405 + }, + { + "epoch": 1.3607670795045945, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 3406 + }, + { + "epoch": 1.361166600079904, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3407 + }, + { + "epoch": 1.3615661206552137, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3408 + }, + { + "epoch": 1.3619656412305234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3409 + }, + { + "epoch": 1.362365161805833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3410 + }, + { + "epoch": 1.3627646823811426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3411 + }, + { + "epoch": 1.3631642029564524, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 3412 + }, + { + "epoch": 1.363563723531762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3413 + }, + { + "epoch": 1.3639632441070715, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3414 + }, + { + "epoch": 1.3643627646823813, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3415 + }, + { + "epoch": 1.3647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3416 + }, + { + "epoch": 1.3651618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3417 + }, + { + "epoch": 1.36556132640831, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3418 + }, + { + "epoch": 1.3659608469836195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3419 + }, + { + "epoch": 1.3663603675589293, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3420 + }, + { + "epoch": 1.3667598881342389, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3421 + }, + { + "epoch": 1.3671594087095484, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3422 + }, + { + "epoch": 1.3675589292848582, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3423 + }, + { + "epoch": 1.3679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3424 + }, + { + "epoch": 1.3683579704354774, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3425 + }, + { + "epoch": 1.3687574910107871, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3426 + }, + { + "epoch": 1.3691570115860967, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3427 + }, + { + "epoch": 1.3695565321614063, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3428 + }, + { + "epoch": 1.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3429 + }, + { + "epoch": 1.3703555733120256, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3430 + }, + { + "epoch": 1.3707550938873352, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3431 + }, + { + "epoch": 1.3711546144626447, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3432 + }, + { + "epoch": 1.3715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3433 + }, + { + "epoch": 1.371953655613264, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 3434 + }, + { + "epoch": 1.3723531761885737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3435 + }, + { + "epoch": 1.3727526967638832, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3436 + }, + { + "epoch": 1.373152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 3437 + }, + { + "epoch": 1.3735517379145026, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3438 + }, + { + "epoch": 1.3739512584898121, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3439 + }, + { + "epoch": 1.374350779065122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3440 + }, + { + "epoch": 1.3747502996404315, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3441 + }, + { + "epoch": 1.375149820215741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3442 + }, + { + "epoch": 1.3755493407910508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3443 + }, + { + "epoch": 1.3759488613663604, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3444 + }, + { + "epoch": 1.37634838194167, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3445 + }, + { + "epoch": 1.3767479025169798, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3446 + }, + { + "epoch": 1.3771474230922893, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3447 + }, + { + "epoch": 1.3775469436675989, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3448 + }, + { + "epoch": 1.3779464642429085, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3449 + }, + { + "epoch": 1.378345984818218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3450 + }, + { + "epoch": 1.3787455053935278, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3451 + }, + { + "epoch": 1.3791450259688374, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3452 + }, + { + "epoch": 1.379544546544147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3453 + }, + { + "epoch": 1.3799440671194567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3454 + }, + { + "epoch": 1.3803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3455 + }, + { + "epoch": 1.3807431082700758, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3456 + }, + { + "epoch": 1.3811426288453856, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3457 + }, + { + "epoch": 1.3815421494206952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3458 + }, + { + "epoch": 1.3819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3459 + }, + { + "epoch": 1.3823411905713145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3460 + }, + { + "epoch": 1.3827407111466241, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3461 + }, + { + "epoch": 1.3831402317219337, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3462 + }, + { + "epoch": 1.3835397522972432, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3463 + }, + { + "epoch": 1.3839392728725528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3464 + }, + { + "epoch": 1.3843387934478626, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3465 + }, + { + "epoch": 1.3847383140231722, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3466 + }, + { + "epoch": 1.3851378345984817, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3467 + }, + { + "epoch": 1.3855373551737915, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3468 + }, + { + "epoch": 1.385936875749101, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3469 + }, + { + "epoch": 1.3863363963244106, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3470 + }, + { + "epoch": 1.3867359168997204, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3471 + }, + { + "epoch": 1.38713543747503, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3472 + }, + { + "epoch": 1.3875349580503396, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3473 + }, + { + "epoch": 1.3879344786256493, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3474 + }, + { + "epoch": 1.388333999200959, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3475 + }, + { + "epoch": 1.3887335197762685, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3476 + }, + { + "epoch": 1.389133040351578, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3477 + }, + { + "epoch": 1.3895325609268878, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3478 + }, + { + "epoch": 1.3899320815021974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3479 + }, + { + "epoch": 1.390331602077507, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3480 + }, + { + "epoch": 1.3907311226528165, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3481 + }, + { + "epoch": 1.3911306432281263, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3482 + }, + { + "epoch": 1.3915301638034359, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3483 + }, + { + "epoch": 1.3919296843787454, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3484 + }, + { + "epoch": 1.3923292049540552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3485 + }, + { + "epoch": 1.3927287255293648, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3486 + }, + { + "epoch": 1.3931282461046743, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3487 + }, + { + "epoch": 1.3935277666799841, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3488 + }, + { + "epoch": 1.3939272872552937, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3489 + }, + { + "epoch": 1.3943268078306033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3490 + }, + { + "epoch": 1.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3491 + }, + { + "epoch": 1.3951258489812226, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3492 + }, + { + "epoch": 1.3955253695565322, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3493 + }, + { + "epoch": 1.3959248901318417, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3494 + }, + { + "epoch": 1.3963244107071513, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3495 + }, + { + "epoch": 1.396723931282461, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3496 + }, + { + "epoch": 1.3971234518577706, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3497 + }, + { + "epoch": 1.3975229724330802, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3498 + }, + { + "epoch": 1.39792249300839, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3499 + }, + { + "epoch": 1.3983220135836996, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3500 + }, + { + "epoch": 1.3987215341590091, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3501 + }, + { + "epoch": 1.399121054734319, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3502 + }, + { + "epoch": 1.3995205753096285, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3503 + }, + { + "epoch": 1.399920095884938, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3504 + }, + { + "epoch": 1.4003196164602478, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3505 + }, + { + "epoch": 1.4007191370355574, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3506 + }, + { + "epoch": 1.401118657610867, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3507 + }, + { + "epoch": 1.4015181781861765, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3508 + }, + { + "epoch": 1.401917698761486, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3509 + }, + { + "epoch": 1.4023172193367959, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3510 + }, + { + "epoch": 1.4027167399121054, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3511 + }, + { + "epoch": 1.403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3512 + }, + { + "epoch": 1.4035157810627248, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3513 + }, + { + "epoch": 1.4039153016380344, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3514 + }, + { + "epoch": 1.404314822213344, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3515 + }, + { + "epoch": 1.4047143427886537, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3516 + }, + { + "epoch": 1.4051138633639633, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3517 + }, + { + "epoch": 1.4055133839392728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3518 + }, + { + "epoch": 1.4059129045145826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3519 + }, + { + "epoch": 1.4063124250898922, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3520 + }, + { + "epoch": 1.4067119456652017, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3521 + }, + { + "epoch": 1.4071114662405113, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3522 + }, + { + "epoch": 1.407510986815821, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3523 + }, + { + "epoch": 1.4079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3524 + }, + { + "epoch": 1.4083100279664402, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3525 + }, + { + "epoch": 1.4087095485417498, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3526 + }, + { + "epoch": 1.4091090691170596, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3527 + }, + { + "epoch": 1.4095085896923691, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3528 + }, + { + "epoch": 1.4099081102676787, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3529 + }, + { + "epoch": 1.4103076308429885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 3530 + }, + { + "epoch": 1.410707151418298, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3531 + }, + { + "epoch": 1.4111066719936076, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 3532 + }, + { + "epoch": 1.4115061925689174, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3533 + }, + { + "epoch": 1.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3534 + }, + { + "epoch": 1.4123052337195365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3535 + }, + { + "epoch": 1.4127047542948463, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3536 + }, + { + "epoch": 1.4131042748701559, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3537 + }, + { + "epoch": 1.4135037954454654, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3538 + }, + { + "epoch": 1.413903316020775, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3539 + }, + { + "epoch": 1.4143028365960846, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3540 + }, + { + "epoch": 1.4147023571713944, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3541 + }, + { + "epoch": 1.415101877746704, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3542 + }, + { + "epoch": 1.4155013983220135, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3543 + }, + { + "epoch": 1.4159009188973233, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3544 + }, + { + "epoch": 1.4163004394726328, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3545 + }, + { + "epoch": 1.4166999600479424, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3546 + }, + { + "epoch": 1.4170994806232522, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3547 + }, + { + "epoch": 1.4174990011985618, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3548 + }, + { + "epoch": 1.4178985217738713, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3549 + }, + { + "epoch": 1.418298042349181, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 3550 + }, + { + "epoch": 1.4186975629244907, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3551 + }, + { + "epoch": 1.4190970834998002, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3552 + }, + { + "epoch": 1.4194966040751098, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3553 + }, + { + "epoch": 1.4198961246504194, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3554 + }, + { + "epoch": 1.4202956452257292, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3555 + }, + { + "epoch": 1.4206951658010387, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3556 + }, + { + "epoch": 1.4210946863763483, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3557 + }, + { + "epoch": 1.421494206951658, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3558 + }, + { + "epoch": 1.4218937275269676, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3559 + }, + { + "epoch": 1.4222932481022772, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3560 + }, + { + "epoch": 1.422692768677587, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3561 + }, + { + "epoch": 1.4230922892528965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3562 + }, + { + "epoch": 1.4234918098282061, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3563 + }, + { + "epoch": 1.423891330403516, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3564 + }, + { + "epoch": 1.4242908509788255, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3565 + }, + { + "epoch": 1.424690371554135, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3566 + }, + { + "epoch": 1.4250898921294446, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3567 + }, + { + "epoch": 1.4254894127047544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3568 + }, + { + "epoch": 1.425888933280064, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3569 + }, + { + "epoch": 1.4262884538553735, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3570 + }, + { + "epoch": 1.426687974430683, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3571 + }, + { + "epoch": 1.4270874950059929, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3572 + }, + { + "epoch": 1.4274870155813024, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3573 + }, + { + "epoch": 1.427886536156612, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3574 + }, + { + "epoch": 1.4282860567319218, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3575 + }, + { + "epoch": 1.4286855773072313, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3576 + }, + { + "epoch": 1.429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 3577 + }, + { + "epoch": 1.4294846184578507, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3578 + }, + { + "epoch": 1.4298841390331603, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3579 + }, + { + "epoch": 1.4302836596084698, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3580 + }, + { + "epoch": 1.4306831801837796, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3581 + }, + { + "epoch": 1.4310827007590892, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3582 + }, + { + "epoch": 1.4314822213343987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3583 + }, + { + "epoch": 1.4318817419097083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3584 + }, + { + "epoch": 1.4322812624850179, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3585 + }, + { + "epoch": 1.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3586 + }, + { + "epoch": 1.4330803036356372, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3587 + }, + { + "epoch": 1.4334798242109468, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3588 + }, + { + "epoch": 1.4338793447862566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3589 + }, + { + "epoch": 1.4342788653615661, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3590 + }, + { + "epoch": 1.4346783859368757, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3591 + }, + { + "epoch": 1.4350779065121855, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3592 + }, + { + "epoch": 1.435477427087495, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3593 + }, + { + "epoch": 1.4358769476628046, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3594 + }, + { + "epoch": 1.4362764682381144, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 3595 + }, + { + "epoch": 1.436675988813424, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3596 + }, + { + "epoch": 1.4370755093887335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3597 + }, + { + "epoch": 1.437475029964043, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3598 + }, + { + "epoch": 1.4378745505393526, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3599 + }, + { + "epoch": 1.4382740711146624, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3600 + }, + { + "epoch": 1.438673591689972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3601 + }, + { + "epoch": 1.4390731122652816, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3602 + }, + { + "epoch": 1.4394726328405913, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3603 + }, + { + "epoch": 1.439872153415901, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3604 + }, + { + "epoch": 1.4402716739912105, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3605 + }, + { + "epoch": 1.4406711945665203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3606 + }, + { + "epoch": 1.4410707151418298, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3607 + }, + { + "epoch": 1.4414702357171394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3608 + }, + { + "epoch": 1.4418697562924492, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3609 + }, + { + "epoch": 1.4422692768677587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3610 + }, + { + "epoch": 1.4426687974430683, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3611 + }, + { + "epoch": 1.4430683180183779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3612 + }, + { + "epoch": 1.4434678385936877, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3613 + }, + { + "epoch": 1.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3614 + }, + { + "epoch": 1.4442668797443068, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3615 + }, + { + "epoch": 1.4446664003196164, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 3616 + }, + { + "epoch": 1.4450659208949261, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3617 + }, + { + "epoch": 1.4454654414702357, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3618 + }, + { + "epoch": 1.4458649620455453, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3619 + }, + { + "epoch": 1.446264482620855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3620 + }, + { + "epoch": 1.4466640031961646, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3621 + }, + { + "epoch": 1.4470635237714742, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3622 + }, + { + "epoch": 1.447463044346784, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3623 + }, + { + "epoch": 1.4478625649220935, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3624 + }, + { + "epoch": 1.448262085497403, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3625 + }, + { + "epoch": 1.4486616060727129, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3626 + }, + { + "epoch": 1.4490611266480224, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 3627 + }, + { + "epoch": 1.449460647223332, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3628 + }, + { + "epoch": 1.4498601677986416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3629 + }, + { + "epoch": 1.4502596883739511, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3630 + }, + { + "epoch": 1.450659208949261, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3631 + }, + { + "epoch": 1.4510587295245705, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3632 + }, + { + "epoch": 1.45145825009988, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3633 + }, + { + "epoch": 1.4518577706751898, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3634 + }, + { + "epoch": 1.4522572912504994, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3635 + }, + { + "epoch": 1.452656811825809, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3636 + }, + { + "epoch": 1.4530563324011188, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3637 + }, + { + "epoch": 1.4534558529764283, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3638 + }, + { + "epoch": 1.4538553735517379, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3639 + }, + { + "epoch": 1.4542548941270477, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3640 + }, + { + "epoch": 1.4546544147023572, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3641 + }, + { + "epoch": 1.4550539352776668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3642 + }, + { + "epoch": 1.4554534558529764, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3643 + }, + { + "epoch": 1.455852976428286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3644 + }, + { + "epoch": 1.4562524970035957, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3645 + }, + { + "epoch": 1.4566520175789053, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 3646 + }, + { + "epoch": 1.4570515381542148, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3647 + }, + { + "epoch": 1.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3648 + }, + { + "epoch": 1.4578505793048342, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3649 + }, + { + "epoch": 1.4582500998801438, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3650 + }, + { + "epoch": 1.4586496204554535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3651 + }, + { + "epoch": 1.459049141030763, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3652 + }, + { + "epoch": 1.4594486616060727, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3653 + }, + { + "epoch": 1.4598481821813825, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3654 + }, + { + "epoch": 1.460247702756692, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 3655 + }, + { + "epoch": 1.4606472233320016, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3656 + }, + { + "epoch": 1.4610467439073112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3657 + }, + { + "epoch": 1.461446264482621, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3658 + }, + { + "epoch": 1.4618457850579305, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3659 + }, + { + "epoch": 1.46224530563324, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3660 + }, + { + "epoch": 1.4626448262085496, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3661 + }, + { + "epoch": 1.4630443467838594, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3662 + }, + { + "epoch": 1.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3663 + }, + { + "epoch": 1.4638433879344785, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3664 + }, + { + "epoch": 1.4642429085097883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3665 + }, + { + "epoch": 1.464642429085098, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3666 + }, + { + "epoch": 1.4650419496604075, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3667 + }, + { + "epoch": 1.4654414702357172, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 3668 + }, + { + "epoch": 1.4658409908110268, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3669 + }, + { + "epoch": 1.4662405113863364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3670 + }, + { + "epoch": 1.4666400319616462, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3671 + }, + { + "epoch": 1.4670395525369557, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3672 + }, + { + "epoch": 1.4674390731122653, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3673 + }, + { + "epoch": 1.4678385936875749, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3674 + }, + { + "epoch": 1.4682381142628844, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3675 + }, + { + "epoch": 1.4686376348381942, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 3676 + }, + { + "epoch": 1.4690371554135038, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 3677 + }, + { + "epoch": 1.4694366759888133, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3678 + }, + { + "epoch": 1.4698361965641231, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3679 + }, + { + "epoch": 1.4702357171394327, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3680 + }, + { + "epoch": 1.4706352377147422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 3681 + }, + { + "epoch": 1.471034758290052, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3682 + }, + { + "epoch": 1.4714342788653616, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3683 + }, + { + "epoch": 1.4718337994406712, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3684 + }, + { + "epoch": 1.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3685 + }, + { + "epoch": 1.4726328405912905, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3686 + }, + { + "epoch": 1.4730323611666, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3687 + }, + { + "epoch": 1.4734318817419096, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3688 + }, + { + "epoch": 1.4738314023172192, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3689 + }, + { + "epoch": 1.474230922892529, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3690 + }, + { + "epoch": 1.4746304434678386, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3691 + }, + { + "epoch": 1.4750299640431481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3692 + }, + { + "epoch": 1.475429484618458, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3693 + }, + { + "epoch": 1.4758290051937675, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3694 + }, + { + "epoch": 1.476228525769077, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3695 + }, + { + "epoch": 1.4766280463443868, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3696 + }, + { + "epoch": 1.4770275669196964, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3697 + }, + { + "epoch": 1.477427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3698 + }, + { + "epoch": 1.4778266080703157, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3699 + }, + { + "epoch": 1.4782261286456253, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3700 + }, + { + "epoch": 1.4786256492209349, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3701 + }, + { + "epoch": 1.4790251697962444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3702 + }, + { + "epoch": 1.479424690371554, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3703 + }, + { + "epoch": 1.4798242109468638, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3704 + }, + { + "epoch": 1.4802237315221733, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3705 + }, + { + "epoch": 1.480623252097483, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3706 + }, + { + "epoch": 1.4810227726727927, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3707 + }, + { + "epoch": 1.4814222932481023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3708 + }, + { + "epoch": 1.4818218138234118, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 3709 + }, + { + "epoch": 1.4822213343987216, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 3710 + }, + { + "epoch": 1.4826208549740312, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3711 + }, + { + "epoch": 1.4830203755493407, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 3712 + }, + { + "epoch": 1.4834198961246505, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3713 + }, + { + "epoch": 1.48381941669996, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3714 + }, + { + "epoch": 1.4842189372752697, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3715 + }, + { + "epoch": 1.4846184578505794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3716 + }, + { + "epoch": 1.485017978425889, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3717 + }, + { + "epoch": 1.4854174990011986, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3718 + }, + { + "epoch": 1.4858170195765081, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3719 + }, + { + "epoch": 1.4862165401518177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3720 + }, + { + "epoch": 1.4866160607271275, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3721 + }, + { + "epoch": 1.487015581302437, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3722 + }, + { + "epoch": 1.4874151018777466, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3723 + }, + { + "epoch": 1.4878146224530564, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3724 + }, + { + "epoch": 1.488214143028366, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3725 + }, + { + "epoch": 1.4886136636036755, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3726 + }, + { + "epoch": 1.4890131841789853, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3727 + }, + { + "epoch": 1.4894127047542949, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3728 + }, + { + "epoch": 1.4898122253296044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3729 + }, + { + "epoch": 1.4902117459049142, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3730 + }, + { + "epoch": 1.4906112664802238, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3731 + }, + { + "epoch": 1.4910107870555334, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3732 + }, + { + "epoch": 1.491410307630843, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3733 + }, + { + "epoch": 1.4918098282061525, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3734 + }, + { + "epoch": 1.4922093487814623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3735 + }, + { + "epoch": 1.4926088693567718, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3736 + }, + { + "epoch": 1.4930083899320814, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3737 + }, + { + "epoch": 1.4934079105073912, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3738 + }, + { + "epoch": 1.4938074310827008, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3739 + }, + { + "epoch": 1.4942069516580103, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3740 + }, + { + "epoch": 1.49460647223332, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3741 + }, + { + "epoch": 1.4950059928086297, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3742 + }, + { + "epoch": 1.4954055133839392, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3743 + }, + { + "epoch": 1.495805033959249, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3744 + }, + { + "epoch": 1.4962045545345586, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3745 + }, + { + "epoch": 1.4966040751098681, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3746 + }, + { + "epoch": 1.4970035956851777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3747 + }, + { + "epoch": 1.4974031162604873, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3748 + }, + { + "epoch": 1.497802636835797, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3749 + }, + { + "epoch": 1.4982021574111066, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3750 + }, + { + "epoch": 1.4986016779864162, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3751 + }, + { + "epoch": 1.499001198561726, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3752 + }, + { + "epoch": 1.4994007191370355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3753 + }, + { + "epoch": 1.499800239712345, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3754 + }, + { + "epoch": 1.500199760287655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3755 + }, + { + "epoch": 1.5005992808629645, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3756 + }, + { + "epoch": 1.500998801438274, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3757 + }, + { + "epoch": 1.5013983220135838, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3758 + }, + { + "epoch": 1.5017978425888934, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3759 + }, + { + "epoch": 1.502197363164203, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 3760 + }, + { + "epoch": 1.5025968837395127, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3761 + }, + { + "epoch": 1.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3762 + }, + { + "epoch": 1.5033959248901319, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3763 + }, + { + "epoch": 1.5037954454654414, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3764 + }, + { + "epoch": 1.504194966040751, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3765 + }, + { + "epoch": 1.5045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3766 + }, + { + "epoch": 1.5049940071913703, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3767 + }, + { + "epoch": 1.50539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3768 + }, + { + "epoch": 1.5057930483419897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3769 + }, + { + "epoch": 1.5061925689172992, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3770 + }, + { + "epoch": 1.5065920894926088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3771 + }, + { + "epoch": 1.5069916100679186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3772 + }, + { + "epoch": 1.5073911306432282, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3773 + }, + { + "epoch": 1.5077906512185377, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3774 + }, + { + "epoch": 1.5081901717938475, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3775 + }, + { + "epoch": 1.5085896923691569, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3776 + }, + { + "epoch": 1.5089892129444666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3777 + }, + { + "epoch": 1.5093887335197764, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3778 + }, + { + "epoch": 1.5097882540950858, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3779 + }, + { + "epoch": 1.5101877746703956, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3780 + }, + { + "epoch": 1.5105872952457051, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3781 + }, + { + "epoch": 1.5109868158210147, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3782 + }, + { + "epoch": 1.5113863363963245, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3783 + }, + { + "epoch": 1.511785856971634, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3784 + }, + { + "epoch": 1.5121853775469436, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3785 + }, + { + "epoch": 1.5125848981222534, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3786 + }, + { + "epoch": 1.512984418697563, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3787 + }, + { + "epoch": 1.5133839392728725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3788 + }, + { + "epoch": 1.5137834598481823, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3789 + }, + { + "epoch": 1.5141829804234919, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3790 + }, + { + "epoch": 1.5145825009988014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3791 + }, + { + "epoch": 1.5149820215741112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3792 + }, + { + "epoch": 1.5153815421494206, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3793 + }, + { + "epoch": 1.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3794 + }, + { + "epoch": 1.51618058330004, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3795 + }, + { + "epoch": 1.5165801038753495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3796 + }, + { + "epoch": 1.5169796244506593, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3797 + }, + { + "epoch": 1.5173791450259688, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3798 + }, + { + "epoch": 1.5177786656012784, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3799 + }, + { + "epoch": 1.5181781861765882, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3800 + }, + { + "epoch": 1.5185777067518977, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3801 + }, + { + "epoch": 1.5189772273272073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3802 + }, + { + "epoch": 1.519376747902517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3803 + }, + { + "epoch": 1.5197762684778267, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3804 + }, + { + "epoch": 1.5201757890531362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3805 + }, + { + "epoch": 1.520575309628446, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 3806 + }, + { + "epoch": 1.5209748302037553, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3807 + }, + { + "epoch": 1.5213743507790651, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3808 + }, + { + "epoch": 1.5217738713543747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3809 + }, + { + "epoch": 1.5221733919296843, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3810 + }, + { + "epoch": 1.522572912504994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3811 + }, + { + "epoch": 1.5229724330803036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3812 + }, + { + "epoch": 1.5233719536556132, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3813 + }, + { + "epoch": 1.523771474230923, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 3814 + }, + { + "epoch": 1.5241709948062325, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3815 + }, + { + "epoch": 1.524570515381542, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3816 + }, + { + "epoch": 1.5249700359568519, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3817 + }, + { + "epoch": 1.5253695565321614, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3818 + }, + { + "epoch": 1.525769077107471, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3819 + }, + { + "epoch": 1.5261685976827808, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3820 + }, + { + "epoch": 1.5265681182580901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 3821 + }, + { + "epoch": 1.5269676388334, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3822 + }, + { + "epoch": 1.5273671594087097, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3823 + }, + { + "epoch": 1.527766679984019, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 3824 + }, + { + "epoch": 1.5281662005593288, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 3825 + }, + { + "epoch": 1.5285657211346384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3826 + }, + { + "epoch": 1.528965241709948, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3827 + }, + { + "epoch": 1.5293647622852578, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3828 + }, + { + "epoch": 1.5297642828605673, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3829 + }, + { + "epoch": 1.5301638034358769, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3830 + }, + { + "epoch": 1.5305633240111867, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3831 + }, + { + "epoch": 1.5309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 3832 + }, + { + "epoch": 1.5313623651618058, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3833 + }, + { + "epoch": 1.5317618857371156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 3834 + }, + { + "epoch": 1.5321614063124251, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3835 + }, + { + "epoch": 1.5325609268877347, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3836 + }, + { + "epoch": 1.5329604474630445, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3837 + }, + { + "epoch": 1.5333599680383538, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3838 + }, + { + "epoch": 1.5337594886136636, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3839 + }, + { + "epoch": 1.5341590091889732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 3840 + }, + { + "epoch": 1.5345585297642828, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3841 + }, + { + "epoch": 1.5349580503395925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3842 + }, + { + "epoch": 1.535357570914902, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3843 + }, + { + "epoch": 1.5357570914902117, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3844 + }, + { + "epoch": 1.5361566120655215, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 3845 + }, + { + "epoch": 1.536556132640831, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3846 + }, + { + "epoch": 1.5369556532161406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3847 + }, + { + "epoch": 1.5373551737914504, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3848 + }, + { + "epoch": 1.53775469436676, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3849 + }, + { + "epoch": 1.5381542149420695, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3850 + }, + { + "epoch": 1.5385537355173793, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3851 + }, + { + "epoch": 1.5389532560926886, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3852 + }, + { + "epoch": 1.5393527766679984, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3853 + }, + { + "epoch": 1.539752297243308, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3854 + }, + { + "epoch": 1.5401518178186175, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 3855 + }, + { + "epoch": 1.5405513383939273, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3856 + }, + { + "epoch": 1.540950858969237, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3857 + }, + { + "epoch": 1.5413503795445465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3858 + }, + { + "epoch": 1.5417499001198562, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3859 + }, + { + "epoch": 1.5421494206951658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3860 + }, + { + "epoch": 1.5425489412704754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3861 + }, + { + "epoch": 1.5429484618457852, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3862 + }, + { + "epoch": 1.5433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3863 + }, + { + "epoch": 1.5437475029964043, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3864 + }, + { + "epoch": 1.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3865 + }, + { + "epoch": 1.5445465441470234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3866 + }, + { + "epoch": 1.5449460647223332, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3867 + }, + { + "epoch": 1.545345585297643, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3868 + }, + { + "epoch": 1.5457451058729523, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3869 + }, + { + "epoch": 1.5461446264482621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 3870 + }, + { + "epoch": 1.5465441470235717, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3871 + }, + { + "epoch": 1.5469436675988812, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3872 + }, + { + "epoch": 1.547343188174191, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3873 + }, + { + "epoch": 1.5477427087495006, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3874 + }, + { + "epoch": 1.5481422293248102, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3875 + }, + { + "epoch": 1.54854174990012, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 3876 + }, + { + "epoch": 1.5489412704754295, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3877 + }, + { + "epoch": 1.549340791050739, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3878 + }, + { + "epoch": 1.5497403116260489, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3879 + }, + { + "epoch": 1.5501398322013584, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3880 + }, + { + "epoch": 1.550539352776668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3881 + }, + { + "epoch": 1.5509388733519778, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3882 + }, + { + "epoch": 1.5513383939272871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3883 + }, + { + "epoch": 1.551737914502597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 3884 + }, + { + "epoch": 1.5521374350779065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3885 + }, + { + "epoch": 1.552536955653216, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 3886 + }, + { + "epoch": 1.5529364762285258, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3887 + }, + { + "epoch": 1.5533359968038354, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3888 + }, + { + "epoch": 1.553735517379145, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3889 + }, + { + "epoch": 1.5541350379544547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3890 + }, + { + "epoch": 1.5545345585297643, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3891 + }, + { + "epoch": 1.5549340791050739, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3892 + }, + { + "epoch": 1.5553335996803836, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3893 + }, + { + "epoch": 1.5557331202556932, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3894 + }, + { + "epoch": 1.5561326408310028, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3895 + }, + { + "epoch": 1.5565321614063126, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3896 + }, + { + "epoch": 1.556931681981622, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3897 + }, + { + "epoch": 1.5573312025569317, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3898 + }, + { + "epoch": 1.5577307231322413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3899 + }, + { + "epoch": 1.5581302437075508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3900 + }, + { + "epoch": 1.5585297642828606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3901 + }, + { + "epoch": 1.5589292848581702, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3902 + }, + { + "epoch": 1.5593288054334797, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3903 + }, + { + "epoch": 1.5597283260087895, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3904 + }, + { + "epoch": 1.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3905 + }, + { + "epoch": 1.5605273671594087, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3906 + }, + { + "epoch": 1.5609268877347184, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3907 + }, + { + "epoch": 1.561326408310028, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3908 + }, + { + "epoch": 1.5617259288853376, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3909 + }, + { + "epoch": 1.5621254494606474, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3910 + }, + { + "epoch": 1.5625249700359567, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3911 + }, + { + "epoch": 1.5629244906112665, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3912 + }, + { + "epoch": 1.5633240111865763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3913 + }, + { + "epoch": 1.5637235317618856, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3914 + }, + { + "epoch": 1.5641230523371954, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3915 + }, + { + "epoch": 1.564522572912505, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3916 + }, + { + "epoch": 1.5649220934878145, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3917 + }, + { + "epoch": 1.5653216140631243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3918 + }, + { + "epoch": 1.5657211346384339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 3919 + }, + { + "epoch": 1.5661206552137434, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3920 + }, + { + "epoch": 1.5665201757890532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3921 + }, + { + "epoch": 1.5669196963643628, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 3922 + }, + { + "epoch": 1.5673192169396724, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3923 + }, + { + "epoch": 1.5677187375149821, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3924 + }, + { + "epoch": 1.5681182580902915, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3925 + }, + { + "epoch": 1.5685177786656013, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3926 + }, + { + "epoch": 1.568917299240911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3927 + }, + { + "epoch": 1.5693168198162204, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3928 + }, + { + "epoch": 1.5697163403915302, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3929 + }, + { + "epoch": 1.5701158609668397, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3930 + }, + { + "epoch": 1.5705153815421493, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3931 + }, + { + "epoch": 1.570914902117459, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3932 + }, + { + "epoch": 1.5713144226927687, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3933 + }, + { + "epoch": 1.5717139432680782, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3934 + }, + { + "epoch": 1.572113463843388, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3935 + }, + { + "epoch": 1.5725129844186976, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3936 + }, + { + "epoch": 1.5729125049940071, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3937 + }, + { + "epoch": 1.573312025569317, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 3938 + }, + { + "epoch": 1.5737115461446265, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3939 + }, + { + "epoch": 1.574111066719936, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3940 + }, + { + "epoch": 1.5745105872952458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3941 + }, + { + "epoch": 1.5749101078705552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3942 + }, + { + "epoch": 1.575309628445865, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3943 + }, + { + "epoch": 1.5757091490211745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3944 + }, + { + "epoch": 1.576108669596484, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3945 + }, + { + "epoch": 1.5765081901717939, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3946 + }, + { + "epoch": 1.5769077107471035, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3947 + }, + { + "epoch": 1.577307231322413, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3948 + }, + { + "epoch": 1.5777067518977228, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3949 + }, + { + "epoch": 1.5781062724730324, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3950 + }, + { + "epoch": 1.578505793048342, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3951 + }, + { + "epoch": 1.5789053136236517, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3952 + }, + { + "epoch": 1.5793048341989613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3953 + }, + { + "epoch": 1.5797043547742708, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3954 + }, + { + "epoch": 1.5801038753495806, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3955 + }, + { + "epoch": 1.58050339592489, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3956 + }, + { + "epoch": 1.5809029165001998, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 3957 + }, + { + "epoch": 1.5813024370755095, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3958 + }, + { + "epoch": 1.581701957650819, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 3959 + }, + { + "epoch": 1.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3960 + }, + { + "epoch": 1.5825009988014382, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3961 + }, + { + "epoch": 1.5829005193767478, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3962 + }, + { + "epoch": 1.5833000399520576, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3963 + }, + { + "epoch": 1.5836995605273672, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 3964 + }, + { + "epoch": 1.5840990811026767, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 3965 + }, + { + "epoch": 1.5844986016779865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3966 + }, + { + "epoch": 1.584898122253296, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 3967 + }, + { + "epoch": 1.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3968 + }, + { + "epoch": 1.5856971634039154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3969 + }, + { + "epoch": 1.5860966839792248, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3970 + }, + { + "epoch": 1.5864962045545346, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3971 + }, + { + "epoch": 1.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3972 + }, + { + "epoch": 1.5872952457051537, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3973 + }, + { + "epoch": 1.5876947662804635, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3974 + }, + { + "epoch": 1.588094286855773, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3975 + }, + { + "epoch": 1.5884938074310826, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3976 + }, + { + "epoch": 1.5888933280063924, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3977 + }, + { + "epoch": 1.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3978 + }, + { + "epoch": 1.5896923691570115, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 3979 + }, + { + "epoch": 1.5900918897323213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3980 + }, + { + "epoch": 1.5904914103076309, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3981 + }, + { + "epoch": 1.5908909308829404, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3982 + }, + { + "epoch": 1.5912904514582502, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3983 + }, + { + "epoch": 1.5916899720335598, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3984 + }, + { + "epoch": 1.5920894926088693, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3985 + }, + { + "epoch": 1.5924890131841791, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3986 + }, + { + "epoch": 1.5928885337594885, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 3987 + }, + { + "epoch": 1.5932880543347983, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3988 + }, + { + "epoch": 1.5936875749101078, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3989 + }, + { + "epoch": 1.5940870954854174, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3990 + }, + { + "epoch": 1.5944866160607272, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3991 + }, + { + "epoch": 1.5948861366360367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3992 + }, + { + "epoch": 1.5952856572113463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3993 + }, + { + "epoch": 1.595685177786656, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3994 + }, + { + "epoch": 1.5960846983619656, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3995 + }, + { + "epoch": 1.5964842189372752, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3996 + }, + { + "epoch": 1.596883739512585, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3997 + }, + { + "epoch": 1.5972832600878946, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3998 + }, + { + "epoch": 1.5976827806632041, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3999 + }, + { + "epoch": 1.598082301238514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4000 + }, + { + "epoch": 1.5984818218138233, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4001 + }, + { + "epoch": 1.598881342389133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4002 + }, + { + "epoch": 1.5992808629644428, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4003 + }, + { + "epoch": 1.5996803835397522, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4004 + }, + { + "epoch": 1.600079904115062, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4005 + }, + { + "epoch": 1.6004794246903715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4006 + }, + { + "epoch": 1.600878945265681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4007 + }, + { + "epoch": 1.6012784658409909, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4008 + }, + { + "epoch": 1.6016779864163004, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4009 + }, + { + "epoch": 1.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4010 + }, + { + "epoch": 1.6024770275669198, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4011 + }, + { + "epoch": 1.6028765481422294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4012 + }, + { + "epoch": 1.603276068717539, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4013 + }, + { + "epoch": 1.6036755892928487, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4014 + }, + { + "epoch": 1.604075109868158, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4015 + }, + { + "epoch": 1.6044746304434678, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 4016 + }, + { + "epoch": 1.6048741510187776, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4017 + }, + { + "epoch": 1.605273671594087, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4018 + }, + { + "epoch": 1.6056731921693967, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4019 + }, + { + "epoch": 1.6060727127447063, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4020 + }, + { + "epoch": 1.6064722333200159, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4021 + }, + { + "epoch": 1.6068717538953257, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4022 + }, + { + "epoch": 1.6072712744706352, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4023 + }, + { + "epoch": 1.6076707950459448, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4024 + }, + { + "epoch": 1.6080703156212546, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4025 + }, + { + "epoch": 1.6084698361965641, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4026 + }, + { + "epoch": 1.6088693567718737, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4027 + }, + { + "epoch": 1.6092688773471835, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4028 + }, + { + "epoch": 1.609668397922493, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4029 + }, + { + "epoch": 1.6100679184978026, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4030 + }, + { + "epoch": 1.6104674390731124, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 4031 + }, + { + "epoch": 1.6108669596484217, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4032 + }, + { + "epoch": 1.6112664802237315, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4033 + }, + { + "epoch": 1.611666000799041, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 4034 + }, + { + "epoch": 1.6120655213743507, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4035 + }, + { + "epoch": 1.6124650419496604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4036 + }, + { + "epoch": 1.61286456252497, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4037 + }, + { + "epoch": 1.6132640831002796, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 4038 + }, + { + "epoch": 1.6136636036755894, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4039 + }, + { + "epoch": 1.614063124250899, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4040 + }, + { + "epoch": 1.6144626448262085, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4041 + }, + { + "epoch": 1.6148621654015183, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4042 + }, + { + "epoch": 1.6152616859768278, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4043 + }, + { + "epoch": 1.6156612065521374, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4044 + }, + { + "epoch": 1.6160607271274472, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4045 + }, + { + "epoch": 1.6164602477027565, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4046 + }, + { + "epoch": 1.6168597682780663, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4047 + }, + { + "epoch": 1.617259288853376, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4048 + }, + { + "epoch": 1.6176588094286855, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4049 + }, + { + "epoch": 1.6180583300039952, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4050 + }, + { + "epoch": 1.6184578505793048, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4051 + }, + { + "epoch": 1.6188573711546144, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4052 + }, + { + "epoch": 1.6192568917299242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4053 + }, + { + "epoch": 1.6196564123052337, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4054 + }, + { + "epoch": 1.6200559328805433, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4055 + }, + { + "epoch": 1.620455453455853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4056 + }, + { + "epoch": 1.6208549740311626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4057 + }, + { + "epoch": 1.6212544946064722, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4058 + }, + { + "epoch": 1.621654015181782, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4059 + }, + { + "epoch": 1.6220535357570913, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4060 + }, + { + "epoch": 1.6224530563324011, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4061 + }, + { + "epoch": 1.622852576907711, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4062 + }, + { + "epoch": 1.6232520974830202, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4063 + }, + { + "epoch": 1.62365161805833, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4064 + }, + { + "epoch": 1.6240511386336396, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4065 + }, + { + "epoch": 1.6244506592089492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4066 + }, + { + "epoch": 1.624850179784259, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4067 + }, + { + "epoch": 1.6252497003595685, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4068 + }, + { + "epoch": 1.625649220934878, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4069 + }, + { + "epoch": 1.6260487415101879, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4070 + }, + { + "epoch": 1.6264482620854974, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4071 + }, + { + "epoch": 1.626847782660807, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4072 + }, + { + "epoch": 1.6272473032361168, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4073 + }, + { + "epoch": 1.6276468238114263, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4074 + }, + { + "epoch": 1.628046344386736, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 4075 + }, + { + "epoch": 1.6284458649620457, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4076 + }, + { + "epoch": 1.628845385537355, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4077 + }, + { + "epoch": 1.6292449061126648, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4078 + }, + { + "epoch": 1.6296444266879744, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 4079 + }, + { + "epoch": 1.630043947263284, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4080 + }, + { + "epoch": 1.6304434678385937, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4081 + }, + { + "epoch": 1.6308429884139033, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4082 + }, + { + "epoch": 1.6312425089892129, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4083 + }, + { + "epoch": 1.6316420295645226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4084 + }, + { + "epoch": 1.6320415501398322, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4085 + }, + { + "epoch": 1.6324410707151418, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4086 + }, + { + "epoch": 1.6328405912904516, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4087 + }, + { + "epoch": 1.6332401118657611, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4088 + }, + { + "epoch": 1.6336396324410707, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4089 + }, + { + "epoch": 1.6340391530163805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4090 + }, + { + "epoch": 1.6344386735916898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4091 + }, + { + "epoch": 1.6348381941669996, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4092 + }, + { + "epoch": 1.6352377147423094, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4093 + }, + { + "epoch": 1.6356372353176187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4094 + }, + { + "epoch": 1.6360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4095 + }, + { + "epoch": 1.636436276468238, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4096 + }, + { + "epoch": 1.6368357970435476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 4097 + }, + { + "epoch": 1.6372353176188574, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4098 + }, + { + "epoch": 1.637634838194167, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4099 + }, + { + "epoch": 1.6380343587694766, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 4100 + }, + { + "epoch": 1.6384338793447863, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4101 + }, + { + "epoch": 1.638833399920096, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4102 + }, + { + "epoch": 1.6392329204954055, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4103 + }, + { + "epoch": 1.6396324410707153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4104 + }, + { + "epoch": 1.6400319616460246, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4105 + }, + { + "epoch": 1.6404314822213344, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4106 + }, + { + "epoch": 1.6408310027966442, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4107 + }, + { + "epoch": 1.6412305233719535, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4108 + }, + { + "epoch": 1.6416300439472633, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4109 + }, + { + "epoch": 1.6420295645225729, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4110 + }, + { + "epoch": 1.6424290850978824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 4111 + }, + { + "epoch": 1.6428286056731922, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4112 + }, + { + "epoch": 1.6432281262485018, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4113 + }, + { + "epoch": 1.6436276468238114, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4114 + }, + { + "epoch": 1.6440271673991211, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4115 + }, + { + "epoch": 1.6444266879744307, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4116 + }, + { + "epoch": 1.6448262085497403, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4117 + }, + { + "epoch": 1.64522572912505, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4118 + }, + { + "epoch": 1.6456252497003596, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4119 + }, + { + "epoch": 1.6460247702756692, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4120 + }, + { + "epoch": 1.646424290850979, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4121 + }, + { + "epoch": 1.6468238114262883, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4122 + }, + { + "epoch": 1.647223332001598, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4123 + }, + { + "epoch": 1.6476228525769077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4124 + }, + { + "epoch": 1.6480223731522172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4125 + }, + { + "epoch": 1.648421893727527, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4126 + }, + { + "epoch": 1.6488214143028366, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4127 + }, + { + "epoch": 1.6492209348781461, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4128 + }, + { + "epoch": 1.649620455453456, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4129 + }, + { + "epoch": 1.6500199760287655, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4130 + }, + { + "epoch": 1.650419496604075, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 4131 + }, + { + "epoch": 1.6508190171793848, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4132 + }, + { + "epoch": 1.6512185377546944, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4133 + }, + { + "epoch": 1.651618058330004, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4134 + }, + { + "epoch": 1.6520175789053138, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4135 + }, + { + "epoch": 1.652417099480623, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 4136 + }, + { + "epoch": 1.6528166200559329, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4137 + }, + { + "epoch": 1.6532161406312427, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4138 + }, + { + "epoch": 1.653615661206552, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4139 + }, + { + "epoch": 1.6540151817818618, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4140 + }, + { + "epoch": 1.6544147023571714, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4141 + }, + { + "epoch": 1.654814222932481, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4142 + }, + { + "epoch": 1.6552137435077907, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4143 + }, + { + "epoch": 1.6556132640831003, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4144 + }, + { + "epoch": 1.6560127846584098, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4145 + }, + { + "epoch": 1.6564123052337196, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4146 + }, + { + "epoch": 1.6568118258090292, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4147 + }, + { + "epoch": 1.6572113463843388, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4148 + }, + { + "epoch": 1.6576108669596485, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4149 + }, + { + "epoch": 1.6580103875349579, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4150 + }, + { + "epoch": 1.6584099081102677, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4151 + }, + { + "epoch": 1.6588094286855775, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4152 + }, + { + "epoch": 1.6592089492608868, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4153 + }, + { + "epoch": 1.6596084698361966, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4154 + }, + { + "epoch": 1.6600079904115062, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4155 + }, + { + "epoch": 1.6604075109868157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4156 + }, + { + "epoch": 1.6608070315621255, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4157 + }, + { + "epoch": 1.661206552137435, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4158 + }, + { + "epoch": 1.6616060727127446, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4159 + }, + { + "epoch": 1.6620055932880544, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4160 + }, + { + "epoch": 1.662405113863364, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4161 + }, + { + "epoch": 1.6628046344386735, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4162 + }, + { + "epoch": 1.6632041550139833, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4163 + }, + { + "epoch": 1.663603675589293, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4164 + }, + { + "epoch": 1.6640031961646025, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4165 + }, + { + "epoch": 1.6644027167399122, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4166 + }, + { + "epoch": 1.6648022373152216, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4167 + }, + { + "epoch": 1.6652017578905314, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4168 + }, + { + "epoch": 1.665601278465841, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 4169 + }, + { + "epoch": 1.6660007990411505, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4170 + }, + { + "epoch": 1.6664003196164603, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4171 + }, + { + "epoch": 1.6667998401917699, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 4172 + }, + { + "epoch": 1.6671993607670794, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4173 + }, + { + "epoch": 1.6675988813423892, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4174 + }, + { + "epoch": 1.6679984019176988, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4175 + }, + { + "epoch": 1.6683979224930083, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 4176 + }, + { + "epoch": 1.6687974430683181, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4177 + }, + { + "epoch": 1.6691969636436277, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4178 + }, + { + "epoch": 1.6695964842189372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4179 + }, + { + "epoch": 1.669996004794247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4180 + }, + { + "epoch": 1.6703955253695564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4181 + }, + { + "epoch": 1.6707950459448662, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4182 + }, + { + "epoch": 1.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4183 + }, + { + "epoch": 1.6715940870954853, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4184 + }, + { + "epoch": 1.671993607670795, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4185 + }, + { + "epoch": 1.6723931282461046, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 4186 + }, + { + "epoch": 1.6727926488214142, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4187 + }, + { + "epoch": 1.673192169396724, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4188 + }, + { + "epoch": 1.6735916899720336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4189 + }, + { + "epoch": 1.6739912105473431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4190 + }, + { + "epoch": 1.674390731122653, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4191 + }, + { + "epoch": 1.6747902516979625, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4192 + }, + { + "epoch": 1.675189772273272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4193 + }, + { + "epoch": 1.6755892928485818, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4194 + }, + { + "epoch": 1.6759888134238912, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4195 + }, + { + "epoch": 1.676388333999201, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4196 + }, + { + "epoch": 1.6767878545745107, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4197 + }, + { + "epoch": 1.67718737514982, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4198 + }, + { + "epoch": 1.6775868957251299, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4199 + }, + { + "epoch": 1.6779864163004394, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4200 + }, + { + "epoch": 1.678385936875749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 4201 + }, + { + "epoch": 1.6787854574510588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4202 + }, + { + "epoch": 1.6791849780263683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4203 + }, + { + "epoch": 1.679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 4204 + }, + { + "epoch": 1.6799840191769877, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4205 + }, + { + "epoch": 1.6803835397522973, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4206 + }, + { + "epoch": 1.6807830603276068, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4207 + }, + { + "epoch": 1.6811825809029166, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4208 + }, + { + "epoch": 1.6815821014782262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4209 + }, + { + "epoch": 1.6819816220535357, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4210 + }, + { + "epoch": 1.6823811426288455, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4211 + }, + { + "epoch": 1.6827806632041549, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4212 + }, + { + "epoch": 1.6831801837794647, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 4213 + }, + { + "epoch": 1.6835797043547742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4214 + }, + { + "epoch": 1.6839792249300838, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 4215 + }, + { + "epoch": 1.6843787455053936, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4216 + }, + { + "epoch": 1.6847782660807031, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 4217 + }, + { + "epoch": 1.6851777866560127, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4218 + }, + { + "epoch": 1.6855773072313225, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4219 + }, + { + "epoch": 1.685976827806632, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 4220 + }, + { + "epoch": 1.6863763483819416, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 4221 + }, + { + "epoch": 1.6867758689572514, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4222 + }, + { + "epoch": 1.687175389532561, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4223 + }, + { + "epoch": 1.6875749101078705, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4224 + }, + { + "epoch": 1.6879744306831803, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4225 + }, + { + "epoch": 1.6883739512584897, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4226 + }, + { + "epoch": 1.6887734718337994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4227 + }, + { + "epoch": 1.6891729924091092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4228 + }, + { + "epoch": 1.6895725129844186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4229 + }, + { + "epoch": 1.6899720335597284, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4230 + }, + { + "epoch": 1.690371554135038, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4231 + }, + { + "epoch": 1.6907710747103475, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4232 + }, + { + "epoch": 1.6911705952856573, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4233 + }, + { + "epoch": 1.6915701158609668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4234 + }, + { + "epoch": 1.6919696364362764, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4235 + }, + { + "epoch": 1.6923691570115862, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4236 + }, + { + "epoch": 1.6927686775868958, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4237 + }, + { + "epoch": 1.6931681981622053, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4238 + }, + { + "epoch": 1.693567718737515, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4239 + }, + { + "epoch": 1.6939672393128244, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4240 + }, + { + "epoch": 1.6943667598881342, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4241 + }, + { + "epoch": 1.694766280463444, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4242 + }, + { + "epoch": 1.6951658010387534, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4243 + }, + { + "epoch": 1.6955653216140631, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4244 + }, + { + "epoch": 1.6959648421893727, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4245 + }, + { + "epoch": 1.6963643627646823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4246 + }, + { + "epoch": 1.696763883339992, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4247 + }, + { + "epoch": 1.6971634039153016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4248 + }, + { + "epoch": 1.6975629244906112, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4249 + }, + { + "epoch": 1.697962445065921, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4250 + }, + { + "epoch": 1.6983619656412305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4251 + }, + { + "epoch": 1.69876148621654, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4252 + }, + { + "epoch": 1.69916100679185, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4253 + }, + { + "epoch": 1.6995605273671595, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 4254 + }, + { + "epoch": 1.699960047942469, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4255 + }, + { + "epoch": 1.7003595685177788, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4256 + }, + { + "epoch": 1.7007590890930881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4257 + }, + { + "epoch": 1.701158609668398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4258 + }, + { + "epoch": 1.7015581302437075, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4259 + }, + { + "epoch": 1.701957650819017, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4260 + }, + { + "epoch": 1.7023571713943269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4261 + }, + { + "epoch": 1.7027566919696364, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4262 + }, + { + "epoch": 1.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4263 + }, + { + "epoch": 1.7035557331202558, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4264 + }, + { + "epoch": 1.7039552536955653, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4265 + }, + { + "epoch": 1.704354774270875, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4266 + }, + { + "epoch": 1.7047542948461847, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4267 + }, + { + "epoch": 1.7051538154214942, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4268 + }, + { + "epoch": 1.7055533359968038, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4269 + }, + { + "epoch": 1.7059528565721136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4270 + }, + { + "epoch": 1.706352377147423, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4271 + }, + { + "epoch": 1.7067518977227327, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4272 + }, + { + "epoch": 1.7071514182980425, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4273 + }, + { + "epoch": 1.7075509388733519, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4274 + }, + { + "epoch": 1.7079504594486616, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4275 + }, + { + "epoch": 1.7083499800239712, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4276 + }, + { + "epoch": 1.7087495005992808, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4277 + }, + { + "epoch": 1.7091490211745906, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4278 + }, + { + "epoch": 1.7095485417499001, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4279 + }, + { + "epoch": 1.7099480623252097, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4280 + }, + { + "epoch": 1.7103475829005195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4281 + }, + { + "epoch": 1.710747103475829, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4282 + }, + { + "epoch": 1.7111466240511386, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4283 + }, + { + "epoch": 1.7115461446264484, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4284 + }, + { + "epoch": 1.7119456652017577, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4285 + }, + { + "epoch": 1.7123451857770675, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4286 + }, + { + "epoch": 1.7127447063523773, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4287 + }, + { + "epoch": 1.7131442269276866, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4288 + }, + { + "epoch": 1.7135437475029964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4289 + }, + { + "epoch": 1.713943268078306, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4290 + }, + { + "epoch": 1.7143427886536156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4291 + }, + { + "epoch": 1.7147423092289253, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4292 + }, + { + "epoch": 1.715141829804235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4293 + }, + { + "epoch": 1.7155413503795445, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4294 + }, + { + "epoch": 1.7159408709548543, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4295 + }, + { + "epoch": 1.7163403915301638, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4296 + }, + { + "epoch": 1.7167399121054734, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4297 + }, + { + "epoch": 1.7171394326807832, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4298 + }, + { + "epoch": 1.7175389532560927, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4299 + }, + { + "epoch": 1.7179384738314023, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 4300 + }, + { + "epoch": 1.718337994406712, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4301 + }, + { + "epoch": 1.7187375149820214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4302 + }, + { + "epoch": 1.7191370355573312, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4303 + }, + { + "epoch": 1.7195365561326408, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4304 + }, + { + "epoch": 1.7199360767079503, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4305 + }, + { + "epoch": 1.7203355972832601, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4306 + }, + { + "epoch": 1.7207351178585697, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4307 + }, + { + "epoch": 1.7211346384338793, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 4308 + }, + { + "epoch": 1.721534159009189, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4309 + }, + { + "epoch": 1.7219336795844986, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4310 + }, + { + "epoch": 1.7223332001598082, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4311 + }, + { + "epoch": 1.722732720735118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4312 + }, + { + "epoch": 1.7231322413104275, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4313 + }, + { + "epoch": 1.723531761885737, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4314 + }, + { + "epoch": 1.7239312824610469, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4315 + }, + { + "epoch": 1.7243308030363562, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4316 + }, + { + "epoch": 1.724730323611666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4317 + }, + { + "epoch": 1.7251298441869758, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4318 + }, + { + "epoch": 1.7255293647622851, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4319 + }, + { + "epoch": 1.725928885337595, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4320 + }, + { + "epoch": 1.7263284059129045, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4321 + }, + { + "epoch": 1.726727926488214, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4322 + }, + { + "epoch": 1.7271274470635238, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4323 + }, + { + "epoch": 1.7275269676388334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4324 + }, + { + "epoch": 1.727926488214143, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4325 + }, + { + "epoch": 1.7283260087894528, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4326 + }, + { + "epoch": 1.7287255293647623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4327 + }, + { + "epoch": 1.7291250499400719, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4328 + }, + { + "epoch": 1.7295245705153817, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4329 + }, + { + "epoch": 1.729924091090691, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4330 + }, + { + "epoch": 1.7303236116660008, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4331 + }, + { + "epoch": 1.7307231322413106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4332 + }, + { + "epoch": 1.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4333 + }, + { + "epoch": 1.7315221733919297, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4334 + }, + { + "epoch": 1.7319216939672393, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4335 + }, + { + "epoch": 1.7323212145425488, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4336 + }, + { + "epoch": 1.7327207351178586, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4337 + }, + { + "epoch": 1.7331202556931682, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4338 + }, + { + "epoch": 1.7335197762684778, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4339 + }, + { + "epoch": 1.7339192968437875, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4340 + }, + { + "epoch": 1.734318817419097, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4341 + }, + { + "epoch": 1.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4342 + }, + { + "epoch": 1.7351178585697165, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4343 + }, + { + "epoch": 1.735517379145026, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4344 + }, + { + "epoch": 1.7359168997203356, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4345 + }, + { + "epoch": 1.7363164202956454, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4346 + }, + { + "epoch": 1.7367159408709547, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4347 + }, + { + "epoch": 1.7371154614462645, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4348 + }, + { + "epoch": 1.737514982021574, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4349 + }, + { + "epoch": 1.7379145025968836, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4350 + }, + { + "epoch": 1.7383140231721934, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4351 + }, + { + "epoch": 1.738713543747503, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4352 + }, + { + "epoch": 1.7391130643228125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4353 + }, + { + "epoch": 1.7395125848981223, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4354 + }, + { + "epoch": 1.739912105473432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4355 + }, + { + "epoch": 1.7403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4356 + }, + { + "epoch": 1.7407111466240512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 4357 + }, + { + "epoch": 1.7411106671993608, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4358 + }, + { + "epoch": 1.7415101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4359 + }, + { + "epoch": 1.7419097083499802, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4360 + }, + { + "epoch": 1.7423092289252895, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4361 + }, + { + "epoch": 1.7427087495005993, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4362 + }, + { + "epoch": 1.743108270075909, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4363 + }, + { + "epoch": 1.7435077906512184, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4364 + }, + { + "epoch": 1.7439073112265282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4365 + }, + { + "epoch": 1.7443068318018378, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4366 + }, + { + "epoch": 1.7447063523771473, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4367 + }, + { + "epoch": 1.7451058729524571, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4368 + }, + { + "epoch": 1.7455053935277667, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4369 + }, + { + "epoch": 1.7459049141030762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4370 + }, + { + "epoch": 1.746304434678386, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4371 + }, + { + "epoch": 1.7467039552536956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4372 + }, + { + "epoch": 1.7471034758290052, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4373 + }, + { + "epoch": 1.747502996404315, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4374 + }, + { + "epoch": 1.7479025169796243, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4375 + }, + { + "epoch": 1.748302037554934, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4376 + }, + { + "epoch": 1.7487015581302439, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4377 + }, + { + "epoch": 1.7491010787055532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4378 + }, + { + "epoch": 1.749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4379 + }, + { + "epoch": 1.7499001198561726, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4380 + }, + { + "epoch": 1.7502996404314821, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4381 + }, + { + "epoch": 1.750699161006792, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4382 + }, + { + "epoch": 1.7510986815821015, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4383 + }, + { + "epoch": 1.751498202157411, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4384 + }, + { + "epoch": 1.7518977227327208, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4385 + }, + { + "epoch": 1.7522972433080304, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4386 + }, + { + "epoch": 1.75269676388334, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4387 + }, + { + "epoch": 1.7530962844586497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4388 + }, + { + "epoch": 1.7534958050339593, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4389 + }, + { + "epoch": 1.7538953256092689, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4390 + }, + { + "epoch": 1.7542948461845786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4391 + }, + { + "epoch": 1.754694366759888, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4392 + }, + { + "epoch": 1.7550938873351978, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 4393 + }, + { + "epoch": 1.7554934079105073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4394 + }, + { + "epoch": 1.755892928485817, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4395 + }, + { + "epoch": 1.7562924490611267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4396 + }, + { + "epoch": 1.7566919696364363, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4397 + }, + { + "epoch": 1.7570914902117458, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4398 + }, + { + "epoch": 1.7574910107870556, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4399 + }, + { + "epoch": 1.7578905313623652, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4400 + }, + { + "epoch": 1.7582900519376747, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4401 + }, + { + "epoch": 1.7586895725129845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4402 + }, + { + "epoch": 1.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4403 + }, + { + "epoch": 1.7594886136636037, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4404 + }, + { + "epoch": 1.7598881342389134, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4405 + }, + { + "epoch": 1.7602876548142228, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 4406 + }, + { + "epoch": 1.7606871753895326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 4407 + }, + { + "epoch": 1.7610866959648424, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4408 + }, + { + "epoch": 1.7614862165401517, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 4409 + }, + { + "epoch": 1.7618857371154615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4410 + }, + { + "epoch": 1.762285257690771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4411 + }, + { + "epoch": 1.7626847782660806, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4412 + }, + { + "epoch": 1.7630842988413904, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4413 + }, + { + "epoch": 1.7634838194167, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4414 + }, + { + "epoch": 1.7638833399920095, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4415 + }, + { + "epoch": 1.7642828605673193, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4416 + }, + { + "epoch": 1.7646823811426289, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4417 + }, + { + "epoch": 1.7650819017179384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4418 + }, + { + "epoch": 1.7654814222932482, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4419 + }, + { + "epoch": 1.7658809428685576, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4420 + }, + { + "epoch": 1.7662804634438674, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4421 + }, + { + "epoch": 1.7666799840191771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4422 + }, + { + "epoch": 1.7670795045944865, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4423 + }, + { + "epoch": 1.7674790251697963, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4424 + }, + { + "epoch": 1.7678785457451058, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4425 + }, + { + "epoch": 1.7682780663204154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4426 + }, + { + "epoch": 1.7686775868957252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4427 + }, + { + "epoch": 1.7690771074710347, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4428 + }, + { + "epoch": 1.7694766280463443, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4429 + }, + { + "epoch": 1.769876148621654, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4430 + }, + { + "epoch": 1.7702756691969637, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4431 + }, + { + "epoch": 1.7706751897722732, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4432 + }, + { + "epoch": 1.771074710347583, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4433 + }, + { + "epoch": 1.7714742309228926, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4434 + }, + { + "epoch": 1.7718737514982021, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4435 + }, + { + "epoch": 1.772273272073512, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4436 + }, + { + "epoch": 1.7726727926488213, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4437 + }, + { + "epoch": 1.773072313224131, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4438 + }, + { + "epoch": 1.7734718337994406, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4439 + }, + { + "epoch": 1.7738713543747502, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4440 + }, + { + "epoch": 1.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4441 + }, + { + "epoch": 1.7746703955253695, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4442 + }, + { + "epoch": 1.775069916100679, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 4443 + }, + { + "epoch": 1.7754694366759889, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 4444 + }, + { + "epoch": 1.7758689572512985, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4445 + }, + { + "epoch": 1.776268477826608, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 4446 + }, + { + "epoch": 1.7766679984019178, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 4447 + }, + { + "epoch": 1.7770675189772274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4448 + }, + { + "epoch": 1.777467039552537, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 4449 + }, + { + "epoch": 1.7778665601278467, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 4450 + }, + { + "epoch": 1.778266080703156, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4451 + }, + { + "epoch": 1.7786656012784658, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4452 + }, + { + "epoch": 1.7790651218537756, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4453 + }, + { + "epoch": 1.779464642429085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4454 + }, + { + "epoch": 1.7798641630043948, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4455 + }, + { + "epoch": 1.7802636835797043, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4456 + }, + { + "epoch": 1.780663204155014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4457 + }, + { + "epoch": 1.7810627247303237, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4458 + }, + { + "epoch": 1.7814622453056332, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4459 + }, + { + "epoch": 1.7818617658809428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4460 + }, + { + "epoch": 1.7822612864562526, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4461 + }, + { + "epoch": 1.7826608070315622, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4462 + }, + { + "epoch": 1.7830603276068717, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4463 + }, + { + "epoch": 1.7834598481821815, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4464 + }, + { + "epoch": 1.7838593687574908, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4465 + }, + { + "epoch": 1.7842588893328006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4466 + }, + { + "epoch": 1.7846584099081104, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4467 + }, + { + "epoch": 1.7850579304834198, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4468 + }, + { + "epoch": 1.7854574510587295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4469 + }, + { + "epoch": 1.7858569716340391, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4470 + }, + { + "epoch": 1.7862564922093487, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4471 + }, + { + "epoch": 1.7866560127846585, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4472 + }, + { + "epoch": 1.787055533359968, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4473 + }, + { + "epoch": 1.7874550539352776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4474 + }, + { + "epoch": 1.7878545745105874, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4475 + }, + { + "epoch": 1.788254095085897, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4476 + }, + { + "epoch": 1.7886536156612065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4477 + }, + { + "epoch": 1.7890531362365163, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4478 + }, + { + "epoch": 1.7894526568118259, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4479 + }, + { + "epoch": 1.7898521773871354, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4480 + }, + { + "epoch": 1.7902516979624452, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4481 + }, + { + "epoch": 1.7906512185377546, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 4482 + }, + { + "epoch": 1.7910507391130643, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4483 + }, + { + "epoch": 1.791450259688374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4484 + }, + { + "epoch": 1.7918497802636835, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4485 + }, + { + "epoch": 1.7922493008389933, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4486 + }, + { + "epoch": 1.7926488214143028, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4487 + }, + { + "epoch": 1.7930483419896124, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4488 + }, + { + "epoch": 1.7934478625649222, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4489 + }, + { + "epoch": 1.7938473831402317, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4490 + }, + { + "epoch": 1.7942469037155413, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4491 + }, + { + "epoch": 1.794646424290851, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4492 + }, + { + "epoch": 1.7950459448661606, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4493 + }, + { + "epoch": 1.7954454654414702, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4494 + }, + { + "epoch": 1.79584498601678, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4495 + }, + { + "epoch": 1.7962445065920893, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4496 + }, + { + "epoch": 1.7966440271673991, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4497 + }, + { + "epoch": 1.797043547742709, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4498 + }, + { + "epoch": 1.7974430683180183, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4499 + }, + { + "epoch": 1.797842588893328, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4500 + }, + { + "epoch": 1.7982421094686376, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4501 + }, + { + "epoch": 1.7986416300439472, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4502 + }, + { + "epoch": 1.799041150619257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4503 + }, + { + "epoch": 1.7994406711945665, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4504 + }, + { + "epoch": 1.799840191769876, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4505 + }, + { + "epoch": 1.8002397123451859, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4506 + }, + { + "epoch": 1.8006392329204954, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4507 + }, + { + "epoch": 1.801038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4508 + }, + { + "epoch": 1.8014382740711148, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4509 + }, + { + "epoch": 1.8018377946464241, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4510 + }, + { + "epoch": 1.802237315221734, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4511 + }, + { + "epoch": 1.8026368357970437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4512 + }, + { + "epoch": 1.803036356372353, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4513 + }, + { + "epoch": 1.8034358769476628, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4514 + }, + { + "epoch": 1.8038353975229724, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4515 + }, + { + "epoch": 1.804234918098282, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 4516 + }, + { + "epoch": 1.8046344386735917, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4517 + }, + { + "epoch": 1.8050339592489013, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4518 + }, + { + "epoch": 1.8054334798242109, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4519 + }, + { + "epoch": 1.8058330003995207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4520 + }, + { + "epoch": 1.8062325209748302, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4521 + }, + { + "epoch": 1.8066320415501398, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 4522 + }, + { + "epoch": 1.8070315621254496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4523 + }, + { + "epoch": 1.8074310827007591, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4524 + }, + { + "epoch": 1.8078306032760687, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4525 + }, + { + "epoch": 1.8082301238513785, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4526 + }, + { + "epoch": 1.8086296444266878, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4527 + }, + { + "epoch": 1.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4528 + }, + { + "epoch": 1.8094286855773072, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4529 + }, + { + "epoch": 1.8098282061526167, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4530 + }, + { + "epoch": 1.8102277267279265, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4531 + }, + { + "epoch": 1.810627247303236, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4532 + }, + { + "epoch": 1.8110267678785457, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4533 + }, + { + "epoch": 1.8114262884538554, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4534 + }, + { + "epoch": 1.811825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4535 + }, + { + "epoch": 1.8122253296044746, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4536 + }, + { + "epoch": 1.8126248501797844, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4537 + }, + { + "epoch": 1.813024370755094, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 4538 + }, + { + "epoch": 1.8134238913304035, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4539 + }, + { + "epoch": 1.8138234119057133, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 4540 + }, + { + "epoch": 1.8142229324810226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 4541 + }, + { + "epoch": 1.8146224530563324, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4542 + }, + { + "epoch": 1.8150219736316422, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4543 + }, + { + "epoch": 1.8154214942069515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4544 + }, + { + "epoch": 1.8158210147822613, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4545 + }, + { + "epoch": 1.8162205353575709, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4546 + }, + { + "epoch": 1.8166200559328805, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4547 + }, + { + "epoch": 1.8170195765081902, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4548 + }, + { + "epoch": 1.8174190970834998, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4549 + }, + { + "epoch": 1.8178186176588094, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4550 + }, + { + "epoch": 1.8182181382341192, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4551 + }, + { + "epoch": 1.8186176588094287, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4552 + }, + { + "epoch": 1.8190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4553 + }, + { + "epoch": 1.819416699960048, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4554 + }, + { + "epoch": 1.8198162205353574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4555 + }, + { + "epoch": 1.8202157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4556 + }, + { + "epoch": 1.820615261685977, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4557 + }, + { + "epoch": 1.8210147822612863, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4558 + }, + { + "epoch": 1.821414302836596, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4559 + }, + { + "epoch": 1.8218138234119057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4560 + }, + { + "epoch": 1.8222133439872152, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4561 + }, + { + "epoch": 1.822612864562525, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4562 + }, + { + "epoch": 1.8230123851378346, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4563 + }, + { + "epoch": 1.8234119057131442, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4564 + }, + { + "epoch": 1.823811426288454, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4565 + }, + { + "epoch": 1.8242109468637635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4566 + }, + { + "epoch": 1.824610467439073, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4567 + }, + { + "epoch": 1.8250099880143829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4568 + }, + { + "epoch": 1.8254095085896924, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4569 + }, + { + "epoch": 1.825809029165002, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4570 + }, + { + "epoch": 1.8262085497403118, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4571 + }, + { + "epoch": 1.8266080703156211, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4572 + }, + { + "epoch": 1.827007590890931, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4573 + }, + { + "epoch": 1.8274071114662405, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4574 + }, + { + "epoch": 1.82780663204155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4575 + }, + { + "epoch": 1.8282061526168598, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4576 + }, + { + "epoch": 1.8286056731921694, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4577 + }, + { + "epoch": 1.829005193767479, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4578 + }, + { + "epoch": 1.8294047143427887, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4579 + }, + { + "epoch": 1.8298042349180983, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4580 + }, + { + "epoch": 1.8302037554934079, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4581 + }, + { + "epoch": 1.8306032760687176, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4582 + }, + { + "epoch": 1.8310027966440272, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4583 + }, + { + "epoch": 1.8314023172193368, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4584 + }, + { + "epoch": 1.8318018377946466, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4585 + }, + { + "epoch": 1.832201358369956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4586 + }, + { + "epoch": 1.8326008789452657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4587 + }, + { + "epoch": 1.8330003995205755, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4588 + }, + { + "epoch": 1.8333999200958848, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4589 + }, + { + "epoch": 1.8337994406711946, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4590 + }, + { + "epoch": 1.8341989612465042, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4591 + }, + { + "epoch": 1.8345984818218137, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4592 + }, + { + "epoch": 1.8349980023971235, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4593 + }, + { + "epoch": 1.835397522972433, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4594 + }, + { + "epoch": 1.8357970435477426, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4595 + }, + { + "epoch": 1.8361965641230524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4596 + }, + { + "epoch": 1.836596084698362, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4597 + }, + { + "epoch": 1.8369956052736716, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4598 + }, + { + "epoch": 1.8373951258489813, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4599 + }, + { + "epoch": 1.8377946464242907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4600 + }, + { + "epoch": 1.8381941669996005, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4601 + }, + { + "epoch": 1.8385936875749103, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4602 + }, + { + "epoch": 1.8389932081502196, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4603 + }, + { + "epoch": 1.8393927287255294, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4604 + }, + { + "epoch": 1.839792249300839, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4605 + }, + { + "epoch": 1.8401917698761485, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4606 + }, + { + "epoch": 1.8405912904514583, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4607 + }, + { + "epoch": 1.8409908110267679, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 4608 + }, + { + "epoch": 1.8413903316020774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4609 + }, + { + "epoch": 1.8417898521773872, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4610 + }, + { + "epoch": 1.8421893727526968, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4611 + }, + { + "epoch": 1.8425888933280063, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4612 + }, + { + "epoch": 1.8429884139033161, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4613 + }, + { + "epoch": 1.8433879344786257, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4614 + }, + { + "epoch": 1.8437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4615 + }, + { + "epoch": 1.844186975629245, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4616 + }, + { + "epoch": 1.8445864962045544, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4617 + }, + { + "epoch": 1.8449860167798642, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4618 + }, + { + "epoch": 1.8453855373551737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4619 + }, + { + "epoch": 1.8457850579304833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4620 + }, + { + "epoch": 1.846184578505793, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4621 + }, + { + "epoch": 1.8465840990811027, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4622 + }, + { + "epoch": 1.8469836196564122, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4623 + }, + { + "epoch": 1.847383140231722, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4624 + }, + { + "epoch": 1.8477826608070316, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4625 + }, + { + "epoch": 1.8481821813823411, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4626 + }, + { + "epoch": 1.848581701957651, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4627 + }, + { + "epoch": 1.8489812225329605, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4628 + }, + { + "epoch": 1.84938074310827, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4629 + }, + { + "epoch": 1.8497802636835798, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4630 + }, + { + "epoch": 1.8501797842588892, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4631 + }, + { + "epoch": 1.850579304834199, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4632 + }, + { + "epoch": 1.8509788254095088, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 4633 + }, + { + "epoch": 1.851378345984818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4634 + }, + { + "epoch": 1.8517778665601279, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4635 + }, + { + "epoch": 1.8521773871354374, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4636 + }, + { + "epoch": 1.852576907710747, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4637 + }, + { + "epoch": 1.8529764282860568, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4638 + }, + { + "epoch": 1.8533759488613664, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4639 + }, + { + "epoch": 1.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4640 + }, + { + "epoch": 1.8541749900119857, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4641 + }, + { + "epoch": 1.8545745105872953, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 4642 + }, + { + "epoch": 1.8549740311626048, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4643 + }, + { + "epoch": 1.8553735517379146, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4644 + }, + { + "epoch": 1.855773072313224, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4645 + }, + { + "epoch": 1.8561725928885338, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4646 + }, + { + "epoch": 1.8565721134638435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4647 + }, + { + "epoch": 1.8569716340391529, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4648 + }, + { + "epoch": 1.8573711546144627, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4649 + }, + { + "epoch": 1.8577706751897722, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4650 + }, + { + "epoch": 1.8581701957650818, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4651 + }, + { + "epoch": 1.8585697163403916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4652 + }, + { + "epoch": 1.8589692369157012, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4653 + }, + { + "epoch": 1.8593687574910107, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4654 + }, + { + "epoch": 1.8597682780663205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4655 + }, + { + "epoch": 1.86016779864163, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4656 + }, + { + "epoch": 1.8605673192169396, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4657 + }, + { + "epoch": 1.8609668397922494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4658 + }, + { + "epoch": 1.861366360367559, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4659 + }, + { + "epoch": 1.8617658809428685, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4660 + }, + { + "epoch": 1.8621654015181783, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4661 + }, + { + "epoch": 1.8625649220934877, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4662 + }, + { + "epoch": 1.8629644426687975, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4663 + }, + { + "epoch": 1.863363963244107, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4664 + }, + { + "epoch": 1.8637634838194166, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4665 + }, + { + "epoch": 1.8641630043947264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4666 + }, + { + "epoch": 1.864562524970036, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4667 + }, + { + "epoch": 1.8649620455453455, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4668 + }, + { + "epoch": 1.8653615661206553, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4669 + }, + { + "epoch": 1.8657610866959649, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4670 + }, + { + "epoch": 1.8661606072712744, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4671 + }, + { + "epoch": 1.8665601278465842, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4672 + }, + { + "epoch": 1.8669596484218938, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4673 + }, + { + "epoch": 1.8673591689972033, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4674 + }, + { + "epoch": 1.8677586895725131, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4675 + }, + { + "epoch": 1.8681582101478225, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4676 + }, + { + "epoch": 1.8685577307231322, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4677 + }, + { + "epoch": 1.868957251298442, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4678 + }, + { + "epoch": 1.8693567718737514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4679 + }, + { + "epoch": 1.8697562924490612, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4680 + }, + { + "epoch": 1.8701558130243707, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4681 + }, + { + "epoch": 1.8705553335996803, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4682 + }, + { + "epoch": 1.87095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4683 + }, + { + "epoch": 1.8713543747502996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4684 + }, + { + "epoch": 1.8717538953256092, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4685 + }, + { + "epoch": 1.872153415900919, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4686 + }, + { + "epoch": 1.8725529364762286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4687 + }, + { + "epoch": 1.8729524570515381, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4688 + }, + { + "epoch": 1.873351977626848, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 4689 + }, + { + "epoch": 1.8737514982021573, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4690 + }, + { + "epoch": 1.874151018777467, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4691 + }, + { + "epoch": 1.8745505393527768, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4692 + }, + { + "epoch": 1.8749500599280862, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4693 + }, + { + "epoch": 1.875349580503396, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4694 + }, + { + "epoch": 1.8757491010787055, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4695 + }, + { + "epoch": 1.876148621654015, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4696 + }, + { + "epoch": 1.8765481422293249, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4697 + }, + { + "epoch": 1.8769476628046344, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4698 + }, + { + "epoch": 1.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4699 + }, + { + "epoch": 1.8777467039552538, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4700 + }, + { + "epoch": 1.8781462245305633, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4701 + }, + { + "epoch": 1.878545745105873, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4702 + }, + { + "epoch": 1.8789452656811827, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4703 + }, + { + "epoch": 1.8793447862564923, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4704 + }, + { + "epoch": 1.8797443068318018, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4705 + }, + { + "epoch": 1.8801438274071116, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4706 + }, + { + "epoch": 1.880543347982421, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4707 + }, + { + "epoch": 1.8809428685577307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4708 + }, + { + "epoch": 1.8813423891330403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4709 + }, + { + "epoch": 1.8817419097083499, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4710 + }, + { + "epoch": 1.8821414302836597, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4711 + }, + { + "epoch": 1.8825409508589692, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4712 + }, + { + "epoch": 1.8829404714342788, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4713 + }, + { + "epoch": 1.8833399920095886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4714 + }, + { + "epoch": 1.8837395125848981, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4715 + }, + { + "epoch": 1.8841390331602077, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4716 + }, + { + "epoch": 1.8845385537355175, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4717 + }, + { + "epoch": 1.884938074310827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4718 + }, + { + "epoch": 1.8853375948861366, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4719 + }, + { + "epoch": 1.8857371154614464, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4720 + }, + { + "epoch": 1.8861366360367557, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4721 + }, + { + "epoch": 1.8865361566120655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4722 + }, + { + "epoch": 1.8869356771873753, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4723 + }, + { + "epoch": 1.8873351977626847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4724 + }, + { + "epoch": 1.8877347183379944, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4725 + }, + { + "epoch": 1.888134238913304, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4726 + }, + { + "epoch": 1.8885337594886136, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4727 + }, + { + "epoch": 1.8889332800639234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4728 + }, + { + "epoch": 1.889332800639233, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4729 + }, + { + "epoch": 1.8897323212145425, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4730 + }, + { + "epoch": 1.8901318417898523, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4731 + }, + { + "epoch": 1.8905313623651618, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4732 + }, + { + "epoch": 1.8909308829404714, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 4733 + }, + { + "epoch": 1.8913304035157812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4734 + }, + { + "epoch": 1.8917299240910905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4735 + }, + { + "epoch": 1.8921294446664003, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 4736 + }, + { + "epoch": 1.89252896524171, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4737 + }, + { + "epoch": 1.8929284858170194, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4738 + }, + { + "epoch": 1.8933280063923292, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4739 + }, + { + "epoch": 1.8937275269676388, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4740 + }, + { + "epoch": 1.8941270475429484, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4741 + }, + { + "epoch": 1.8945265681182581, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4742 + }, + { + "epoch": 1.8949260886935677, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4743 + }, + { + "epoch": 1.8953256092688773, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4744 + }, + { + "epoch": 1.895725129844187, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4745 + }, + { + "epoch": 1.8961246504194966, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4746 + }, + { + "epoch": 1.8965241709948062, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4747 + }, + { + "epoch": 1.896923691570116, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4748 + }, + { + "epoch": 1.8973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4749 + }, + { + "epoch": 1.897722732720735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4750 + }, + { + "epoch": 1.898122253296045, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4751 + }, + { + "epoch": 1.8985217738713542, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 4752 + }, + { + "epoch": 1.898921294446664, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4753 + }, + { + "epoch": 1.8993208150219736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4754 + }, + { + "epoch": 1.8997203355972831, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4755 + }, + { + "epoch": 1.900119856172593, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4756 + }, + { + "epoch": 1.9005193767479025, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4757 + }, + { + "epoch": 1.900918897323212, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4758 + }, + { + "epoch": 1.9013184178985219, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4759 + }, + { + "epoch": 1.9017179384738314, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4760 + }, + { + "epoch": 1.902117459049141, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4761 + }, + { + "epoch": 1.9025169796244508, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4762 + }, + { + "epoch": 1.9029165001997603, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4763 + }, + { + "epoch": 1.90331602077507, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4764 + }, + { + "epoch": 1.9037155413503797, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4765 + }, + { + "epoch": 1.904115061925689, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4766 + }, + { + "epoch": 1.9045145825009988, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4767 + }, + { + "epoch": 1.9049141030763086, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4768 + }, + { + "epoch": 1.905313623651618, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4769 + }, + { + "epoch": 1.9057131442269277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4770 + }, + { + "epoch": 1.9061126648022373, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4771 + }, + { + "epoch": 1.9065121853775469, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4772 + }, + { + "epoch": 1.9069117059528566, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4773 + }, + { + "epoch": 1.9073112265281662, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4774 + }, + { + "epoch": 1.9077107471034758, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4775 + }, + { + "epoch": 1.9081102676787856, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4776 + }, + { + "epoch": 1.9085097882540951, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4777 + }, + { + "epoch": 1.9089093088294047, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4778 + }, + { + "epoch": 1.9093088294047145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4779 + }, + { + "epoch": 1.9097083499800238, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4780 + }, + { + "epoch": 1.9101078705553336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4781 + }, + { + "epoch": 1.9105073911306434, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4782 + }, + { + "epoch": 1.9109069117059527, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4783 + }, + { + "epoch": 1.9113064322812625, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4784 + }, + { + "epoch": 1.911705952856572, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4785 + }, + { + "epoch": 1.9121054734318816, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4786 + }, + { + "epoch": 1.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4787 + }, + { + "epoch": 1.912904514582501, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4788 + }, + { + "epoch": 1.9133040351578106, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4789 + }, + { + "epoch": 1.9137035557331203, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 4790 + }, + { + "epoch": 1.91410307630843, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4791 + }, + { + "epoch": 1.9145025968837395, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4792 + }, + { + "epoch": 1.9149021174590493, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4793 + }, + { + "epoch": 1.9153016380343588, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4794 + }, + { + "epoch": 1.9157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4795 + }, + { + "epoch": 1.9161006791849782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4796 + }, + { + "epoch": 1.9165001997602875, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4797 + }, + { + "epoch": 1.9168997203355973, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4798 + }, + { + "epoch": 1.9172992409109069, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4799 + }, + { + "epoch": 1.9176987614862164, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 4800 + }, + { + "epoch": 1.9180982820615262, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4801 + }, + { + "epoch": 1.9184978026368358, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4802 + }, + { + "epoch": 1.9188973232121453, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 4803 + }, + { + "epoch": 1.9192968437874551, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 4804 + }, + { + "epoch": 1.9196963643627647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4805 + }, + { + "epoch": 1.9200958849380743, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4806 + }, + { + "epoch": 1.920495405513384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4807 + }, + { + "epoch": 1.9208949260886936, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4808 + }, + { + "epoch": 1.9212944466640032, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4809 + }, + { + "epoch": 1.921693967239313, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4810 + }, + { + "epoch": 1.9220934878146223, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4811 + }, + { + "epoch": 1.922493008389932, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4812 + }, + { + "epoch": 1.9228925289652419, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4813 + }, + { + "epoch": 1.9232920495405512, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4814 + }, + { + "epoch": 1.923691570115861, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 4815 + }, + { + "epoch": 1.9240910906911706, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4816 + }, + { + "epoch": 1.9244906112664801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4817 + }, + { + "epoch": 1.92489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4818 + }, + { + "epoch": 1.9252896524170995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4819 + }, + { + "epoch": 1.925689172992409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4820 + }, + { + "epoch": 1.9260886935677188, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4821 + }, + { + "epoch": 1.9264882141430284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4822 + }, + { + "epoch": 1.926887734718338, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4823 + }, + { + "epoch": 1.9272872552936477, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4824 + }, + { + "epoch": 1.927686775868957, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4825 + }, + { + "epoch": 1.9280862964442669, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4826 + }, + { + "epoch": 1.9284858170195767, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4827 + }, + { + "epoch": 1.928885337594886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4828 + }, + { + "epoch": 1.9292848581701958, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4829 + }, + { + "epoch": 1.9296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4830 + }, + { + "epoch": 1.930083899320815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4831 + }, + { + "epoch": 1.9304834198961247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4832 + }, + { + "epoch": 1.9308829404714343, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4833 + }, + { + "epoch": 1.9312824610467438, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4834 + }, + { + "epoch": 1.9316819816220536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4835 + }, + { + "epoch": 1.9320815021973632, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4836 + }, + { + "epoch": 1.9324810227726728, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4837 + }, + { + "epoch": 1.9328805433479825, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4838 + }, + { + "epoch": 1.933280063923292, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4839 + }, + { + "epoch": 1.9336795844986017, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4840 + }, + { + "epoch": 1.9340791050739115, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4841 + }, + { + "epoch": 1.9344786256492208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4842 + }, + { + "epoch": 1.9348781462245306, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4843 + }, + { + "epoch": 1.9352776667998401, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4844 + }, + { + "epoch": 1.9356771873751497, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4845 + }, + { + "epoch": 1.9360767079504595, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4846 + }, + { + "epoch": 1.936476228525769, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4847 + }, + { + "epoch": 1.9368757491010786, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4848 + }, + { + "epoch": 1.9372752696763884, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4849 + }, + { + "epoch": 1.937674790251698, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4850 + }, + { + "epoch": 1.9380743108270075, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 4851 + }, + { + "epoch": 1.9384738314023173, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4852 + }, + { + "epoch": 1.938873351977627, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4853 + }, + { + "epoch": 1.9392728725529365, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4854 + }, + { + "epoch": 1.9396723931282462, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4855 + }, + { + "epoch": 1.9400719137035556, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4856 + }, + { + "epoch": 1.9404714342788654, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4857 + }, + { + "epoch": 1.940870954854175, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4858 + }, + { + "epoch": 1.9412704754294845, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4859 + }, + { + "epoch": 1.9416699960047943, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4860 + }, + { + "epoch": 1.9420695165801038, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4861 + }, + { + "epoch": 1.9424690371554134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4862 + }, + { + "epoch": 1.9428685577307232, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4863 + }, + { + "epoch": 1.9432680783060328, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4864 + }, + { + "epoch": 1.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4865 + }, + { + "epoch": 1.9440671194566521, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4866 + }, + { + "epoch": 1.9444666400319617, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4867 + }, + { + "epoch": 1.9448661606072712, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4868 + }, + { + "epoch": 1.945265681182581, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4869 + }, + { + "epoch": 1.9456652017578904, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4870 + }, + { + "epoch": 1.9460647223332002, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4871 + }, + { + "epoch": 1.94646424290851, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4872 + }, + { + "epoch": 1.9468637634838193, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4873 + }, + { + "epoch": 1.947263284059129, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4874 + }, + { + "epoch": 1.9476628046344386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4875 + }, + { + "epoch": 1.9480623252097482, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4876 + }, + { + "epoch": 1.948461845785058, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4877 + }, + { + "epoch": 1.9488613663603676, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4878 + }, + { + "epoch": 1.9492608869356771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4879 + }, + { + "epoch": 1.949660407510987, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4880 + }, + { + "epoch": 1.9500599280862965, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4881 + }, + { + "epoch": 1.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4882 + }, + { + "epoch": 1.9508589692369158, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4883 + }, + { + "epoch": 1.9512584898122254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4884 + }, + { + "epoch": 1.951658010387535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4885 + }, + { + "epoch": 1.9520575309628447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4886 + }, + { + "epoch": 1.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4887 + }, + { + "epoch": 1.9528565721134639, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4888 + }, + { + "epoch": 1.9532560926887734, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4889 + }, + { + "epoch": 1.953655613264083, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 4890 + }, + { + "epoch": 1.9540551338393928, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4891 + }, + { + "epoch": 1.9544546544147023, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4892 + }, + { + "epoch": 1.954854174990012, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4893 + }, + { + "epoch": 1.9552536955653217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4894 + }, + { + "epoch": 1.9556532161406313, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4895 + }, + { + "epoch": 1.9560527367159408, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4896 + }, + { + "epoch": 1.9564522572912506, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4897 + }, + { + "epoch": 1.9568517778665602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4898 + }, + { + "epoch": 1.9572512984418697, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4899 + }, + { + "epoch": 1.9576508190171795, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4900 + }, + { + "epoch": 1.9580503395924889, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4901 + }, + { + "epoch": 1.9584498601677987, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4902 + }, + { + "epoch": 1.9588493807431082, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4903 + }, + { + "epoch": 1.9592489013184178, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4904 + }, + { + "epoch": 1.9596484218937276, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4905 + }, + { + "epoch": 1.9600479424690371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4906 + }, + { + "epoch": 1.9604474630443467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4907 + }, + { + "epoch": 1.9608469836196565, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4908 + }, + { + "epoch": 1.961246504194966, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4909 + }, + { + "epoch": 1.9616460247702756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4910 + }, + { + "epoch": 1.9620455453455854, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4911 + }, + { + "epoch": 1.962445065920895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4912 + }, + { + "epoch": 1.9628445864962045, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 4913 + }, + { + "epoch": 1.9632441070715143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4914 + }, + { + "epoch": 1.9636436276468237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 4915 + }, + { + "epoch": 1.9640431482221334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4916 + }, + { + "epoch": 1.9644426687974432, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 4917 + }, + { + "epoch": 1.9648421893727526, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4918 + }, + { + "epoch": 1.9652417099480624, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4919 + }, + { + "epoch": 1.965641230523372, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4920 + }, + { + "epoch": 1.9660407510986815, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4921 + }, + { + "epoch": 1.9664402716739913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4922 + }, + { + "epoch": 1.9668397922493008, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4923 + }, + { + "epoch": 1.9672393128246104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4924 + }, + { + "epoch": 1.9676388333999202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 4925 + }, + { + "epoch": 1.9680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4926 + }, + { + "epoch": 1.9684378745505393, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4927 + }, + { + "epoch": 1.968837395125849, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4928 + }, + { + "epoch": 1.9692369157011587, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4929 + }, + { + "epoch": 1.9696364362764682, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4930 + }, + { + "epoch": 1.970035956851778, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4931 + }, + { + "epoch": 1.9704354774270874, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4932 + }, + { + "epoch": 1.9708349980023971, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4933 + }, + { + "epoch": 1.9712345185777067, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 4934 + }, + { + "epoch": 1.9716340391530163, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4935 + }, + { + "epoch": 1.972033559728326, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4936 + }, + { + "epoch": 1.9724330803036356, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4937 + }, + { + "epoch": 1.9728326008789452, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4938 + }, + { + "epoch": 1.973232121454255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4939 + }, + { + "epoch": 1.9736316420295645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4940 + }, + { + "epoch": 1.974031162604874, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4941 + }, + { + "epoch": 1.9744306831801839, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4942 + }, + { + "epoch": 1.9748302037554935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4943 + }, + { + "epoch": 1.975229724330803, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4944 + }, + { + "epoch": 1.9756292449061128, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4945 + }, + { + "epoch": 1.9760287654814221, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4946 + }, + { + "epoch": 1.976428286056732, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4947 + }, + { + "epoch": 1.9768278066320415, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4948 + }, + { + "epoch": 1.977227327207351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4949 + }, + { + "epoch": 1.9776268477826608, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4950 + }, + { + "epoch": 1.9780263683579704, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4951 + }, + { + "epoch": 1.97842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 4952 + }, + { + "epoch": 1.9788254095085898, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4953 + }, + { + "epoch": 1.9792249300838993, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4954 + }, + { + "epoch": 1.979624450659209, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4955 + }, + { + "epoch": 1.9800239712345187, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4956 + }, + { + "epoch": 1.9804234918098282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4957 + }, + { + "epoch": 1.9808230123851378, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4958 + }, + { + "epoch": 1.9812225329604476, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4959 + }, + { + "epoch": 1.981622053535757, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4960 + }, + { + "epoch": 1.9820215741110667, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4961 + }, + { + "epoch": 1.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4962 + }, + { + "epoch": 1.9828206152616858, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 4963 + }, + { + "epoch": 1.9832201358369956, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4964 + }, + { + "epoch": 1.9836196564123052, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4965 + }, + { + "epoch": 1.9840191769876148, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4966 + }, + { + "epoch": 1.9844186975629245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4967 + }, + { + "epoch": 1.9848182181382341, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4968 + }, + { + "epoch": 1.9852177387135437, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4969 + }, + { + "epoch": 1.9856172592888535, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4970 + }, + { + "epoch": 1.986016779864163, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4971 + }, + { + "epoch": 1.9864163004394726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4972 + }, + { + "epoch": 1.9868158210147824, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4973 + }, + { + "epoch": 1.987215341590092, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4974 + }, + { + "epoch": 1.9876148621654015, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4975 + }, + { + "epoch": 1.9880143827407113, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4976 + }, + { + "epoch": 1.9884139033160206, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4977 + }, + { + "epoch": 1.9888134238913304, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4978 + }, + { + "epoch": 1.98921294446664, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4979 + }, + { + "epoch": 1.9896124650419496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4980 + }, + { + "epoch": 1.9900119856172593, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4981 + }, + { + "epoch": 1.990411506192569, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4982 + }, + { + "epoch": 1.9908110267678785, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4983 + }, + { + "epoch": 1.9912105473431883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4984 + }, + { + "epoch": 1.9916100679184978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4985 + }, + { + "epoch": 1.9920095884938074, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4986 + }, + { + "epoch": 1.9924091090691172, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4987 + }, + { + "epoch": 1.9928086296444267, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4988 + }, + { + "epoch": 1.9932081502197363, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4989 + }, + { + "epoch": 1.993607670795046, + "grad_norm": 0.04541015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4990 + }, + { + "epoch": 1.9940071913703554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4991 + }, + { + "epoch": 1.9944067119456652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4992 + }, + { + "epoch": 1.9948062325209748, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 4993 + }, + { + "epoch": 1.9952057530962843, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4994 + }, + { + "epoch": 1.9956052736715941, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4995 + }, + { + "epoch": 1.9960047942469037, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4996 + }, + { + "epoch": 1.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4997 + }, + { + "epoch": 1.996803835397523, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4998 + }, + { + "epoch": 1.9972033559728326, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4999 + }, + { + "epoch": 1.9976028765481422, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5000 + }, + { + "epoch": 1.998002397123452, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5001 + }, + { + "epoch": 1.9984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5002 + }, + { + "epoch": 1.998801438274071, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5003 + }, + { + "epoch": 1.9992009588493809, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5004 + }, + { + "epoch": 1.9996004794246902, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5005 + }, + { + "epoch": 2.0, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5006 + }, + { + "epoch": 2.00039952057531, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5007 + }, + { + "epoch": 2.000799041150619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5008 + }, + { + "epoch": 2.001198561725929, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5009 + }, + { + "epoch": 2.0015980823012387, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5010 + }, + { + "epoch": 2.001997602876548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5011 + }, + { + "epoch": 2.002397123451858, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5012 + }, + { + "epoch": 2.0027966440271676, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5013 + }, + { + "epoch": 2.003196164602477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 5014 + }, + { + "epoch": 2.0035956851777867, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5015 + }, + { + "epoch": 2.003995205753096, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5016 + }, + { + "epoch": 2.004394726328406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5017 + }, + { + "epoch": 2.0047942469037157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5018 + }, + { + "epoch": 2.005193767479025, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5019 + }, + { + "epoch": 2.005593288054335, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5020 + }, + { + "epoch": 2.0059928086296446, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5021 + }, + { + "epoch": 2.006392329204954, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5022 + }, + { + "epoch": 2.0067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5023 + }, + { + "epoch": 2.0071913703555735, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 5024 + }, + { + "epoch": 2.007590890930883, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5025 + }, + { + "epoch": 2.0079904115061926, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5026 + }, + { + "epoch": 2.0083899320815024, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5027 + }, + { + "epoch": 2.0087894526568117, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 5028 + }, + { + "epoch": 2.0091889732321215, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 5029 + }, + { + "epoch": 2.009588493807431, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5030 + }, + { + "epoch": 2.0099880143827407, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5031 + }, + { + "epoch": 2.0103875349580504, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5032 + }, + { + "epoch": 2.01078705553336, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5033 + }, + { + "epoch": 2.0111865761086696, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5034 + }, + { + "epoch": 2.0115860966839794, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5035 + }, + { + "epoch": 2.0119856172592887, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5036 + }, + { + "epoch": 2.0123851378345985, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5037 + }, + { + "epoch": 2.0127846584099083, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5038 + }, + { + "epoch": 2.0131841789852176, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5039 + }, + { + "epoch": 2.0135836995605274, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5040 + }, + { + "epoch": 2.013983220135837, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 5041 + }, + { + "epoch": 2.0143827407111465, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 5042 + }, + { + "epoch": 2.0147822612864563, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 5043 + }, + { + "epoch": 2.0151817818617657, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5044 + }, + { + "epoch": 2.0155813024370755, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5045 + }, + { + "epoch": 2.0159808230123852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5046 + }, + { + "epoch": 2.0163803435876946, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5047 + }, + { + "epoch": 2.0167798641630044, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5048 + }, + { + "epoch": 2.017179384738314, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5049 + }, + { + "epoch": 2.0175789053136235, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5050 + }, + { + "epoch": 2.0179784258889333, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5051 + }, + { + "epoch": 2.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5052 + }, + { + "epoch": 2.0187774670395524, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5053 + }, + { + "epoch": 2.019176987614862, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5054 + }, + { + "epoch": 2.019576508190172, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5055 + }, + { + "epoch": 2.0199760287654813, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5056 + }, + { + "epoch": 2.020375549340791, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5057 + }, + { + "epoch": 2.020775069916101, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5058 + }, + { + "epoch": 2.0211745904914102, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5059 + }, + { + "epoch": 2.02157411106672, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5060 + }, + { + "epoch": 2.0219736316420294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5061 + }, + { + "epoch": 2.022373152217339, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5062 + }, + { + "epoch": 2.022772672792649, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5063 + }, + { + "epoch": 2.0231721933679583, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5064 + }, + { + "epoch": 2.023571713943268, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5065 + }, + { + "epoch": 2.023971234518578, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5066 + }, + { + "epoch": 2.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5067 + }, + { + "epoch": 2.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5068 + }, + { + "epoch": 2.0251697962445068, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5069 + }, + { + "epoch": 2.025569316819816, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5070 + }, + { + "epoch": 2.025968837395126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5071 + }, + { + "epoch": 2.0263683579704357, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5072 + }, + { + "epoch": 2.026767878545745, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5073 + }, + { + "epoch": 2.027167399121055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5074 + }, + { + "epoch": 2.027566919696364, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5075 + }, + { + "epoch": 2.027966440271674, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5076 + }, + { + "epoch": 2.0283659608469837, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5077 + }, + { + "epoch": 2.028765481422293, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5078 + }, + { + "epoch": 2.029165001997603, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5079 + }, + { + "epoch": 2.0295645225729126, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5080 + }, + { + "epoch": 2.029964043148222, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5081 + }, + { + "epoch": 2.0303635637235318, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5082 + }, + { + "epoch": 2.0307630842988416, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5083 + }, + { + "epoch": 2.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5084 + }, + { + "epoch": 2.0315621254494607, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 5085 + }, + { + "epoch": 2.0319616460247705, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5086 + }, + { + "epoch": 2.03236116660008, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5087 + }, + { + "epoch": 2.0327606871753896, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5088 + }, + { + "epoch": 2.033160207750699, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5089 + }, + { + "epoch": 2.0335597283260087, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5090 + }, + { + "epoch": 2.0339592489013185, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5091 + }, + { + "epoch": 2.034358769476628, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5092 + }, + { + "epoch": 2.0347582900519376, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5093 + }, + { + "epoch": 2.0351578106272474, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5094 + }, + { + "epoch": 2.0355573312025568, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5095 + }, + { + "epoch": 2.0359568517778666, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5096 + }, + { + "epoch": 2.0363563723531763, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5097 + }, + { + "epoch": 2.0367558929284857, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5098 + }, + { + "epoch": 2.0371554135037955, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5099 + }, + { + "epoch": 2.0375549340791053, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 5100 + }, + { + "epoch": 2.0379544546544146, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5101 + }, + { + "epoch": 2.0383539752297244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5102 + }, + { + "epoch": 2.038753495805034, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5103 + }, + { + "epoch": 2.0391530163803435, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5104 + }, + { + "epoch": 2.0395525369556533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5105 + }, + { + "epoch": 2.0399520575309626, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5106 + }, + { + "epoch": 2.0403515781062724, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5107 + }, + { + "epoch": 2.040751098681582, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 5108 + }, + { + "epoch": 2.0411506192568916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 5109 + }, + { + "epoch": 2.0415501398322013, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5110 + }, + { + "epoch": 2.041949660407511, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5111 + }, + { + "epoch": 2.0423491809828205, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5112 + }, + { + "epoch": 2.0427487015581303, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5113 + }, + { + "epoch": 2.04314822213344, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5114 + }, + { + "epoch": 2.0435477427087494, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5115 + }, + { + "epoch": 2.043947263284059, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 5116 + }, + { + "epoch": 2.044346783859369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5117 + }, + { + "epoch": 2.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5118 + }, + { + "epoch": 2.045145825009988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5119 + }, + { + "epoch": 2.0455453455852974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 5120 + }, + { + "epoch": 2.0459448661606072, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5121 + }, + { + "epoch": 2.046344386735917, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5122 + }, + { + "epoch": 2.0467439073112264, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5123 + }, + { + "epoch": 2.047143427886536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5124 + }, + { + "epoch": 2.047542948461846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5125 + }, + { + "epoch": 2.0479424690371553, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5126 + }, + { + "epoch": 2.048341989612465, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5127 + }, + { + "epoch": 2.048741510187775, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5128 + }, + { + "epoch": 2.049141030763084, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5129 + }, + { + "epoch": 2.049540551338394, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5130 + }, + { + "epoch": 2.0499400719137038, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5131 + }, + { + "epoch": 2.050339592489013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5132 + }, + { + "epoch": 2.050739113064323, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5133 + }, + { + "epoch": 2.0511386336396322, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5134 + }, + { + "epoch": 2.051538154214942, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5135 + }, + { + "epoch": 2.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 5136 + }, + { + "epoch": 2.052337195365561, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5137 + }, + { + "epoch": 2.052736715940871, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 5138 + }, + { + "epoch": 2.0531362365161807, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5139 + }, + { + "epoch": 2.05353575709149, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5140 + }, + { + "epoch": 2.0539352776668, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5141 + }, + { + "epoch": 2.0543347982421096, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5142 + }, + { + "epoch": 2.054734318817419, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5143 + }, + { + "epoch": 2.0551338393927288, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5144 + }, + { + "epoch": 2.0555333599680385, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 5145 + }, + { + "epoch": 2.055932880543348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5146 + }, + { + "epoch": 2.0563324011186577, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5147 + }, + { + "epoch": 2.0567319216939675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5148 + }, + { + "epoch": 2.057131442269277, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5149 + }, + { + "epoch": 2.0575309628445866, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5150 + }, + { + "epoch": 2.057930483419896, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5151 + }, + { + "epoch": 2.0583300039952057, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5152 + }, + { + "epoch": 2.0587295245705155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5153 + }, + { + "epoch": 2.059129045145825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5154 + }, + { + "epoch": 2.0595285657211346, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5155 + }, + { + "epoch": 2.0599280862964444, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5156 + }, + { + "epoch": 2.0603276068717538, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5157 + }, + { + "epoch": 2.0607271274470635, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5158 + }, + { + "epoch": 2.0611266480223733, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5159 + }, + { + "epoch": 2.0615261685976827, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5160 + }, + { + "epoch": 2.0619256891729925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5161 + }, + { + "epoch": 2.0623252097483022, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5162 + }, + { + "epoch": 2.0627247303236116, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5163 + }, + { + "epoch": 2.0631242508989214, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5164 + }, + { + "epoch": 2.0635237714742307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5165 + }, + { + "epoch": 2.0639232920495405, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5166 + }, + { + "epoch": 2.0643228126248503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5167 + }, + { + "epoch": 2.0647223332001596, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5168 + }, + { + "epoch": 2.0651218537754694, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5169 + }, + { + "epoch": 2.065521374350779, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5170 + }, + { + "epoch": 2.0659208949260885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5171 + }, + { + "epoch": 2.0663204155013983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5172 + }, + { + "epoch": 2.066719936076708, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5173 + }, + { + "epoch": 2.0671194566520175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 5174 + }, + { + "epoch": 2.0675189772273272, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5175 + }, + { + "epoch": 2.067918497802637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5176 + }, + { + "epoch": 2.0683180183779464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 5177 + }, + { + "epoch": 2.068717538953256, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5178 + }, + { + "epoch": 2.0691170595285655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5179 + }, + { + "epoch": 2.0695165801038753, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5180 + }, + { + "epoch": 2.069916100679185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5181 + }, + { + "epoch": 2.0703156212544944, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5182 + }, + { + "epoch": 2.070715141829804, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5183 + }, + { + "epoch": 2.071114662405114, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5184 + }, + { + "epoch": 2.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5185 + }, + { + "epoch": 2.071913703555733, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5186 + }, + { + "epoch": 2.072313224131043, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5187 + }, + { + "epoch": 2.0727127447063523, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5188 + }, + { + "epoch": 2.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5189 + }, + { + "epoch": 2.073511785856972, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 5190 + }, + { + "epoch": 2.073911306432281, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5191 + }, + { + "epoch": 2.074310827007591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5192 + }, + { + "epoch": 2.0747103475829007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5193 + }, + { + "epoch": 2.07510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5194 + }, + { + "epoch": 2.07550938873352, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5195 + }, + { + "epoch": 2.075908909308829, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5196 + }, + { + "epoch": 2.076308429884139, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5197 + }, + { + "epoch": 2.076707950459449, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5198 + }, + { + "epoch": 2.077107471034758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5199 + }, + { + "epoch": 2.077506991610068, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5200 + }, + { + "epoch": 2.0779065121853777, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5201 + }, + { + "epoch": 2.078306032760687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5202 + }, + { + "epoch": 2.078705553335997, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5203 + }, + { + "epoch": 2.0791050739113066, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5204 + }, + { + "epoch": 2.079504594486616, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 5205 + }, + { + "epoch": 2.0799041150619257, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5206 + }, + { + "epoch": 2.0803036356372355, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5207 + }, + { + "epoch": 2.080703156212545, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5208 + }, + { + "epoch": 2.0811026767878547, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5209 + }, + { + "epoch": 2.081502197363164, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5210 + }, + { + "epoch": 2.081901717938474, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5211 + }, + { + "epoch": 2.0823012385137836, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5212 + }, + { + "epoch": 2.082700759089093, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5213 + }, + { + "epoch": 2.0831002796644027, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 5214 + }, + { + "epoch": 2.0834998002397125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5215 + }, + { + "epoch": 2.083899320815022, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5216 + }, + { + "epoch": 2.0842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5217 + }, + { + "epoch": 2.0846983619656414, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5218 + }, + { + "epoch": 2.0850978825409507, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5219 + }, + { + "epoch": 2.0854974031162605, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5220 + }, + { + "epoch": 2.0858969236915703, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5221 + }, + { + "epoch": 2.0862964442668797, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 5222 + }, + { + "epoch": 2.0866959648421894, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5223 + }, + { + "epoch": 2.087095485417499, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5224 + }, + { + "epoch": 2.0874950059928086, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5225 + }, + { + "epoch": 2.0878945265681184, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5226 + }, + { + "epoch": 2.0882940471434277, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5227 + }, + { + "epoch": 2.0886935677187375, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5228 + }, + { + "epoch": 2.0890930882940473, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5229 + }, + { + "epoch": 2.0894926088693566, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5230 + }, + { + "epoch": 2.0898921294446664, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5231 + }, + { + "epoch": 2.090291650019976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5232 + }, + { + "epoch": 2.0906911705952855, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5233 + }, + { + "epoch": 2.0910906911705953, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5234 + }, + { + "epoch": 2.091490211745905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5235 + }, + { + "epoch": 2.0918897323212144, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5236 + }, + { + "epoch": 2.0922892528965242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5237 + }, + { + "epoch": 2.092688773471834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5238 + }, + { + "epoch": 2.0930882940471434, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5239 + }, + { + "epoch": 2.093487814622453, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5240 + }, + { + "epoch": 2.0938873351977625, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5241 + }, + { + "epoch": 2.0942868557730723, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5242 + }, + { + "epoch": 2.094686376348382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5243 + }, + { + "epoch": 2.0950858969236914, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5244 + }, + { + "epoch": 2.095485417499001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5245 + }, + { + "epoch": 2.095884938074311, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5246 + }, + { + "epoch": 2.0962844586496203, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5247 + }, + { + "epoch": 2.09668397922493, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5248 + }, + { + "epoch": 2.09708349980024, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5249 + }, + { + "epoch": 2.0974830203755492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5250 + }, + { + "epoch": 2.097882540950859, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5251 + }, + { + "epoch": 2.098282061526169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 5252 + }, + { + "epoch": 2.098681582101478, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5253 + }, + { + "epoch": 2.099081102676788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5254 + }, + { + "epoch": 2.0994806232520973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5255 + }, + { + "epoch": 2.099880143827407, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5256 + }, + { + "epoch": 2.100279664402717, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5257 + }, + { + "epoch": 2.100679184978026, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5258 + }, + { + "epoch": 2.101078705553336, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5259 + }, + { + "epoch": 2.1014782261286458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5260 + }, + { + "epoch": 2.101877746703955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5261 + }, + { + "epoch": 2.102277267279265, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5262 + }, + { + "epoch": 2.1026767878545747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5263 + }, + { + "epoch": 2.103076308429884, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5264 + }, + { + "epoch": 2.103475829005194, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5265 + }, + { + "epoch": 2.1038753495805036, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5266 + }, + { + "epoch": 2.104274870155813, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5267 + }, + { + "epoch": 2.1046743907311227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5268 + }, + { + "epoch": 2.105073911306432, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5269 + }, + { + "epoch": 2.105473431881742, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5270 + }, + { + "epoch": 2.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5271 + }, + { + "epoch": 2.106272473032361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5272 + }, + { + "epoch": 2.1066719936076708, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5273 + }, + { + "epoch": 2.1070715141829806, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5274 + }, + { + "epoch": 2.10747103475829, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5275 + }, + { + "epoch": 2.1078705553335997, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5276 + }, + { + "epoch": 2.1082700759089095, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5277 + }, + { + "epoch": 2.108669596484219, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5278 + }, + { + "epoch": 2.1090691170595286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5279 + }, + { + "epoch": 2.1094686376348384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5280 + }, + { + "epoch": 2.1098681582101477, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5281 + }, + { + "epoch": 2.1102676787854575, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5282 + }, + { + "epoch": 2.1106671993607673, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5283 + }, + { + "epoch": 2.1110667199360766, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5284 + }, + { + "epoch": 2.1114662405113864, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5285 + }, + { + "epoch": 2.1118657610866958, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5286 + }, + { + "epoch": 2.1122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5287 + }, + { + "epoch": 2.1126648022373153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5288 + }, + { + "epoch": 2.1130643228126247, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5289 + }, + { + "epoch": 2.1134638433879345, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5290 + }, + { + "epoch": 2.1138633639632443, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5291 + }, + { + "epoch": 2.1142628845385536, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5292 + }, + { + "epoch": 2.1146624051138634, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5293 + }, + { + "epoch": 2.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5294 + }, + { + "epoch": 2.1154614462644825, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5295 + }, + { + "epoch": 2.1158609668397923, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5296 + }, + { + "epoch": 2.116260487415102, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 5297 + }, + { + "epoch": 2.1166600079904114, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5298 + }, + { + "epoch": 2.117059528565721, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5299 + }, + { + "epoch": 2.1174590491410306, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5300 + }, + { + "epoch": 2.1178585697163403, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5301 + }, + { + "epoch": 2.11825809029165, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5302 + }, + { + "epoch": 2.1186576108669595, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5303 + }, + { + "epoch": 2.1190571314422693, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5304 + }, + { + "epoch": 2.119456652017579, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5305 + }, + { + "epoch": 2.1198561725928884, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5306 + }, + { + "epoch": 2.120255693168198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5307 + }, + { + "epoch": 2.120655213743508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5308 + }, + { + "epoch": 2.1210547343188173, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5309 + }, + { + "epoch": 2.121454254894127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5310 + }, + { + "epoch": 2.121853775469437, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5311 + }, + { + "epoch": 2.122253296044746, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5312 + }, + { + "epoch": 2.122652816620056, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5313 + }, + { + "epoch": 2.1230523371953653, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5314 + }, + { + "epoch": 2.123451857770675, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5315 + }, + { + "epoch": 2.123851378345985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5316 + }, + { + "epoch": 2.1242508989212943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5317 + }, + { + "epoch": 2.124650419496604, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5318 + }, + { + "epoch": 2.125049940071914, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5319 + }, + { + "epoch": 2.125449460647223, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5320 + }, + { + "epoch": 2.125848981222533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5321 + }, + { + "epoch": 2.1262485017978427, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5322 + }, + { + "epoch": 2.126648022373152, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5323 + }, + { + "epoch": 2.127047542948462, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5324 + }, + { + "epoch": 2.1274470635237717, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5325 + }, + { + "epoch": 2.127846584099081, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5326 + }, + { + "epoch": 2.128246104674391, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5327 + }, + { + "epoch": 2.1286456252497006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5328 + }, + { + "epoch": 2.12904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5329 + }, + { + "epoch": 2.1294446664003197, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5330 + }, + { + "epoch": 2.129844186975629, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5331 + }, + { + "epoch": 2.130243707550939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5332 + }, + { + "epoch": 2.1306432281262486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5333 + }, + { + "epoch": 2.131042748701558, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5334 + }, + { + "epoch": 2.1314422692768678, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5335 + }, + { + "epoch": 2.1318417898521775, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5336 + }, + { + "epoch": 2.132241310427487, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5337 + }, + { + "epoch": 2.1326408310027967, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5338 + }, + { + "epoch": 2.1330403515781065, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5339 + }, + { + "epoch": 2.133439872153416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5340 + }, + { + "epoch": 2.1338393927287256, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5341 + }, + { + "epoch": 2.134238913304035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5342 + }, + { + "epoch": 2.1346384338793447, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5343 + }, + { + "epoch": 2.1350379544546545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5344 + }, + { + "epoch": 2.135437475029964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5345 + }, + { + "epoch": 2.1358369956052736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5346 + }, + { + "epoch": 2.1362365161805834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5347 + }, + { + "epoch": 2.1366360367558928, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5348 + }, + { + "epoch": 2.1370355573312025, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5349 + }, + { + "epoch": 2.1374350779065123, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5350 + }, + { + "epoch": 2.1378345984818217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 5351 + }, + { + "epoch": 2.1382341190571315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5352 + }, + { + "epoch": 2.1386336396324412, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5353 + }, + { + "epoch": 2.1390331602077506, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5354 + }, + { + "epoch": 2.1394326807830604, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5355 + }, + { + "epoch": 2.13983220135837, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5356 + }, + { + "epoch": 2.1402317219336795, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 5357 + }, + { + "epoch": 2.1406312425089893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5358 + }, + { + "epoch": 2.1410307630842986, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5359 + }, + { + "epoch": 2.1414302836596084, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5360 + }, + { + "epoch": 2.141829804234918, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5361 + }, + { + "epoch": 2.1422293248102275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5362 + }, + { + "epoch": 2.1426288453855373, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5363 + }, + { + "epoch": 2.143028365960847, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5364 + }, + { + "epoch": 2.1434278865361565, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5365 + }, + { + "epoch": 2.1438274071114662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5366 + }, + { + "epoch": 2.144226927686776, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5367 + }, + { + "epoch": 2.1446264482620854, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5368 + }, + { + "epoch": 2.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5369 + }, + { + "epoch": 2.145425489412705, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5370 + }, + { + "epoch": 2.1458250099880143, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5371 + }, + { + "epoch": 2.146224530563324, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5372 + }, + { + "epoch": 2.146624051138634, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5373 + }, + { + "epoch": 2.147023571713943, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5374 + }, + { + "epoch": 2.147423092289253, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5375 + }, + { + "epoch": 2.1478226128645623, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5376 + }, + { + "epoch": 2.148222133439872, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5377 + }, + { + "epoch": 2.148621654015182, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5378 + }, + { + "epoch": 2.1490211745904912, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5379 + }, + { + "epoch": 2.149420695165801, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5380 + }, + { + "epoch": 2.149820215741111, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5381 + }, + { + "epoch": 2.15021973631642, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5382 + }, + { + "epoch": 2.15061925689173, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5383 + }, + { + "epoch": 2.1510187774670397, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5384 + }, + { + "epoch": 2.151418298042349, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5385 + }, + { + "epoch": 2.151817818617659, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5386 + }, + { + "epoch": 2.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5387 + }, + { + "epoch": 2.152616859768278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5388 + }, + { + "epoch": 2.1530163803435878, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5389 + }, + { + "epoch": 2.153415900918897, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5390 + }, + { + "epoch": 2.153815421494207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5391 + }, + { + "epoch": 2.1542149420695167, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5392 + }, + { + "epoch": 2.154614462644826, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5393 + }, + { + "epoch": 2.155013983220136, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5394 + }, + { + "epoch": 2.1554135037954456, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5395 + }, + { + "epoch": 2.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5396 + }, + { + "epoch": 2.1562125449460647, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5397 + }, + { + "epoch": 2.1566120655213745, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5398 + }, + { + "epoch": 2.157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5399 + }, + { + "epoch": 2.1574111066719937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5400 + }, + { + "epoch": 2.1578106272473034, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5401 + }, + { + "epoch": 2.158210147822613, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5402 + }, + { + "epoch": 2.1586096683979226, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5403 + }, + { + "epoch": 2.159009188973232, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5404 + }, + { + "epoch": 2.1594087095485417, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5405 + }, + { + "epoch": 2.1598082301238515, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5406 + }, + { + "epoch": 2.160207750699161, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5407 + }, + { + "epoch": 2.1606072712744706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5408 + }, + { + "epoch": 2.1610067918497804, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5409 + }, + { + "epoch": 2.1614063124250897, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5410 + }, + { + "epoch": 2.1618058330003995, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5411 + }, + { + "epoch": 2.1622053535757093, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5412 + }, + { + "epoch": 2.1626048741510187, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5413 + }, + { + "epoch": 2.1630043947263284, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5414 + }, + { + "epoch": 2.1634039153016382, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5415 + }, + { + "epoch": 2.1638034358769476, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5416 + }, + { + "epoch": 2.1642029564522574, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5417 + }, + { + "epoch": 2.164602477027567, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5418 + }, + { + "epoch": 2.1650019976028765, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5419 + }, + { + "epoch": 2.1654015181781863, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5420 + }, + { + "epoch": 2.1658010387534956, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5421 + }, + { + "epoch": 2.1662005593288054, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5422 + }, + { + "epoch": 2.166600079904115, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5423 + }, + { + "epoch": 2.1669996004794245, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5424 + }, + { + "epoch": 2.1673991210547343, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5425 + }, + { + "epoch": 2.167798641630044, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5426 + }, + { + "epoch": 2.1681981622053534, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5427 + }, + { + "epoch": 2.1685976827806632, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5428 + }, + { + "epoch": 2.168997203355973, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5429 + }, + { + "epoch": 2.1693967239312824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5430 + }, + { + "epoch": 2.169796244506592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5431 + }, + { + "epoch": 2.1701957650819015, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5432 + }, + { + "epoch": 2.1705952856572113, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5433 + }, + { + "epoch": 2.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5434 + }, + { + "epoch": 2.1713943268078304, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5435 + }, + { + "epoch": 2.17179384738314, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5436 + }, + { + "epoch": 2.17219336795845, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 5437 + }, + { + "epoch": 2.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5438 + }, + { + "epoch": 2.172992409109069, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5439 + }, + { + "epoch": 2.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5440 + }, + { + "epoch": 2.1737914502596882, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5441 + }, + { + "epoch": 2.174190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5442 + }, + { + "epoch": 2.174590491410308, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5443 + }, + { + "epoch": 2.174990011985617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5444 + }, + { + "epoch": 2.175389532560927, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5445 + }, + { + "epoch": 2.1757890531362367, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5446 + }, + { + "epoch": 2.176188573711546, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5447 + }, + { + "epoch": 2.176588094286856, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5448 + }, + { + "epoch": 2.176987614862165, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5449 + }, + { + "epoch": 2.177387135437475, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5450 + }, + { + "epoch": 2.1777866560127848, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5451 + }, + { + "epoch": 2.178186176588094, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 5452 + }, + { + "epoch": 2.178585697163404, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5453 + }, + { + "epoch": 2.1789852177387137, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5454 + }, + { + "epoch": 2.179384738314023, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5455 + }, + { + "epoch": 2.179784258889333, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5456 + }, + { + "epoch": 2.1801837794646426, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5457 + }, + { + "epoch": 2.180583300039952, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5458 + }, + { + "epoch": 2.1809828206152617, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5459 + }, + { + "epoch": 2.1813823411905715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5460 + }, + { + "epoch": 2.181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5461 + }, + { + "epoch": 2.1821813823411906, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5462 + }, + { + "epoch": 2.1825809029165004, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5463 + }, + { + "epoch": 2.1829804234918098, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5464 + }, + { + "epoch": 2.1833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5465 + }, + { + "epoch": 2.183779464642429, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 5466 + }, + { + "epoch": 2.1841789852177387, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 5467 + }, + { + "epoch": 2.1845785057930485, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5468 + }, + { + "epoch": 2.184978026368358, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5469 + }, + { + "epoch": 2.1853775469436676, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5470 + }, + { + "epoch": 2.1857770675189774, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5471 + }, + { + "epoch": 2.1861765880942867, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5472 + }, + { + "epoch": 2.1865761086695965, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 5473 + }, + { + "epoch": 2.1869756292449063, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5474 + }, + { + "epoch": 2.1873751498202156, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5475 + }, + { + "epoch": 2.1877746703955254, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5476 + }, + { + "epoch": 2.1881741909708348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5477 + }, + { + "epoch": 2.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5478 + }, + { + "epoch": 2.1889732321214543, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5479 + }, + { + "epoch": 2.1893727526967637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5480 + }, + { + "epoch": 2.1897722732720735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 5481 + }, + { + "epoch": 2.1901717938473833, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5482 + }, + { + "epoch": 2.1905713144226926, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5483 + }, + { + "epoch": 2.1909708349980024, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5484 + }, + { + "epoch": 2.191370355573312, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5485 + }, + { + "epoch": 2.1917698761486215, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5486 + }, + { + "epoch": 2.1921693967239313, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 5487 + }, + { + "epoch": 2.192568917299241, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5488 + }, + { + "epoch": 2.1929684378745504, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 5489 + }, + { + "epoch": 2.19336795844986, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5490 + }, + { + "epoch": 2.19376747902517, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5491 + }, + { + "epoch": 2.1941669996004793, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5492 + }, + { + "epoch": 2.194566520175789, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5493 + }, + { + "epoch": 2.1949660407510985, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5494 + }, + { + "epoch": 2.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5495 + }, + { + "epoch": 2.195765081901718, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5496 + }, + { + "epoch": 2.1961646024770274, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5497 + }, + { + "epoch": 2.196564123052337, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5498 + }, + { + "epoch": 2.196963643627647, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5499 + }, + { + "epoch": 2.1973631642029563, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5500 + }, + { + "epoch": 2.197762684778266, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5501 + }, + { + "epoch": 2.198162205353576, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5502 + }, + { + "epoch": 2.198561725928885, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5503 + }, + { + "epoch": 2.198961246504195, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5504 + }, + { + "epoch": 2.199360767079505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5505 + }, + { + "epoch": 2.199760287654814, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5506 + }, + { + "epoch": 2.200159808230124, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5507 + }, + { + "epoch": 2.2005593288054337, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5508 + }, + { + "epoch": 2.200958849380743, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5509 + }, + { + "epoch": 2.201358369956053, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5510 + }, + { + "epoch": 2.201757890531362, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5511 + }, + { + "epoch": 2.202157411106672, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5512 + }, + { + "epoch": 2.2025569316819817, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5513 + }, + { + "epoch": 2.202956452257291, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5514 + }, + { + "epoch": 2.203355972832601, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5515 + }, + { + "epoch": 2.2037554934079107, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5516 + }, + { + "epoch": 2.20415501398322, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5517 + }, + { + "epoch": 2.20455453455853, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5518 + }, + { + "epoch": 2.2049540551338396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5519 + }, + { + "epoch": 2.205353575709149, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5520 + }, + { + "epoch": 2.2057530962844587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5521 + }, + { + "epoch": 2.206152616859768, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5522 + }, + { + "epoch": 2.206552137435078, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5523 + }, + { + "epoch": 2.2069516580103876, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5524 + }, + { + "epoch": 2.207351178585697, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5525 + }, + { + "epoch": 2.2077506991610067, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5526 + }, + { + "epoch": 2.2081502197363165, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5527 + }, + { + "epoch": 2.208549740311626, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5528 + }, + { + "epoch": 2.2089492608869357, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5529 + }, + { + "epoch": 2.2093487814622454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5530 + }, + { + "epoch": 2.209748302037555, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5531 + }, + { + "epoch": 2.2101478226128646, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5532 + }, + { + "epoch": 2.2105473431881744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5533 + }, + { + "epoch": 2.2109468637634837, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5534 + }, + { + "epoch": 2.2113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5535 + }, + { + "epoch": 2.2117459049141033, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5536 + }, + { + "epoch": 2.2121454254894126, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5537 + }, + { + "epoch": 2.2125449460647224, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5538 + }, + { + "epoch": 2.2129444666400317, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5539 + }, + { + "epoch": 2.2133439872153415, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5540 + }, + { + "epoch": 2.2137435077906513, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5541 + }, + { + "epoch": 2.2141430283659607, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5542 + }, + { + "epoch": 2.2145425489412705, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5543 + }, + { + "epoch": 2.2149420695165802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5544 + }, + { + "epoch": 2.2153415900918896, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5545 + }, + { + "epoch": 2.2157411106671994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5546 + }, + { + "epoch": 2.216140631242509, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5547 + }, + { + "epoch": 2.2165401518178185, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5548 + }, + { + "epoch": 2.2169396723931283, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5549 + }, + { + "epoch": 2.217339192968438, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5550 + }, + { + "epoch": 2.2177387135437474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5551 + }, + { + "epoch": 2.218138234119057, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5552 + }, + { + "epoch": 2.218537754694367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5553 + }, + { + "epoch": 2.2189372752696763, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 5554 + }, + { + "epoch": 2.219336795844986, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5555 + }, + { + "epoch": 2.2197363164202955, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5556 + }, + { + "epoch": 2.2201358369956052, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5557 + }, + { + "epoch": 2.220535357570915, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5558 + }, + { + "epoch": 2.2209348781462244, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5559 + }, + { + "epoch": 2.221334398721534, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5560 + }, + { + "epoch": 2.221733919296844, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5561 + }, + { + "epoch": 2.2221334398721533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5562 + }, + { + "epoch": 2.222532960447463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5563 + }, + { + "epoch": 2.222932481022773, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5564 + }, + { + "epoch": 2.223332001598082, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5565 + }, + { + "epoch": 2.223731522173392, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5566 + }, + { + "epoch": 2.2241310427487013, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5567 + }, + { + "epoch": 2.224530563324011, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5568 + }, + { + "epoch": 2.224930083899321, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5569 + }, + { + "epoch": 2.2253296044746302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5570 + }, + { + "epoch": 2.22572912504994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5571 + }, + { + "epoch": 2.22612864562525, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5572 + }, + { + "epoch": 2.226528166200559, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5573 + }, + { + "epoch": 2.226927686775869, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5574 + }, + { + "epoch": 2.2273272073511787, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5575 + }, + { + "epoch": 2.227726727926488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5576 + }, + { + "epoch": 2.228126248501798, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5577 + }, + { + "epoch": 2.2285257690771076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5578 + }, + { + "epoch": 2.228925289652417, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5579 + }, + { + "epoch": 2.2293248102277268, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5580 + }, + { + "epoch": 2.2297243308030366, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5581 + }, + { + "epoch": 2.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5582 + }, + { + "epoch": 2.2305233719536557, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5583 + }, + { + "epoch": 2.230922892528965, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5584 + }, + { + "epoch": 2.231322413104275, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5585 + }, + { + "epoch": 2.2317219336795846, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5586 + }, + { + "epoch": 2.232121454254894, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 5587 + }, + { + "epoch": 2.2325209748302037, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5588 + }, + { + "epoch": 2.2329204954055135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 5589 + }, + { + "epoch": 2.233320015980823, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5590 + }, + { + "epoch": 2.2337195365561326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5591 + }, + { + "epoch": 2.2341190571314424, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5592 + }, + { + "epoch": 2.2345185777067518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 5593 + }, + { + "epoch": 2.2349180982820616, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5594 + }, + { + "epoch": 2.2353176188573713, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5595 + }, + { + "epoch": 2.2357171394326807, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5596 + }, + { + "epoch": 2.2361166600079905, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5597 + }, + { + "epoch": 2.2365161805833003, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5598 + }, + { + "epoch": 2.2369157011586096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5599 + }, + { + "epoch": 2.2373152217339194, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5600 + }, + { + "epoch": 2.2377147423092287, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5601 + }, + { + "epoch": 2.2381142628845385, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5602 + }, + { + "epoch": 2.2385137834598483, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5603 + }, + { + "epoch": 2.2389133040351576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5604 + }, + { + "epoch": 2.2393128246104674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 5605 + }, + { + "epoch": 2.239712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5606 + }, + { + "epoch": 2.2401118657610866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5607 + }, + { + "epoch": 2.2405113863363963, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5608 + }, + { + "epoch": 2.240910906911706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5609 + }, + { + "epoch": 2.2413104274870155, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5610 + }, + { + "epoch": 2.2417099480623253, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5611 + }, + { + "epoch": 2.2421094686376346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5612 + }, + { + "epoch": 2.2425089892129444, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5613 + }, + { + "epoch": 2.242908509788254, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5614 + }, + { + "epoch": 2.2433080303635635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5615 + }, + { + "epoch": 2.2437075509388733, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5616 + }, + { + "epoch": 2.244107071514183, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5617 + }, + { + "epoch": 2.2445065920894924, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 5618 + }, + { + "epoch": 2.2449061126648022, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5619 + }, + { + "epoch": 2.245305633240112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5620 + }, + { + "epoch": 2.2457051538154214, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5621 + }, + { + "epoch": 2.246104674390731, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5622 + }, + { + "epoch": 2.246504194966041, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5623 + }, + { + "epoch": 2.2469037155413503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5624 + }, + { + "epoch": 2.24730323611666, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5625 + }, + { + "epoch": 2.24770275669197, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5626 + }, + { + "epoch": 2.248102277267279, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5627 + }, + { + "epoch": 2.248501797842589, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5628 + }, + { + "epoch": 2.2489013184178983, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5629 + }, + { + "epoch": 2.249300838993208, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5630 + }, + { + "epoch": 2.249700359568518, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5631 + }, + { + "epoch": 2.2500998801438272, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 5632 + }, + { + "epoch": 2.250499400719137, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5633 + }, + { + "epoch": 2.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5634 + }, + { + "epoch": 2.251298441869756, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5635 + }, + { + "epoch": 2.251697962445066, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5636 + }, + { + "epoch": 2.2520974830203757, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5637 + }, + { + "epoch": 2.252497003595685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 5638 + }, + { + "epoch": 2.252896524170995, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5639 + }, + { + "epoch": 2.2532960447463046, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 5640 + }, + { + "epoch": 2.253695565321614, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5641 + }, + { + "epoch": 2.2540950858969238, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5642 + }, + { + "epoch": 2.2544946064722335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5643 + }, + { + "epoch": 2.254894127047543, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5644 + }, + { + "epoch": 2.2552936476228527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5645 + }, + { + "epoch": 2.255693168198162, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5646 + }, + { + "epoch": 2.256092688773472, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5647 + }, + { + "epoch": 2.2564922093487816, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5648 + }, + { + "epoch": 2.256891729924091, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5649 + }, + { + "epoch": 2.2572912504994007, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5650 + }, + { + "epoch": 2.2576907710747105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 5651 + }, + { + "epoch": 2.25809029165002, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5652 + }, + { + "epoch": 2.2584898122253296, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5653 + }, + { + "epoch": 2.2588893328006394, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5654 + }, + { + "epoch": 2.2592888533759488, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5655 + }, + { + "epoch": 2.2596883739512585, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5656 + }, + { + "epoch": 2.260087894526568, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5657 + }, + { + "epoch": 2.2604874151018777, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5658 + }, + { + "epoch": 2.2608869356771875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5659 + }, + { + "epoch": 2.2612864562524972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5660 + }, + { + "epoch": 2.2616859768278066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5661 + }, + { + "epoch": 2.2620854974031164, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5662 + }, + { + "epoch": 2.2624850179784257, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5663 + }, + { + "epoch": 2.2628845385537355, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5664 + }, + { + "epoch": 2.2632840591290453, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5665 + }, + { + "epoch": 2.2636835797043546, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5666 + }, + { + "epoch": 2.2640831002796644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 5667 + }, + { + "epoch": 2.264482620854974, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5668 + }, + { + "epoch": 2.2648821414302835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5669 + }, + { + "epoch": 2.2652816620055933, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5670 + }, + { + "epoch": 2.265681182580903, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5671 + }, + { + "epoch": 2.2660807031562125, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5672 + }, + { + "epoch": 2.2664802237315222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5673 + }, + { + "epoch": 2.2668797443068316, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5674 + }, + { + "epoch": 2.2672792648821414, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5675 + }, + { + "epoch": 2.267678785457451, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5676 + }, + { + "epoch": 2.2680783060327605, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5677 + }, + { + "epoch": 2.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5678 + }, + { + "epoch": 2.26887734718338, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 5679 + }, + { + "epoch": 2.2692768677586894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5680 + }, + { + "epoch": 2.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5681 + }, + { + "epoch": 2.270075908909309, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5682 + }, + { + "epoch": 2.2704754294846183, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5683 + }, + { + "epoch": 2.270874950059928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5684 + }, + { + "epoch": 2.271274470635238, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5685 + }, + { + "epoch": 2.2716739912105472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5686 + }, + { + "epoch": 2.272073511785857, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5687 + }, + { + "epoch": 2.272473032361167, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5688 + }, + { + "epoch": 2.272872552936476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5689 + }, + { + "epoch": 2.273272073511786, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5690 + }, + { + "epoch": 2.2736715940870953, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5691 + }, + { + "epoch": 2.274071114662405, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5692 + }, + { + "epoch": 2.274470635237715, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5693 + }, + { + "epoch": 2.274870155813024, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5694 + }, + { + "epoch": 2.275269676388334, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5695 + }, + { + "epoch": 2.275669196963644, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 5696 + }, + { + "epoch": 2.276068717538953, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5697 + }, + { + "epoch": 2.276468238114263, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5698 + }, + { + "epoch": 2.2768677586895727, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5699 + }, + { + "epoch": 2.277267279264882, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 5700 + }, + { + "epoch": 2.277666799840192, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5701 + }, + { + "epoch": 2.278066320415501, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5702 + }, + { + "epoch": 2.278465840990811, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5703 + }, + { + "epoch": 2.2788653615661207, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5704 + }, + { + "epoch": 2.2792648821414305, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5705 + }, + { + "epoch": 2.27966440271674, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5706 + }, + { + "epoch": 2.2800639232920497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5707 + }, + { + "epoch": 2.280463443867359, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5708 + }, + { + "epoch": 2.280862964442669, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5709 + }, + { + "epoch": 2.2812624850179786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5710 + }, + { + "epoch": 2.281662005593288, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5711 + }, + { + "epoch": 2.2820615261685977, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5712 + }, + { + "epoch": 2.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5713 + }, + { + "epoch": 2.282860567319217, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5714 + }, + { + "epoch": 2.2832600878945266, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5715 + }, + { + "epoch": 2.2836596084698364, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5716 + }, + { + "epoch": 2.2840591290451457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5717 + }, + { + "epoch": 2.2844586496204555, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5718 + }, + { + "epoch": 2.284858170195765, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5719 + }, + { + "epoch": 2.2852576907710747, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5720 + }, + { + "epoch": 2.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5721 + }, + { + "epoch": 2.286056731921694, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5722 + }, + { + "epoch": 2.2864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5723 + }, + { + "epoch": 2.2868557730723134, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5724 + }, + { + "epoch": 2.2872552936476227, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5725 + }, + { + "epoch": 2.2876548142229325, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5726 + }, + { + "epoch": 2.2880543347982423, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5727 + }, + { + "epoch": 2.2884538553735516, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5728 + }, + { + "epoch": 2.2888533759488614, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5729 + }, + { + "epoch": 2.289252896524171, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5730 + }, + { + "epoch": 2.2896524170994805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5731 + }, + { + "epoch": 2.2900519376747903, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5732 + }, + { + "epoch": 2.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5733 + }, + { + "epoch": 2.2908509788254094, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5734 + }, + { + "epoch": 2.2912504994007192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 5735 + }, + { + "epoch": 2.2916500199760286, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 5736 + }, + { + "epoch": 2.2920495405513384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5737 + }, + { + "epoch": 2.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5738 + }, + { + "epoch": 2.2928485817019575, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5739 + }, + { + "epoch": 2.2932481022772673, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5740 + }, + { + "epoch": 2.293647622852577, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5741 + }, + { + "epoch": 2.2940471434278864, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5742 + }, + { + "epoch": 2.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5743 + }, + { + "epoch": 2.294846184578506, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5744 + }, + { + "epoch": 2.2952457051538153, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5745 + }, + { + "epoch": 2.295645225729125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5746 + }, + { + "epoch": 2.2960447463044344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5747 + }, + { + "epoch": 2.2964442668797442, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5748 + }, + { + "epoch": 2.296843787455054, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5749 + }, + { + "epoch": 2.297243308030364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5750 + }, + { + "epoch": 2.297642828605673, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5751 + }, + { + "epoch": 2.298042349180983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5752 + }, + { + "epoch": 2.2984418697562923, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5753 + }, + { + "epoch": 2.298841390331602, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5754 + }, + { + "epoch": 2.299240910906912, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5755 + }, + { + "epoch": 2.299640431482221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5756 + }, + { + "epoch": 2.300039952057531, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5757 + }, + { + "epoch": 2.3004394726328408, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5758 + }, + { + "epoch": 2.30083899320815, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 5759 + }, + { + "epoch": 2.30123851378346, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5760 + }, + { + "epoch": 2.3016380343587697, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5761 + }, + { + "epoch": 2.302037554934079, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5762 + }, + { + "epoch": 2.302437075509389, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5763 + }, + { + "epoch": 2.302836596084698, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5764 + }, + { + "epoch": 2.303236116660008, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5765 + }, + { + "epoch": 2.3036356372353177, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5766 + }, + { + "epoch": 2.304035157810627, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5767 + }, + { + "epoch": 2.304434678385937, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5768 + }, + { + "epoch": 2.3048341989612466, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5769 + }, + { + "epoch": 2.305233719536556, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5770 + }, + { + "epoch": 2.3056332401118658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5771 + }, + { + "epoch": 2.3060327606871756, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5772 + }, + { + "epoch": 2.306432281262485, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5773 + }, + { + "epoch": 2.3068318018377947, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5774 + }, + { + "epoch": 2.3072313224131045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5775 + }, + { + "epoch": 2.307630842988414, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5776 + }, + { + "epoch": 2.3080303635637236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5777 + }, + { + "epoch": 2.3084298841390334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5778 + }, + { + "epoch": 2.3088294047143427, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5779 + }, + { + "epoch": 2.3092289252896525, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5780 + }, + { + "epoch": 2.309628445864962, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5781 + }, + { + "epoch": 2.3100279664402716, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5782 + }, + { + "epoch": 2.3104274870155814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5783 + }, + { + "epoch": 2.3108270075908908, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5784 + }, + { + "epoch": 2.3112265281662006, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5785 + }, + { + "epoch": 2.3116260487415103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5786 + }, + { + "epoch": 2.3120255693168197, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5787 + }, + { + "epoch": 2.3124250898921295, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5788 + }, + { + "epoch": 2.3128246104674393, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5789 + }, + { + "epoch": 2.3132241310427486, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5790 + }, + { + "epoch": 2.3136236516180584, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5791 + }, + { + "epoch": 2.3140231721933677, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5792 + }, + { + "epoch": 2.3144226927686775, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5793 + }, + { + "epoch": 2.3148222133439873, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 5794 + }, + { + "epoch": 2.315221733919297, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5795 + }, + { + "epoch": 2.3156212544946064, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5796 + }, + { + "epoch": 2.316020775069916, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5797 + }, + { + "epoch": 2.3164202956452256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5798 + }, + { + "epoch": 2.3168198162205353, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5799 + }, + { + "epoch": 2.317219336795845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5800 + }, + { + "epoch": 2.3176188573711545, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5801 + }, + { + "epoch": 2.3180183779464643, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5802 + }, + { + "epoch": 2.318417898521774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5803 + }, + { + "epoch": 2.3188174190970834, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5804 + }, + { + "epoch": 2.319216939672393, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5805 + }, + { + "epoch": 2.319616460247703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5806 + }, + { + "epoch": 2.3200159808230123, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5807 + }, + { + "epoch": 2.320415501398322, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5808 + }, + { + "epoch": 2.3208150219736314, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5809 + }, + { + "epoch": 2.321214542548941, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5810 + }, + { + "epoch": 2.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5811 + }, + { + "epoch": 2.3220135836995603, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5812 + }, + { + "epoch": 2.32241310427487, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5813 + }, + { + "epoch": 2.32281262485018, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5814 + }, + { + "epoch": 2.3232121454254893, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5815 + }, + { + "epoch": 2.323611666000799, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5816 + }, + { + "epoch": 2.324011186576109, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5817 + }, + { + "epoch": 2.324410707151418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5818 + }, + { + "epoch": 2.324810227726728, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5819 + }, + { + "epoch": 2.3252097483020377, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5820 + }, + { + "epoch": 2.325609268877347, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5821 + }, + { + "epoch": 2.326008789452657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5822 + }, + { + "epoch": 2.3264083100279667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5823 + }, + { + "epoch": 2.326807830603276, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5824 + }, + { + "epoch": 2.327207351178586, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5825 + }, + { + "epoch": 2.327606871753895, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5826 + }, + { + "epoch": 2.328006392329205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5827 + }, + { + "epoch": 2.3284059129045147, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5828 + }, + { + "epoch": 2.328805433479824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5829 + }, + { + "epoch": 2.329204954055134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5830 + }, + { + "epoch": 2.3296044746304436, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5831 + }, + { + "epoch": 2.330003995205753, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5832 + }, + { + "epoch": 2.3304035157810628, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 5833 + }, + { + "epoch": 2.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5834 + }, + { + "epoch": 2.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5835 + }, + { + "epoch": 2.3316020775069917, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5836 + }, + { + "epoch": 2.332001598082301, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5837 + }, + { + "epoch": 2.332401118657611, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 5838 + }, + { + "epoch": 2.3328006392329206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5839 + }, + { + "epoch": 2.3332001598082304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5840 + }, + { + "epoch": 2.3335996803835397, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5841 + }, + { + "epoch": 2.3339992009588495, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5842 + }, + { + "epoch": 2.334398721534159, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5843 + }, + { + "epoch": 2.3347982421094686, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5844 + }, + { + "epoch": 2.3351977626847784, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5845 + }, + { + "epoch": 2.3355972832600878, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5846 + }, + { + "epoch": 2.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5847 + }, + { + "epoch": 2.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5848 + }, + { + "epoch": 2.3367958449860167, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5849 + }, + { + "epoch": 2.3371953655613265, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5850 + }, + { + "epoch": 2.3375948861366362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5851 + }, + { + "epoch": 2.3379944067119456, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5852 + }, + { + "epoch": 2.3383939272872554, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5853 + }, + { + "epoch": 2.3387934478625647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5854 + }, + { + "epoch": 2.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5855 + }, + { + "epoch": 2.3395924890131843, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5856 + }, + { + "epoch": 2.3399920095884936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5857 + }, + { + "epoch": 2.3403915301638034, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5858 + }, + { + "epoch": 2.340791050739113, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5859 + }, + { + "epoch": 2.3411905713144225, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5860 + }, + { + "epoch": 2.3415900918897323, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5861 + }, + { + "epoch": 2.341989612465042, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5862 + }, + { + "epoch": 2.3423891330403515, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5863 + }, + { + "epoch": 2.3427886536156612, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5864 + }, + { + "epoch": 2.343188174190971, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5865 + }, + { + "epoch": 2.3435876947662804, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5866 + }, + { + "epoch": 2.34398721534159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5867 + }, + { + "epoch": 2.3443867359169, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5868 + }, + { + "epoch": 2.3447862564922093, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5869 + }, + { + "epoch": 2.345185777067519, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5870 + }, + { + "epoch": 2.3455852976428284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5871 + }, + { + "epoch": 2.345984818218138, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5872 + }, + { + "epoch": 2.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5873 + }, + { + "epoch": 2.3467838593687573, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5874 + }, + { + "epoch": 2.347183379944067, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5875 + }, + { + "epoch": 2.347582900519377, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5876 + }, + { + "epoch": 2.3479824210946862, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 5877 + }, + { + "epoch": 2.348381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5878 + }, + { + "epoch": 2.348781462245306, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5879 + }, + { + "epoch": 2.349180982820615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5880 + }, + { + "epoch": 2.349580503395925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 5881 + }, + { + "epoch": 2.3499800239712343, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5882 + }, + { + "epoch": 2.350379544546544, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5883 + }, + { + "epoch": 2.350779065121854, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5884 + }, + { + "epoch": 2.3511785856971636, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5885 + }, + { + "epoch": 2.351578106272473, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5886 + }, + { + "epoch": 2.3519776268477828, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5887 + }, + { + "epoch": 2.352377147423092, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5888 + }, + { + "epoch": 2.352776667998402, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5889 + }, + { + "epoch": 2.3531761885737117, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5890 + }, + { + "epoch": 2.353575709149021, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5891 + }, + { + "epoch": 2.353975229724331, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5892 + }, + { + "epoch": 2.3543747502996406, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5893 + }, + { + "epoch": 2.35477427087495, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5894 + }, + { + "epoch": 2.3551737914502597, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5895 + }, + { + "epoch": 2.3555733120255695, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5896 + }, + { + "epoch": 2.355972832600879, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 5897 + }, + { + "epoch": 2.3563723531761886, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5898 + }, + { + "epoch": 2.356771873751498, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5899 + }, + { + "epoch": 2.357171394326808, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5900 + }, + { + "epoch": 2.3575709149021176, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5901 + }, + { + "epoch": 2.357970435477427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5902 + }, + { + "epoch": 2.3583699560527367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5903 + }, + { + "epoch": 2.3587694766280465, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5904 + }, + { + "epoch": 2.359168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5905 + }, + { + "epoch": 2.3595685177786656, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 5906 + }, + { + "epoch": 2.3599680383539754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5907 + }, + { + "epoch": 2.3603675589292847, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5908 + }, + { + "epoch": 2.3607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5909 + }, + { + "epoch": 2.361166600079904, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5910 + }, + { + "epoch": 2.3615661206552137, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5911 + }, + { + "epoch": 2.3619656412305234, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5912 + }, + { + "epoch": 2.3623651618058332, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5913 + }, + { + "epoch": 2.3627646823811426, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5914 + }, + { + "epoch": 2.3631642029564524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5915 + }, + { + "epoch": 2.3635637235317617, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5916 + }, + { + "epoch": 2.3639632441070715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5917 + }, + { + "epoch": 2.3643627646823813, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5918 + }, + { + "epoch": 2.3647622852576906, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5919 + }, + { + "epoch": 2.3651618058330004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 5920 + }, + { + "epoch": 2.36556132640831, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5921 + }, + { + "epoch": 2.3659608469836195, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5922 + }, + { + "epoch": 2.3663603675589293, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5923 + }, + { + "epoch": 2.366759888134239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5924 + }, + { + "epoch": 2.3671594087095484, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5925 + }, + { + "epoch": 2.3675589292848582, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5926 + }, + { + "epoch": 2.3679584498601676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5927 + }, + { + "epoch": 2.3683579704354774, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5928 + }, + { + "epoch": 2.368757491010787, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5929 + }, + { + "epoch": 2.369157011586097, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5930 + }, + { + "epoch": 2.3695565321614063, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5931 + }, + { + "epoch": 2.369956052736716, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5932 + }, + { + "epoch": 2.3703555733120254, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5933 + }, + { + "epoch": 2.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5934 + }, + { + "epoch": 2.371154614462645, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5935 + }, + { + "epoch": 2.3715541350379543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5936 + }, + { + "epoch": 2.371953655613264, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5937 + }, + { + "epoch": 2.372353176188574, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5938 + }, + { + "epoch": 2.3727526967638832, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5939 + }, + { + "epoch": 2.373152217339193, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5940 + }, + { + "epoch": 2.373551737914503, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5941 + }, + { + "epoch": 2.373951258489812, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5942 + }, + { + "epoch": 2.374350779065122, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 5943 + }, + { + "epoch": 2.3747502996404313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5944 + }, + { + "epoch": 2.375149820215741, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5945 + }, + { + "epoch": 2.375549340791051, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5946 + }, + { + "epoch": 2.37594886136636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5947 + }, + { + "epoch": 2.37634838194167, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5948 + }, + { + "epoch": 2.3767479025169798, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5949 + }, + { + "epoch": 2.377147423092289, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 5950 + }, + { + "epoch": 2.377546943667599, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5951 + }, + { + "epoch": 2.3779464642429087, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5952 + }, + { + "epoch": 2.378345984818218, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 5953 + }, + { + "epoch": 2.378745505393528, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5954 + }, + { + "epoch": 2.379145025968837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5955 + }, + { + "epoch": 2.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5956 + }, + { + "epoch": 2.3799440671194567, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5957 + }, + { + "epoch": 2.3803435876947665, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5958 + }, + { + "epoch": 2.380743108270076, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5959 + }, + { + "epoch": 2.3811426288453856, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5960 + }, + { + "epoch": 2.381542149420695, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5961 + }, + { + "epoch": 2.3819416699960048, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5962 + }, + { + "epoch": 2.3823411905713145, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5963 + }, + { + "epoch": 2.382740711146624, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5964 + }, + { + "epoch": 2.3831402317219337, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5965 + }, + { + "epoch": 2.3835397522972435, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5966 + }, + { + "epoch": 2.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5967 + }, + { + "epoch": 2.3843387934478626, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5968 + }, + { + "epoch": 2.3847383140231724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5969 + }, + { + "epoch": 2.3851378345984817, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5970 + }, + { + "epoch": 2.3855373551737915, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 5971 + }, + { + "epoch": 2.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5972 + }, + { + "epoch": 2.3863363963244106, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 5973 + }, + { + "epoch": 2.3867359168997204, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5974 + }, + { + "epoch": 2.38713543747503, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5975 + }, + { + "epoch": 2.3875349580503396, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5976 + }, + { + "epoch": 2.3879344786256493, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5977 + }, + { + "epoch": 2.3883339992009587, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5978 + }, + { + "epoch": 2.3887335197762685, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5979 + }, + { + "epoch": 2.3891330403515783, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5980 + }, + { + "epoch": 2.3895325609268876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5981 + }, + { + "epoch": 2.3899320815021974, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5982 + }, + { + "epoch": 2.390331602077507, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5983 + }, + { + "epoch": 2.3907311226528165, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5984 + }, + { + "epoch": 2.3911306432281263, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5985 + }, + { + "epoch": 2.391530163803436, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 5986 + }, + { + "epoch": 2.3919296843787454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5987 + }, + { + "epoch": 2.392329204954055, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5988 + }, + { + "epoch": 2.3927287255293646, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5989 + }, + { + "epoch": 2.3931282461046743, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5990 + }, + { + "epoch": 2.393527766679984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5991 + }, + { + "epoch": 2.3939272872552935, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5992 + }, + { + "epoch": 2.3943268078306033, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5993 + }, + { + "epoch": 2.394726328405913, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5994 + }, + { + "epoch": 2.3951258489812224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5995 + }, + { + "epoch": 2.395525369556532, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5996 + }, + { + "epoch": 2.395924890131842, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5997 + }, + { + "epoch": 2.3963244107071513, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5998 + }, + { + "epoch": 2.396723931282461, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5999 + }, + { + "epoch": 2.3971234518577704, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6000 + }, + { + "epoch": 2.39752297243308, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6001 + }, + { + "epoch": 2.39792249300839, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6002 + }, + { + "epoch": 2.3983220135837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6003 + }, + { + "epoch": 2.398721534159009, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6004 + }, + { + "epoch": 2.399121054734319, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6005 + }, + { + "epoch": 2.3995205753096283, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6006 + }, + { + "epoch": 2.399920095884938, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6007 + }, + { + "epoch": 2.400319616460248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6008 + }, + { + "epoch": 2.400719137035557, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6009 + }, + { + "epoch": 2.401118657610867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6010 + }, + { + "epoch": 2.4015181781861767, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6011 + }, + { + "epoch": 2.401917698761486, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6012 + }, + { + "epoch": 2.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6013 + }, + { + "epoch": 2.4027167399121057, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6014 + }, + { + "epoch": 2.403116260487415, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6015 + }, + { + "epoch": 2.403515781062725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6016 + }, + { + "epoch": 2.403915301638034, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6017 + }, + { + "epoch": 2.404314822213344, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6018 + }, + { + "epoch": 2.4047143427886537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6019 + }, + { + "epoch": 2.4051138633639635, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6020 + }, + { + "epoch": 2.405513383939273, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6021 + }, + { + "epoch": 2.4059129045145826, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6022 + }, + { + "epoch": 2.406312425089892, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6023 + }, + { + "epoch": 2.4067119456652017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6024 + }, + { + "epoch": 2.4071114662405115, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6025 + }, + { + "epoch": 2.407510986815821, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6026 + }, + { + "epoch": 2.4079105073911307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6027 + }, + { + "epoch": 2.4083100279664404, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6028 + }, + { + "epoch": 2.40870954854175, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 6029 + }, + { + "epoch": 2.4091090691170596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6030 + }, + { + "epoch": 2.4095085896923694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6031 + }, + { + "epoch": 2.4099081102676787, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6032 + }, + { + "epoch": 2.4103076308429885, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 6033 + }, + { + "epoch": 2.410707151418298, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 6034 + }, + { + "epoch": 2.4111066719936076, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6035 + }, + { + "epoch": 2.4115061925689174, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6036 + }, + { + "epoch": 2.4119057131442267, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6037 + }, + { + "epoch": 2.4123052337195365, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6038 + }, + { + "epoch": 2.4127047542948463, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6039 + }, + { + "epoch": 2.4131042748701557, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6040 + }, + { + "epoch": 2.4135037954454654, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6041 + }, + { + "epoch": 2.4139033160207752, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6042 + }, + { + "epoch": 2.4143028365960846, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 6043 + }, + { + "epoch": 2.4147023571713944, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6044 + }, + { + "epoch": 2.4151018777467037, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6045 + }, + { + "epoch": 2.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6046 + }, + { + "epoch": 2.4159009188973233, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6047 + }, + { + "epoch": 2.416300439472633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 6048 + }, + { + "epoch": 2.4166999600479424, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6049 + }, + { + "epoch": 2.417099480623252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6050 + }, + { + "epoch": 2.4174990011985615, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6051 + }, + { + "epoch": 2.4178985217738713, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6052 + }, + { + "epoch": 2.418298042349181, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6053 + }, + { + "epoch": 2.4186975629244905, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6054 + }, + { + "epoch": 2.4190970834998002, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6055 + }, + { + "epoch": 2.41949660407511, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6056 + }, + { + "epoch": 2.4198961246504194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6057 + }, + { + "epoch": 2.420295645225729, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6058 + }, + { + "epoch": 2.420695165801039, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6059 + }, + { + "epoch": 2.4210946863763483, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6060 + }, + { + "epoch": 2.421494206951658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6061 + }, + { + "epoch": 2.4218937275269674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 6062 + }, + { + "epoch": 2.422293248102277, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6063 + }, + { + "epoch": 2.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6064 + }, + { + "epoch": 2.4230922892528968, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6065 + }, + { + "epoch": 2.423491809828206, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6066 + }, + { + "epoch": 2.423891330403516, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6067 + }, + { + "epoch": 2.4242908509788252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6068 + }, + { + "epoch": 2.424690371554135, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6069 + }, + { + "epoch": 2.425089892129445, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6070 + }, + { + "epoch": 2.425489412704754, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6071 + }, + { + "epoch": 2.425888933280064, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6072 + }, + { + "epoch": 2.4262884538553737, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6073 + }, + { + "epoch": 2.426687974430683, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6074 + }, + { + "epoch": 2.427087495005993, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6075 + }, + { + "epoch": 2.4274870155813026, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6076 + }, + { + "epoch": 2.427886536156612, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6077 + }, + { + "epoch": 2.4282860567319218, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6078 + }, + { + "epoch": 2.428685577307231, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6079 + }, + { + "epoch": 2.429085097882541, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6080 + }, + { + "epoch": 2.4294846184578507, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6081 + }, + { + "epoch": 2.42988413903316, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6082 + }, + { + "epoch": 2.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6083 + }, + { + "epoch": 2.4306831801837796, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6084 + }, + { + "epoch": 2.431082700759089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6085 + }, + { + "epoch": 2.4314822213343987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6086 + }, + { + "epoch": 2.4318817419097085, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6087 + }, + { + "epoch": 2.432281262485018, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6088 + }, + { + "epoch": 2.4326807830603276, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6089 + }, + { + "epoch": 2.433080303635637, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6090 + }, + { + "epoch": 2.4334798242109468, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6091 + }, + { + "epoch": 2.4338793447862566, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 6092 + }, + { + "epoch": 2.4342788653615663, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6093 + }, + { + "epoch": 2.4346783859368757, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6094 + }, + { + "epoch": 2.4350779065121855, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 6095 + }, + { + "epoch": 2.435477427087495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 6096 + }, + { + "epoch": 2.4358769476628046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 6097 + }, + { + "epoch": 2.4362764682381144, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6098 + }, + { + "epoch": 2.4366759888134237, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6099 + }, + { + "epoch": 2.4370755093887335, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6100 + }, + { + "epoch": 2.4374750299640433, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6101 + }, + { + "epoch": 2.4378745505393526, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6102 + }, + { + "epoch": 2.4382740711146624, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6103 + }, + { + "epoch": 2.438673591689972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6104 + }, + { + "epoch": 2.4390731122652816, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6105 + }, + { + "epoch": 2.4394726328405913, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6106 + }, + { + "epoch": 2.4398721534159007, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6107 + }, + { + "epoch": 2.4402716739912105, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6108 + }, + { + "epoch": 2.4406711945665203, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6109 + }, + { + "epoch": 2.44107071514183, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6110 + }, + { + "epoch": 2.4414702357171394, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6111 + }, + { + "epoch": 2.441869756292449, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6112 + }, + { + "epoch": 2.4422692768677585, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6113 + }, + { + "epoch": 2.4426687974430683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6114 + }, + { + "epoch": 2.443068318018378, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6115 + }, + { + "epoch": 2.4434678385936874, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6116 + }, + { + "epoch": 2.443867359168997, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6117 + }, + { + "epoch": 2.444266879744307, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 6118 + }, + { + "epoch": 2.4446664003196164, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6119 + }, + { + "epoch": 2.445065920894926, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6120 + }, + { + "epoch": 2.445465441470236, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6121 + }, + { + "epoch": 2.4458649620455453, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6122 + }, + { + "epoch": 2.446264482620855, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6123 + }, + { + "epoch": 2.4466640031961644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6124 + }, + { + "epoch": 2.447063523771474, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6125 + }, + { + "epoch": 2.447463044346784, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6126 + }, + { + "epoch": 2.4478625649220933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6127 + }, + { + "epoch": 2.448262085497403, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6128 + }, + { + "epoch": 2.448661606072713, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6129 + }, + { + "epoch": 2.4490611266480222, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6130 + }, + { + "epoch": 2.449460647223332, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6131 + }, + { + "epoch": 2.449860167798642, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6132 + }, + { + "epoch": 2.450259688373951, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6133 + }, + { + "epoch": 2.450659208949261, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6134 + }, + { + "epoch": 2.4510587295245703, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 6135 + }, + { + "epoch": 2.45145825009988, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6136 + }, + { + "epoch": 2.45185777067519, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6137 + }, + { + "epoch": 2.4522572912504996, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6138 + }, + { + "epoch": 2.452656811825809, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6139 + }, + { + "epoch": 2.4530563324011188, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6140 + }, + { + "epoch": 2.453455852976428, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6141 + }, + { + "epoch": 2.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6142 + }, + { + "epoch": 2.4542548941270477, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6143 + }, + { + "epoch": 2.454654414702357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6144 + }, + { + "epoch": 2.455053935277667, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6145 + }, + { + "epoch": 2.4554534558529766, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6146 + }, + { + "epoch": 2.455852976428286, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6147 + }, + { + "epoch": 2.4562524970035957, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6148 + }, + { + "epoch": 2.4566520175789055, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6149 + }, + { + "epoch": 2.457051538154215, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6150 + }, + { + "epoch": 2.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6151 + }, + { + "epoch": 2.457850579304834, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6152 + }, + { + "epoch": 2.4582500998801438, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6153 + }, + { + "epoch": 2.4586496204554535, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6154 + }, + { + "epoch": 2.4590491410307633, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6155 + }, + { + "epoch": 2.4594486616060727, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6156 + }, + { + "epoch": 2.4598481821813825, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6157 + }, + { + "epoch": 2.460247702756692, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6158 + }, + { + "epoch": 2.4606472233320016, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6159 + }, + { + "epoch": 2.4610467439073114, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6160 + }, + { + "epoch": 2.4614462644826207, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6161 + }, + { + "epoch": 2.4618457850579305, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6162 + }, + { + "epoch": 2.4622453056332403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6163 + }, + { + "epoch": 2.4626448262085496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6164 + }, + { + "epoch": 2.4630443467838594, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6165 + }, + { + "epoch": 2.463443867359169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6166 + }, + { + "epoch": 2.4638433879344785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6167 + }, + { + "epoch": 2.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6168 + }, + { + "epoch": 2.4646424290850977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 6169 + }, + { + "epoch": 2.4650419496604075, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6170 + }, + { + "epoch": 2.4654414702357172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6171 + }, + { + "epoch": 2.4658409908110266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6172 + }, + { + "epoch": 2.4662405113863364, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6173 + }, + { + "epoch": 2.466640031961646, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6174 + }, + { + "epoch": 2.4670395525369555, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6175 + }, + { + "epoch": 2.4674390731122653, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6176 + }, + { + "epoch": 2.467838593687575, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6177 + }, + { + "epoch": 2.4682381142628844, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6178 + }, + { + "epoch": 2.468637634838194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6179 + }, + { + "epoch": 2.4690371554135035, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6180 + }, + { + "epoch": 2.4694366759888133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6181 + }, + { + "epoch": 2.469836196564123, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6182 + }, + { + "epoch": 2.470235717139433, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6183 + }, + { + "epoch": 2.4706352377147422, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6184 + }, + { + "epoch": 2.471034758290052, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6185 + }, + { + "epoch": 2.4714342788653614, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 6186 + }, + { + "epoch": 2.471833799440671, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6187 + }, + { + "epoch": 2.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6188 + }, + { + "epoch": 2.4726328405912903, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6189 + }, + { + "epoch": 2.4730323611666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6190 + }, + { + "epoch": 2.47343188174191, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6191 + }, + { + "epoch": 2.473831402317219, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6192 + }, + { + "epoch": 2.474230922892529, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6193 + }, + { + "epoch": 2.474630443467839, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6194 + }, + { + "epoch": 2.475029964043148, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6195 + }, + { + "epoch": 2.475429484618458, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6196 + }, + { + "epoch": 2.4758290051937673, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6197 + }, + { + "epoch": 2.476228525769077, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6198 + }, + { + "epoch": 2.476628046344387, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6199 + }, + { + "epoch": 2.4770275669196966, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6200 + }, + { + "epoch": 2.477427087495006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6201 + }, + { + "epoch": 2.4778266080703157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6202 + }, + { + "epoch": 2.478226128645625, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6203 + }, + { + "epoch": 2.478625649220935, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6204 + }, + { + "epoch": 2.4790251697962447, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6205 + }, + { + "epoch": 2.479424690371554, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6206 + }, + { + "epoch": 2.479824210946864, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6207 + }, + { + "epoch": 2.4802237315221736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6208 + }, + { + "epoch": 2.480623252097483, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6209 + }, + { + "epoch": 2.4810227726727927, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6210 + }, + { + "epoch": 2.4814222932481025, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6211 + }, + { + "epoch": 2.481821813823412, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6212 + }, + { + "epoch": 2.4822213343987216, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6213 + }, + { + "epoch": 2.482620854974031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 6214 + }, + { + "epoch": 2.4830203755493407, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6215 + }, + { + "epoch": 2.4834198961246505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6216 + }, + { + "epoch": 2.48381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 6217 + }, + { + "epoch": 2.4842189372752697, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6218 + }, + { + "epoch": 2.4846184578505794, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6219 + }, + { + "epoch": 2.485017978425889, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6220 + }, + { + "epoch": 2.4854174990011986, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6221 + }, + { + "epoch": 2.4858170195765084, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6222 + }, + { + "epoch": 2.4862165401518177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6223 + }, + { + "epoch": 2.4866160607271275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6224 + }, + { + "epoch": 2.487015581302437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6225 + }, + { + "epoch": 2.4874151018777466, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6226 + }, + { + "epoch": 2.4878146224530564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6227 + }, + { + "epoch": 2.488214143028366, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6228 + }, + { + "epoch": 2.4886136636036755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6229 + }, + { + "epoch": 2.4890131841789853, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6230 + }, + { + "epoch": 2.4894127047542947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6231 + }, + { + "epoch": 2.4898122253296044, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6232 + }, + { + "epoch": 2.4902117459049142, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6233 + }, + { + "epoch": 2.4906112664802236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6234 + }, + { + "epoch": 2.4910107870555334, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6235 + }, + { + "epoch": 2.491410307630843, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6236 + }, + { + "epoch": 2.4918098282061525, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6237 + }, + { + "epoch": 2.4922093487814623, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6238 + }, + { + "epoch": 2.492608869356772, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6239 + }, + { + "epoch": 2.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6240 + }, + { + "epoch": 2.493407910507391, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6241 + }, + { + "epoch": 2.4938074310827005, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6242 + }, + { + "epoch": 2.4942069516580103, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6243 + }, + { + "epoch": 2.49460647223332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6244 + }, + { + "epoch": 2.49500599280863, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6245 + }, + { + "epoch": 2.4954055133839392, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 6246 + }, + { + "epoch": 2.495805033959249, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6247 + }, + { + "epoch": 2.4962045545345584, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6248 + }, + { + "epoch": 2.496604075109868, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6249 + }, + { + "epoch": 2.497003595685178, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6250 + }, + { + "epoch": 2.4974031162604873, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6251 + }, + { + "epoch": 2.497802636835797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6252 + }, + { + "epoch": 2.498202157411107, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6253 + }, + { + "epoch": 2.498601677986416, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6254 + }, + { + "epoch": 2.499001198561726, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 6255 + }, + { + "epoch": 2.4994007191370358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6256 + }, + { + "epoch": 2.499800239712345, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6257 + }, + { + "epoch": 2.500199760287655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 6258 + }, + { + "epoch": 2.5005992808629642, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6259 + }, + { + "epoch": 2.500998801438274, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6260 + }, + { + "epoch": 2.501398322013584, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6261 + }, + { + "epoch": 2.5017978425888936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6262 + }, + { + "epoch": 2.502197363164203, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6263 + }, + { + "epoch": 2.5025968837395127, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6264 + }, + { + "epoch": 2.502996404314822, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6265 + }, + { + "epoch": 2.503395924890132, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6266 + }, + { + "epoch": 2.5037954454654416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6267 + }, + { + "epoch": 2.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6268 + }, + { + "epoch": 2.5045944866160608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6269 + }, + { + "epoch": 2.50499400719137, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6270 + }, + { + "epoch": 2.50539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6271 + }, + { + "epoch": 2.5057930483419897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 6272 + }, + { + "epoch": 2.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6273 + }, + { + "epoch": 2.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6274 + }, + { + "epoch": 2.5069916100679186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6275 + }, + { + "epoch": 2.507391130643228, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 6276 + }, + { + "epoch": 2.5077906512185377, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6277 + }, + { + "epoch": 2.5081901717938475, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6278 + }, + { + "epoch": 2.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6279 + }, + { + "epoch": 2.5089892129444666, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6280 + }, + { + "epoch": 2.5093887335197764, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6281 + }, + { + "epoch": 2.5097882540950858, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6282 + }, + { + "epoch": 2.5101877746703956, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6283 + }, + { + "epoch": 2.5105872952457053, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6284 + }, + { + "epoch": 2.5109868158210147, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6285 + }, + { + "epoch": 2.5113863363963245, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6286 + }, + { + "epoch": 2.511785856971634, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6287 + }, + { + "epoch": 2.5121853775469436, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6288 + }, + { + "epoch": 2.5125848981222534, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6289 + }, + { + "epoch": 2.512984418697563, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6290 + }, + { + "epoch": 2.5133839392728725, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6291 + }, + { + "epoch": 2.5137834598481823, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6292 + }, + { + "epoch": 2.5141829804234916, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6293 + }, + { + "epoch": 2.5145825009988014, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6294 + }, + { + "epoch": 2.514982021574111, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6295 + }, + { + "epoch": 2.5153815421494206, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6296 + }, + { + "epoch": 2.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6297 + }, + { + "epoch": 2.5161805833000397, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6298 + }, + { + "epoch": 2.5165801038753495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6299 + }, + { + "epoch": 2.5169796244506593, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6300 + }, + { + "epoch": 2.517379145025969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6301 + }, + { + "epoch": 2.5177786656012784, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6302 + }, + { + "epoch": 2.518178186176588, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6303 + }, + { + "epoch": 2.5185777067518975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6304 + }, + { + "epoch": 2.5189772273272073, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6305 + }, + { + "epoch": 2.519376747902517, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6306 + }, + { + "epoch": 2.519776268477827, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6307 + }, + { + "epoch": 2.520175789053136, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6308 + }, + { + "epoch": 2.520575309628446, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6309 + }, + { + "epoch": 2.5209748302037553, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6310 + }, + { + "epoch": 2.521374350779065, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 6311 + }, + { + "epoch": 2.521773871354375, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6312 + }, + { + "epoch": 2.5221733919296843, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 6313 + }, + { + "epoch": 2.522572912504994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6314 + }, + { + "epoch": 2.5229724330803034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6315 + }, + { + "epoch": 2.523371953655613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6316 + }, + { + "epoch": 2.523771474230923, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6317 + }, + { + "epoch": 2.5241709948062327, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6318 + }, + { + "epoch": 2.524570515381542, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6319 + }, + { + "epoch": 2.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6320 + }, + { + "epoch": 2.525369556532161, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6321 + }, + { + "epoch": 2.525769077107471, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6322 + }, + { + "epoch": 2.526168597682781, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 6323 + }, + { + "epoch": 2.52656811825809, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6324 + }, + { + "epoch": 2.5269676388334, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6325 + }, + { + "epoch": 2.5273671594087097, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6326 + }, + { + "epoch": 2.527766679984019, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 6327 + }, + { + "epoch": 2.528166200559329, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6328 + }, + { + "epoch": 2.5285657211346386, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6329 + }, + { + "epoch": 2.528965241709948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6330 + }, + { + "epoch": 2.5293647622852578, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6331 + }, + { + "epoch": 2.529764282860567, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6332 + }, + { + "epoch": 2.530163803435877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6333 + }, + { + "epoch": 2.5305633240111867, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 6334 + }, + { + "epoch": 2.5309628445864965, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6335 + }, + { + "epoch": 2.531362365161806, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6336 + }, + { + "epoch": 2.5317618857371156, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6337 + }, + { + "epoch": 2.532161406312425, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6338 + }, + { + "epoch": 2.5325609268877347, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6339 + }, + { + "epoch": 2.5329604474630445, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6340 + }, + { + "epoch": 2.533359968038354, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6341 + }, + { + "epoch": 2.5337594886136636, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6342 + }, + { + "epoch": 2.534159009188973, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6343 + }, + { + "epoch": 2.5345585297642828, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6344 + }, + { + "epoch": 2.5349580503395925, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6345 + }, + { + "epoch": 2.5353575709149023, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6346 + }, + { + "epoch": 2.5357570914902117, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 6347 + }, + { + "epoch": 2.5361566120655215, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6348 + }, + { + "epoch": 2.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6349 + }, + { + "epoch": 2.5369556532161406, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 6350 + }, + { + "epoch": 2.5373551737914504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6351 + }, + { + "epoch": 2.53775469436676, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6352 + }, + { + "epoch": 2.5381542149420695, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6353 + }, + { + "epoch": 2.5385537355173793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6354 + }, + { + "epoch": 2.5389532560926886, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6355 + }, + { + "epoch": 2.5393527766679984, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6356 + }, + { + "epoch": 2.539752297243308, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6357 + }, + { + "epoch": 2.5401518178186175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6358 + }, + { + "epoch": 2.5405513383939273, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 6359 + }, + { + "epoch": 2.5409508589692367, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6360 + }, + { + "epoch": 2.5413503795445465, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6361 + }, + { + "epoch": 2.5417499001198562, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 6362 + }, + { + "epoch": 2.542149420695166, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6363 + }, + { + "epoch": 2.5425489412704754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6364 + }, + { + "epoch": 2.542948461845785, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6365 + }, + { + "epoch": 2.5433479824210945, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 6366 + }, + { + "epoch": 2.5437475029964043, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6367 + }, + { + "epoch": 2.544147023571714, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6368 + }, + { + "epoch": 2.5445465441470234, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6369 + }, + { + "epoch": 2.544946064722333, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6370 + }, + { + "epoch": 2.545345585297643, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6371 + }, + { + "epoch": 2.5457451058729523, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6372 + }, + { + "epoch": 2.546144626448262, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6373 + }, + { + "epoch": 2.546544147023572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6374 + }, + { + "epoch": 2.5469436675988812, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6375 + }, + { + "epoch": 2.547343188174191, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6376 + }, + { + "epoch": 2.5477427087495004, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6377 + }, + { + "epoch": 2.54814222932481, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6378 + }, + { + "epoch": 2.54854174990012, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6379 + }, + { + "epoch": 2.5489412704754297, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6380 + }, + { + "epoch": 2.549340791050739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6381 + }, + { + "epoch": 2.549740311626049, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6382 + }, + { + "epoch": 2.550139832201358, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6383 + }, + { + "epoch": 2.550539352776668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6384 + }, + { + "epoch": 2.5509388733519778, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6385 + }, + { + "epoch": 2.551338393927287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 6386 + }, + { + "epoch": 2.551737914502597, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6387 + }, + { + "epoch": 2.5521374350779062, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 6388 + }, + { + "epoch": 2.552536955653216, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6389 + }, + { + "epoch": 2.552936476228526, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6390 + }, + { + "epoch": 2.5533359968038356, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6391 + }, + { + "epoch": 2.553735517379145, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6392 + }, + { + "epoch": 2.5541350379544547, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6393 + }, + { + "epoch": 2.554534558529764, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6394 + }, + { + "epoch": 2.554934079105074, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 6395 + }, + { + "epoch": 2.5553335996803836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6396 + }, + { + "epoch": 2.5557331202556934, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6397 + }, + { + "epoch": 2.556132640831003, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6398 + }, + { + "epoch": 2.5565321614063126, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6399 + }, + { + "epoch": 2.556931681981622, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6400 + }, + { + "epoch": 2.5573312025569317, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6401 + }, + { + "epoch": 2.5577307231322415, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6402 + }, + { + "epoch": 2.558130243707551, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6403 + }, + { + "epoch": 2.5585297642828606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6404 + }, + { + "epoch": 2.55892928485817, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 6405 + }, + { + "epoch": 2.5593288054334797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 6406 + }, + { + "epoch": 2.5597283260087895, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 6407 + }, + { + "epoch": 2.5601278465840993, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6408 + }, + { + "epoch": 2.5605273671594087, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6409 + }, + { + "epoch": 2.5609268877347184, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6410 + }, + { + "epoch": 2.561326408310028, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6411 + }, + { + "epoch": 2.5617259288853376, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 6412 + }, + { + "epoch": 2.5621254494606474, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6413 + }, + { + "epoch": 2.5625249700359567, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6414 + }, + { + "epoch": 2.5629244906112665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6415 + }, + { + "epoch": 2.5633240111865763, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6416 + }, + { + "epoch": 2.5637235317618856, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 6417 + }, + { + "epoch": 2.5641230523371954, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6418 + }, + { + "epoch": 2.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 6419 + }, + { + "epoch": 2.5649220934878145, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6420 + }, + { + "epoch": 2.5653216140631243, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6421 + }, + { + "epoch": 2.5657211346384337, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6422 + }, + { + "epoch": 2.5661206552137434, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6423 + }, + { + "epoch": 2.5665201757890532, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6424 + }, + { + "epoch": 2.566919696364363, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6425 + }, + { + "epoch": 2.5673192169396724, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6426 + }, + { + "epoch": 2.567718737514982, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6427 + }, + { + "epoch": 2.5681182580902915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6428 + }, + { + "epoch": 2.5685177786656013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6429 + }, + { + "epoch": 2.568917299240911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6430 + }, + { + "epoch": 2.5693168198162204, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6431 + }, + { + "epoch": 2.56971634039153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6432 + }, + { + "epoch": 2.5701158609668395, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6433 + }, + { + "epoch": 2.5705153815421493, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 6434 + }, + { + "epoch": 2.570914902117459, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6435 + }, + { + "epoch": 2.571314422692769, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6436 + }, + { + "epoch": 2.5717139432680782, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6437 + }, + { + "epoch": 2.572113463843388, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6438 + }, + { + "epoch": 2.5725129844186974, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6439 + }, + { + "epoch": 2.572912504994007, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6440 + }, + { + "epoch": 2.573312025569317, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6441 + }, + { + "epoch": 2.5737115461446267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6442 + }, + { + "epoch": 2.574111066719936, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6443 + }, + { + "epoch": 2.574510587295246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 6444 + }, + { + "epoch": 2.574910107870555, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6445 + }, + { + "epoch": 2.575309628445865, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6446 + }, + { + "epoch": 2.5757091490211748, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6447 + }, + { + "epoch": 2.576108669596484, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6448 + }, + { + "epoch": 2.576508190171794, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6449 + }, + { + "epoch": 2.5769077107471032, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6450 + }, + { + "epoch": 2.577307231322413, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6451 + }, + { + "epoch": 2.577706751897723, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6452 + }, + { + "epoch": 2.5781062724730326, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6453 + }, + { + "epoch": 2.578505793048342, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6454 + }, + { + "epoch": 2.5789053136236517, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6455 + }, + { + "epoch": 2.579304834198961, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6456 + }, + { + "epoch": 2.579704354774271, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 6457 + }, + { + "epoch": 2.5801038753495806, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6458 + }, + { + "epoch": 2.58050339592489, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6459 + }, + { + "epoch": 2.5809029165001998, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6460 + }, + { + "epoch": 2.5813024370755095, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6461 + }, + { + "epoch": 2.581701957650819, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6462 + }, + { + "epoch": 2.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6463 + }, + { + "epoch": 2.5825009988014385, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6464 + }, + { + "epoch": 2.582900519376748, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6465 + }, + { + "epoch": 2.5833000399520576, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6466 + }, + { + "epoch": 2.583699560527367, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6467 + }, + { + "epoch": 2.5840990811026767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6468 + }, + { + "epoch": 2.5844986016779865, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6469 + }, + { + "epoch": 2.5848981222532963, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6470 + }, + { + "epoch": 2.5852976428286056, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6471 + }, + { + "epoch": 2.5856971634039154, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6472 + }, + { + "epoch": 2.5860966839792248, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6473 + }, + { + "epoch": 2.5864962045545346, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6474 + }, + { + "epoch": 2.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 6475 + }, + { + "epoch": 2.5872952457051537, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6476 + }, + { + "epoch": 2.5876947662804635, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6477 + }, + { + "epoch": 2.588094286855773, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6478 + }, + { + "epoch": 2.5884938074310826, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6479 + }, + { + "epoch": 2.5888933280063924, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6480 + }, + { + "epoch": 2.589292848581702, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6481 + }, + { + "epoch": 2.5896923691570115, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6482 + }, + { + "epoch": 2.5900918897323213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6483 + }, + { + "epoch": 2.5904914103076306, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 6484 + }, + { + "epoch": 2.5908909308829404, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6485 + }, + { + "epoch": 2.59129045145825, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6486 + }, + { + "epoch": 2.59168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6487 + }, + { + "epoch": 2.5920894926088693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6488 + }, + { + "epoch": 2.592489013184179, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6489 + }, + { + "epoch": 2.5928885337594885, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6490 + }, + { + "epoch": 2.5932880543347983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6491 + }, + { + "epoch": 2.593687574910108, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6492 + }, + { + "epoch": 2.5940870954854174, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 6493 + }, + { + "epoch": 2.594486616060727, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6494 + }, + { + "epoch": 2.5948861366360365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6495 + }, + { + "epoch": 2.5952856572113463, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6496 + }, + { + "epoch": 2.595685177786656, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6497 + }, + { + "epoch": 2.596084698361966, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6498 + }, + { + "epoch": 2.596484218937275, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6499 + }, + { + "epoch": 2.596883739512585, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6500 + }, + { + "epoch": 2.5972832600878943, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6501 + }, + { + "epoch": 2.597682780663204, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6502 + }, + { + "epoch": 2.598082301238514, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 6503 + }, + { + "epoch": 2.5984818218138233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6504 + }, + { + "epoch": 2.598881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6505 + }, + { + "epoch": 2.599280862964443, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6506 + }, + { + "epoch": 2.599680383539752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6507 + }, + { + "epoch": 2.600079904115062, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 6508 + }, + { + "epoch": 2.6004794246903717, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6509 + }, + { + "epoch": 2.600878945265681, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6510 + }, + { + "epoch": 2.601278465840991, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6511 + }, + { + "epoch": 2.6016779864163, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6512 + }, + { + "epoch": 2.60207750699161, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6513 + }, + { + "epoch": 2.60247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6514 + }, + { + "epoch": 2.6028765481422296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6515 + }, + { + "epoch": 2.603276068717539, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6516 + }, + { + "epoch": 2.6036755892928487, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6517 + }, + { + "epoch": 2.604075109868158, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6518 + }, + { + "epoch": 2.604474630443468, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6519 + }, + { + "epoch": 2.6048741510187776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6520 + }, + { + "epoch": 2.605273671594087, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6521 + }, + { + "epoch": 2.6056731921693967, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6522 + }, + { + "epoch": 2.606072712744706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 6523 + }, + { + "epoch": 2.606472233320016, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6524 + }, + { + "epoch": 2.6068717538953257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6525 + }, + { + "epoch": 2.6072712744706354, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6526 + }, + { + "epoch": 2.607670795045945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6527 + }, + { + "epoch": 2.6080703156212546, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6528 + }, + { + "epoch": 2.608469836196564, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6529 + }, + { + "epoch": 2.6088693567718737, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 6530 + }, + { + "epoch": 2.6092688773471835, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6531 + }, + { + "epoch": 2.6096683979224933, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6532 + }, + { + "epoch": 2.6100679184978026, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6533 + }, + { + "epoch": 2.6104674390731124, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6534 + }, + { + "epoch": 2.6108669596484217, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6535 + }, + { + "epoch": 2.6112664802237315, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6536 + }, + { + "epoch": 2.6116660007990413, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6537 + }, + { + "epoch": 2.6120655213743507, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6538 + }, + { + "epoch": 2.6124650419496604, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6539 + }, + { + "epoch": 2.61286456252497, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6540 + }, + { + "epoch": 2.6132640831002796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 6541 + }, + { + "epoch": 2.6136636036755894, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6542 + }, + { + "epoch": 2.614063124250899, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6543 + }, + { + "epoch": 2.6144626448262085, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6544 + }, + { + "epoch": 2.6148621654015183, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6545 + }, + { + "epoch": 2.6152616859768276, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6546 + }, + { + "epoch": 2.6156612065521374, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6547 + }, + { + "epoch": 2.616060727127447, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 6548 + }, + { + "epoch": 2.6164602477027565, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6549 + }, + { + "epoch": 2.6168597682780663, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6550 + }, + { + "epoch": 2.617259288853376, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6551 + }, + { + "epoch": 2.6176588094286855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6552 + }, + { + "epoch": 2.6180583300039952, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6553 + }, + { + "epoch": 2.618457850579305, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6554 + }, + { + "epoch": 2.6188573711546144, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6555 + }, + { + "epoch": 2.619256891729924, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6556 + }, + { + "epoch": 2.6196564123052335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6557 + }, + { + "epoch": 2.6200559328805433, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6558 + }, + { + "epoch": 2.620455453455853, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 6559 + }, + { + "epoch": 2.620854974031163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6560 + }, + { + "epoch": 2.621254494606472, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6561 + }, + { + "epoch": 2.621654015181782, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6562 + }, + { + "epoch": 2.6220535357570913, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6563 + }, + { + "epoch": 2.622453056332401, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 6564 + }, + { + "epoch": 2.622852576907711, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6565 + }, + { + "epoch": 2.6232520974830202, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6566 + }, + { + "epoch": 2.62365161805833, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6567 + }, + { + "epoch": 2.6240511386336394, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6568 + }, + { + "epoch": 2.624450659208949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6569 + }, + { + "epoch": 2.624850179784259, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6570 + }, + { + "epoch": 2.6252497003595687, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6571 + }, + { + "epoch": 2.625649220934878, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6572 + }, + { + "epoch": 2.626048741510188, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6573 + }, + { + "epoch": 2.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6574 + }, + { + "epoch": 2.626847782660807, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6575 + }, + { + "epoch": 2.6272473032361168, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6576 + }, + { + "epoch": 2.6276468238114266, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 6577 + }, + { + "epoch": 2.628046344386736, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6578 + }, + { + "epoch": 2.6284458649620457, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6579 + }, + { + "epoch": 2.628845385537355, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6580 + }, + { + "epoch": 2.629244906112665, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6581 + }, + { + "epoch": 2.6296444266879746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6582 + }, + { + "epoch": 2.630043947263284, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6583 + }, + { + "epoch": 2.6304434678385937, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6584 + }, + { + "epoch": 2.630842988413903, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 6585 + }, + { + "epoch": 2.631242508989213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6586 + }, + { + "epoch": 2.6316420295645226, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6587 + }, + { + "epoch": 2.6320415501398324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6588 + }, + { + "epoch": 2.6324410707151418, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6589 + }, + { + "epoch": 2.6328405912904516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6590 + }, + { + "epoch": 2.633240111865761, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6591 + }, + { + "epoch": 2.6336396324410707, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6592 + }, + { + "epoch": 2.6340391530163805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6593 + }, + { + "epoch": 2.63443867359169, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6594 + }, + { + "epoch": 2.6348381941669996, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6595 + }, + { + "epoch": 2.6352377147423094, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6596 + }, + { + "epoch": 2.6356372353176187, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6597 + }, + { + "epoch": 2.6360367558929285, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6598 + }, + { + "epoch": 2.6364362764682383, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6599 + }, + { + "epoch": 2.6368357970435476, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6600 + }, + { + "epoch": 2.6372353176188574, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6601 + }, + { + "epoch": 2.6376348381941668, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6602 + }, + { + "epoch": 2.6380343587694766, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6603 + }, + { + "epoch": 2.6384338793447863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6604 + }, + { + "epoch": 2.638833399920096, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6605 + }, + { + "epoch": 2.6392329204954055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6606 + }, + { + "epoch": 2.6396324410707153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6607 + }, + { + "epoch": 2.6400319616460246, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6608 + }, + { + "epoch": 2.6404314822213344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6609 + }, + { + "epoch": 2.640831002796644, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6610 + }, + { + "epoch": 2.6412305233719535, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6611 + }, + { + "epoch": 2.6416300439472633, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6612 + }, + { + "epoch": 2.6420295645225726, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6613 + }, + { + "epoch": 2.6424290850978824, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6614 + }, + { + "epoch": 2.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6615 + }, + { + "epoch": 2.643228126248502, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6616 + }, + { + "epoch": 2.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6617 + }, + { + "epoch": 2.644027167399121, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6618 + }, + { + "epoch": 2.6444266879744305, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6619 + }, + { + "epoch": 2.6448262085497403, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6620 + }, + { + "epoch": 2.64522572912505, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6621 + }, + { + "epoch": 2.64562524970036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6622 + }, + { + "epoch": 2.646024770275669, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6623 + }, + { + "epoch": 2.646424290850979, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6624 + }, + { + "epoch": 2.6468238114262883, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6625 + }, + { + "epoch": 2.647223332001598, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6626 + }, + { + "epoch": 2.647622852576908, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6627 + }, + { + "epoch": 2.6480223731522172, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6628 + }, + { + "epoch": 2.648421893727527, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6629 + }, + { + "epoch": 2.6488214143028364, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6630 + }, + { + "epoch": 2.649220934878146, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6631 + }, + { + "epoch": 2.649620455453456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 6632 + }, + { + "epoch": 2.6500199760287657, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6633 + }, + { + "epoch": 2.650419496604075, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6634 + }, + { + "epoch": 2.650819017179385, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6635 + }, + { + "epoch": 2.651218537754694, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6636 + }, + { + "epoch": 2.651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6637 + }, + { + "epoch": 2.6520175789053138, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6638 + }, + { + "epoch": 2.652417099480623, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 6639 + }, + { + "epoch": 2.652816620055933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6640 + }, + { + "epoch": 2.6532161406312427, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6641 + }, + { + "epoch": 2.653615661206552, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6642 + }, + { + "epoch": 2.654015181781862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6643 + }, + { + "epoch": 2.6544147023571716, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6644 + }, + { + "epoch": 2.654814222932481, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6645 + }, + { + "epoch": 2.6552137435077907, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6646 + }, + { + "epoch": 2.6556132640831, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 6647 + }, + { + "epoch": 2.65601278465841, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6648 + }, + { + "epoch": 2.6564123052337196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6649 + }, + { + "epoch": 2.6568118258090294, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6650 + }, + { + "epoch": 2.6572113463843388, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6651 + }, + { + "epoch": 2.6576108669596485, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6652 + }, + { + "epoch": 2.658010387534958, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6653 + }, + { + "epoch": 2.6584099081102677, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6654 + }, + { + "epoch": 2.6588094286855775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6655 + }, + { + "epoch": 2.659208949260887, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6656 + }, + { + "epoch": 2.6596084698361966, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6657 + }, + { + "epoch": 2.660007990411506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6658 + }, + { + "epoch": 2.6604075109868157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6659 + }, + { + "epoch": 2.6608070315621255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6660 + }, + { + "epoch": 2.6612065521374353, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6661 + }, + { + "epoch": 2.6616060727127446, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6662 + }, + { + "epoch": 2.6620055932880544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6663 + }, + { + "epoch": 2.6624051138633638, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6664 + }, + { + "epoch": 2.6628046344386735, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6665 + }, + { + "epoch": 2.6632041550139833, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 6666 + }, + { + "epoch": 2.663603675589293, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6667 + }, + { + "epoch": 2.6640031961646025, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6668 + }, + { + "epoch": 2.6644027167399122, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6669 + }, + { + "epoch": 2.6648022373152216, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6670 + }, + { + "epoch": 2.6652017578905314, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6671 + }, + { + "epoch": 2.665601278465841, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6672 + }, + { + "epoch": 2.6660007990411505, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6673 + }, + { + "epoch": 2.6664003196164603, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6674 + }, + { + "epoch": 2.6667998401917696, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6675 + }, + { + "epoch": 2.6671993607670794, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6676 + }, + { + "epoch": 2.667598881342389, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6677 + }, + { + "epoch": 2.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6678 + }, + { + "epoch": 2.6683979224930083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6679 + }, + { + "epoch": 2.668797443068318, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6680 + }, + { + "epoch": 2.6691969636436275, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6681 + }, + { + "epoch": 2.6695964842189372, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6682 + }, + { + "epoch": 2.669996004794247, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6683 + }, + { + "epoch": 2.6703955253695564, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6684 + }, + { + "epoch": 2.670795045944866, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6685 + }, + { + "epoch": 2.671194566520176, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6686 + }, + { + "epoch": 2.6715940870954853, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 6687 + }, + { + "epoch": 2.671993607670795, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6688 + }, + { + "epoch": 2.672393128246105, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6689 + }, + { + "epoch": 2.672792648821414, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6690 + }, + { + "epoch": 2.673192169396724, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6691 + }, + { + "epoch": 2.6735916899720333, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6692 + }, + { + "epoch": 2.673991210547343, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6693 + }, + { + "epoch": 2.674390731122653, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6694 + }, + { + "epoch": 2.6747902516979627, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6695 + }, + { + "epoch": 2.675189772273272, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6696 + }, + { + "epoch": 2.675589292848582, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6697 + }, + { + "epoch": 2.675988813423891, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6698 + }, + { + "epoch": 2.676388333999201, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6699 + }, + { + "epoch": 2.6767878545745107, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6700 + }, + { + "epoch": 2.67718737514982, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6701 + }, + { + "epoch": 2.67758689572513, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6702 + }, + { + "epoch": 2.677986416300439, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 6703 + }, + { + "epoch": 2.678385936875749, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6704 + }, + { + "epoch": 2.678785457451059, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6705 + }, + { + "epoch": 2.6791849780263686, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6706 + }, + { + "epoch": 2.679584498601678, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6707 + }, + { + "epoch": 2.6799840191769877, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6708 + }, + { + "epoch": 2.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6709 + }, + { + "epoch": 2.680783060327607, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6710 + }, + { + "epoch": 2.6811825809029166, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6711 + }, + { + "epoch": 2.6815821014782264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6712 + }, + { + "epoch": 2.6819816220535357, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 6713 + }, + { + "epoch": 2.6823811426288455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6714 + }, + { + "epoch": 2.682780663204155, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6715 + }, + { + "epoch": 2.6831801837794647, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6716 + }, + { + "epoch": 2.6835797043547744, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6717 + }, + { + "epoch": 2.683979224930084, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6718 + }, + { + "epoch": 2.6843787455053936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6719 + }, + { + "epoch": 2.684778266080703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 6720 + }, + { + "epoch": 2.6851777866560127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6721 + }, + { + "epoch": 2.6855773072313225, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6722 + }, + { + "epoch": 2.6859768278066323, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6723 + }, + { + "epoch": 2.6863763483819416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6724 + }, + { + "epoch": 2.6867758689572514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6725 + }, + { + "epoch": 2.6871753895325607, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6726 + }, + { + "epoch": 2.6875749101078705, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6727 + }, + { + "epoch": 2.6879744306831803, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 6728 + }, + { + "epoch": 2.6883739512584897, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6729 + }, + { + "epoch": 2.6887734718337994, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6730 + }, + { + "epoch": 2.6891729924091092, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6731 + }, + { + "epoch": 2.6895725129844186, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6732 + }, + { + "epoch": 2.6899720335597284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6733 + }, + { + "epoch": 2.690371554135038, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6734 + }, + { + "epoch": 2.6907710747103475, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6735 + }, + { + "epoch": 2.6911705952856573, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6736 + }, + { + "epoch": 2.6915701158609666, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6737 + }, + { + "epoch": 2.6919696364362764, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6738 + }, + { + "epoch": 2.692369157011586, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6739 + }, + { + "epoch": 2.692768677586896, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 6740 + }, + { + "epoch": 2.6931681981622053, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6741 + }, + { + "epoch": 2.693567718737515, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6742 + }, + { + "epoch": 2.6939672393128244, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6743 + }, + { + "epoch": 2.6943667598881342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6744 + }, + { + "epoch": 2.694766280463444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6745 + }, + { + "epoch": 2.6951658010387534, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 6746 + }, + { + "epoch": 2.695565321614063, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6747 + }, + { + "epoch": 2.6959648421893725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6748 + }, + { + "epoch": 2.6963643627646823, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6749 + }, + { + "epoch": 2.696763883339992, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 6750 + }, + { + "epoch": 2.697163403915302, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6751 + }, + { + "epoch": 2.697562924490611, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6752 + }, + { + "epoch": 2.697962445065921, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6753 + }, + { + "epoch": 2.6983619656412303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6754 + }, + { + "epoch": 2.69876148621654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6755 + }, + { + "epoch": 2.69916100679185, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6756 + }, + { + "epoch": 2.6995605273671597, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6757 + }, + { + "epoch": 2.699960047942469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6758 + }, + { + "epoch": 2.700359568517779, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6759 + }, + { + "epoch": 2.700759089093088, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6760 + }, + { + "epoch": 2.701158609668398, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6761 + }, + { + "epoch": 2.7015581302437077, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6762 + }, + { + "epoch": 2.701957650819017, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6763 + }, + { + "epoch": 2.702357171394327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6764 + }, + { + "epoch": 2.702756691969636, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6765 + }, + { + "epoch": 2.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6766 + }, + { + "epoch": 2.7035557331202558, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6767 + }, + { + "epoch": 2.7039552536955656, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6768 + }, + { + "epoch": 2.704354774270875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6769 + }, + { + "epoch": 2.7047542948461847, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 6770 + }, + { + "epoch": 2.705153815421494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6771 + }, + { + "epoch": 2.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6772 + }, + { + "epoch": 2.7059528565721136, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6773 + }, + { + "epoch": 2.706352377147423, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6774 + }, + { + "epoch": 2.7067518977227327, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6775 + }, + { + "epoch": 2.7071514182980425, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6776 + }, + { + "epoch": 2.707550938873352, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6777 + }, + { + "epoch": 2.7079504594486616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6778 + }, + { + "epoch": 2.7083499800239714, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6779 + }, + { + "epoch": 2.7087495005992808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6780 + }, + { + "epoch": 2.7091490211745906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 6781 + }, + { + "epoch": 2.7095485417499, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6782 + }, + { + "epoch": 2.7099480623252097, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6783 + }, + { + "epoch": 2.7103475829005195, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 6784 + }, + { + "epoch": 2.7107471034758293, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6785 + }, + { + "epoch": 2.7111466240511386, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6786 + }, + { + "epoch": 2.7115461446264484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 6787 + }, + { + "epoch": 2.7119456652017577, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6788 + }, + { + "epoch": 2.7123451857770675, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6789 + }, + { + "epoch": 2.7127447063523773, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6790 + }, + { + "epoch": 2.7131442269276866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 6791 + }, + { + "epoch": 2.7135437475029964, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6792 + }, + { + "epoch": 2.7139432680783058, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6793 + }, + { + "epoch": 2.7143427886536156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6794 + }, + { + "epoch": 2.7147423092289253, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6795 + }, + { + "epoch": 2.715141829804235, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6796 + }, + { + "epoch": 2.7155413503795445, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6797 + }, + { + "epoch": 2.7159408709548543, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 6798 + }, + { + "epoch": 2.7163403915301636, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6799 + }, + { + "epoch": 2.7167399121054734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6800 + }, + { + "epoch": 2.717139432680783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6801 + }, + { + "epoch": 2.717538953256093, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6802 + }, + { + "epoch": 2.7179384738314023, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6803 + }, + { + "epoch": 2.718337994406712, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6804 + }, + { + "epoch": 2.7187375149820214, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6805 + }, + { + "epoch": 2.719137035557331, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6806 + }, + { + "epoch": 2.719536556132641, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6807 + }, + { + "epoch": 2.7199360767079503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6808 + }, + { + "epoch": 2.72033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6809 + }, + { + "epoch": 2.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6810 + }, + { + "epoch": 2.7211346384338793, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6811 + }, + { + "epoch": 2.721534159009189, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6812 + }, + { + "epoch": 2.721933679584499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6813 + }, + { + "epoch": 2.722333200159808, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6814 + }, + { + "epoch": 2.722732720735118, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6815 + }, + { + "epoch": 2.7231322413104273, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6816 + }, + { + "epoch": 2.723531761885737, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6817 + }, + { + "epoch": 2.723931282461047, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6818 + }, + { + "epoch": 2.724330803036356, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6819 + }, + { + "epoch": 2.724730323611666, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6820 + }, + { + "epoch": 2.725129844186976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6821 + }, + { + "epoch": 2.725529364762285, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6822 + }, + { + "epoch": 2.725928885337595, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6823 + }, + { + "epoch": 2.7263284059129047, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6824 + }, + { + "epoch": 2.726727926488214, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6825 + }, + { + "epoch": 2.727127447063524, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6826 + }, + { + "epoch": 2.727526967638833, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6827 + }, + { + "epoch": 2.727926488214143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6828 + }, + { + "epoch": 2.7283260087894528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6829 + }, + { + "epoch": 2.7287255293647625, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6830 + }, + { + "epoch": 2.729125049940072, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6831 + }, + { + "epoch": 2.7295245705153817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6832 + }, + { + "epoch": 2.729924091090691, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6833 + }, + { + "epoch": 2.730323611666001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6834 + }, + { + "epoch": 2.7307231322413106, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6835 + }, + { + "epoch": 2.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 6836 + }, + { + "epoch": 2.7315221733919297, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6837 + }, + { + "epoch": 2.731921693967239, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6838 + }, + { + "epoch": 2.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6839 + }, + { + "epoch": 2.7327207351178586, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6840 + }, + { + "epoch": 2.7331202556931684, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6841 + }, + { + "epoch": 2.7335197762684778, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6842 + }, + { + "epoch": 2.7339192968437875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6843 + }, + { + "epoch": 2.734318817419097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6844 + }, + { + "epoch": 2.7347183379944067, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6845 + }, + { + "epoch": 2.7351178585697165, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6846 + }, + { + "epoch": 2.7355173791450262, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6847 + }, + { + "epoch": 2.7359168997203356, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6848 + }, + { + "epoch": 2.7363164202956454, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6849 + }, + { + "epoch": 2.7367159408709547, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6850 + }, + { + "epoch": 2.7371154614462645, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6851 + }, + { + "epoch": 2.7375149820215743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6852 + }, + { + "epoch": 2.7379145025968836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6853 + }, + { + "epoch": 2.7383140231721934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6854 + }, + { + "epoch": 2.7387135437475028, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6855 + }, + { + "epoch": 2.7391130643228125, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6856 + }, + { + "epoch": 2.7395125848981223, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6857 + }, + { + "epoch": 2.739912105473432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6858 + }, + { + "epoch": 2.7403116260487415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6859 + }, + { + "epoch": 2.7407111466240512, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6860 + }, + { + "epoch": 2.7411106671993606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6861 + }, + { + "epoch": 2.7415101877746704, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6862 + }, + { + "epoch": 2.74190970834998, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6863 + }, + { + "epoch": 2.7423092289252895, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6864 + }, + { + "epoch": 2.7427087495005993, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6865 + }, + { + "epoch": 2.743108270075909, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6866 + }, + { + "epoch": 2.7435077906512184, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6867 + }, + { + "epoch": 2.743907311226528, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6868 + }, + { + "epoch": 2.744306831801838, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6869 + }, + { + "epoch": 2.7447063523771473, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6870 + }, + { + "epoch": 2.745105872952457, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6871 + }, + { + "epoch": 2.7455053935277665, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6872 + }, + { + "epoch": 2.7459049141030762, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6873 + }, + { + "epoch": 2.746304434678386, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6874 + }, + { + "epoch": 2.746703955253696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6875 + }, + { + "epoch": 2.747103475829005, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6876 + }, + { + "epoch": 2.747502996404315, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 6877 + }, + { + "epoch": 2.7479025169796243, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6878 + }, + { + "epoch": 2.748302037554934, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6879 + }, + { + "epoch": 2.748701558130244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6880 + }, + { + "epoch": 2.749101078705553, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6881 + }, + { + "epoch": 2.749500599280863, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6882 + }, + { + "epoch": 2.7499001198561723, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6883 + }, + { + "epoch": 2.750299640431482, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6884 + }, + { + "epoch": 2.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6885 + }, + { + "epoch": 2.7510986815821017, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6886 + }, + { + "epoch": 2.751498202157411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6887 + }, + { + "epoch": 2.751897722732721, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6888 + }, + { + "epoch": 2.75229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6889 + }, + { + "epoch": 2.75269676388334, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6890 + }, + { + "epoch": 2.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6891 + }, + { + "epoch": 2.7534958050339595, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6892 + }, + { + "epoch": 2.753895325609269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6893 + }, + { + "epoch": 2.7542948461845786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6894 + }, + { + "epoch": 2.754694366759888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6895 + }, + { + "epoch": 2.7550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6896 + }, + { + "epoch": 2.7554934079105076, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6897 + }, + { + "epoch": 2.755892928485817, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6898 + }, + { + "epoch": 2.7562924490611267, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6899 + }, + { + "epoch": 2.756691969636436, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6900 + }, + { + "epoch": 2.757091490211746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6901 + }, + { + "epoch": 2.7574910107870556, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6902 + }, + { + "epoch": 2.7578905313623654, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6903 + }, + { + "epoch": 2.7582900519376747, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6904 + }, + { + "epoch": 2.7586895725129845, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6905 + }, + { + "epoch": 2.759089093088294, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6906 + }, + { + "epoch": 2.7594886136636037, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6907 + }, + { + "epoch": 2.7598881342389134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6908 + }, + { + "epoch": 2.760287654814223, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6909 + }, + { + "epoch": 2.7606871753895326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6910 + }, + { + "epoch": 2.7610866959648424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6911 + }, + { + "epoch": 2.7614862165401517, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6912 + }, + { + "epoch": 2.7618857371154615, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6913 + }, + { + "epoch": 2.7622852576907713, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6914 + }, + { + "epoch": 2.7626847782660806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6915 + }, + { + "epoch": 2.7630842988413904, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6916 + }, + { + "epoch": 2.7634838194166997, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6917 + }, + { + "epoch": 2.7638833399920095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6918 + }, + { + "epoch": 2.7642828605673193, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6919 + }, + { + "epoch": 2.764682381142629, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6920 + }, + { + "epoch": 2.7650819017179384, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6921 + }, + { + "epoch": 2.7654814222932482, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6922 + }, + { + "epoch": 2.7658809428685576, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6923 + }, + { + "epoch": 2.7662804634438674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6924 + }, + { + "epoch": 2.766679984019177, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6925 + }, + { + "epoch": 2.7670795045944865, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6926 + }, + { + "epoch": 2.7674790251697963, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6927 + }, + { + "epoch": 2.7678785457451056, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6928 + }, + { + "epoch": 2.7682780663204154, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6929 + }, + { + "epoch": 2.768677586895725, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6930 + }, + { + "epoch": 2.769077107471035, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6931 + }, + { + "epoch": 2.7694766280463443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6932 + }, + { + "epoch": 2.769876148621654, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6933 + }, + { + "epoch": 2.7702756691969634, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6934 + }, + { + "epoch": 2.7706751897722732, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6935 + }, + { + "epoch": 2.771074710347583, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6936 + }, + { + "epoch": 2.771474230922893, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6937 + }, + { + "epoch": 2.771873751498202, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6938 + }, + { + "epoch": 2.772273272073512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6939 + }, + { + "epoch": 2.7726727926488213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6940 + }, + { + "epoch": 2.773072313224131, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6941 + }, + { + "epoch": 2.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6942 + }, + { + "epoch": 2.77387135437475, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6943 + }, + { + "epoch": 2.77427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6944 + }, + { + "epoch": 2.7746703955253693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 6945 + }, + { + "epoch": 2.775069916100679, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6946 + }, + { + "epoch": 2.775469436675989, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6947 + }, + { + "epoch": 2.7758689572512987, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 6948 + }, + { + "epoch": 2.776268477826608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6949 + }, + { + "epoch": 2.776667998401918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6950 + }, + { + "epoch": 2.777067518977227, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6951 + }, + { + "epoch": 2.777467039552537, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6952 + }, + { + "epoch": 2.7778665601278467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6953 + }, + { + "epoch": 2.778266080703156, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6954 + }, + { + "epoch": 2.778665601278466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6955 + }, + { + "epoch": 2.7790651218537756, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6956 + }, + { + "epoch": 2.779464642429085, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6957 + }, + { + "epoch": 2.7798641630043948, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6958 + }, + { + "epoch": 2.7802636835797045, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6959 + }, + { + "epoch": 2.780663204155014, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6960 + }, + { + "epoch": 2.7810627247303237, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6961 + }, + { + "epoch": 2.781462245305633, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6962 + }, + { + "epoch": 2.781861765880943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6963 + }, + { + "epoch": 2.7822612864562526, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6964 + }, + { + "epoch": 2.7826608070315624, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6965 + }, + { + "epoch": 2.7830603276068717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6966 + }, + { + "epoch": 2.7834598481821815, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6967 + }, + { + "epoch": 2.783859368757491, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6968 + }, + { + "epoch": 2.7842588893328006, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6969 + }, + { + "epoch": 2.7846584099081104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6970 + }, + { + "epoch": 2.7850579304834198, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 6971 + }, + { + "epoch": 2.7854574510587295, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6972 + }, + { + "epoch": 2.785856971634039, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6973 + }, + { + "epoch": 2.7862564922093487, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6974 + }, + { + "epoch": 2.7866560127846585, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6975 + }, + { + "epoch": 2.7870555333599683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6976 + }, + { + "epoch": 2.7874550539352776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6977 + }, + { + "epoch": 2.7878545745105874, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 6978 + }, + { + "epoch": 2.7882540950858967, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6979 + }, + { + "epoch": 2.7886536156612065, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6980 + }, + { + "epoch": 2.7890531362365163, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6981 + }, + { + "epoch": 2.789452656811826, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6982 + }, + { + "epoch": 2.7898521773871354, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6983 + }, + { + "epoch": 2.790251697962445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 6984 + }, + { + "epoch": 2.7906512185377546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6985 + }, + { + "epoch": 2.7910507391130643, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6986 + }, + { + "epoch": 2.791450259688374, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6987 + }, + { + "epoch": 2.7918497802636835, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6988 + }, + { + "epoch": 2.7922493008389933, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6989 + }, + { + "epoch": 2.7926488214143026, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6990 + }, + { + "epoch": 2.7930483419896124, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6991 + }, + { + "epoch": 2.793447862564922, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6992 + }, + { + "epoch": 2.793847383140232, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6993 + }, + { + "epoch": 2.7942469037155413, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6994 + }, + { + "epoch": 2.794646424290851, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6995 + }, + { + "epoch": 2.7950459448661604, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 6996 + }, + { + "epoch": 2.79544546544147, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6997 + }, + { + "epoch": 2.79584498601678, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6998 + }, + { + "epoch": 2.7962445065920893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6999 + }, + { + "epoch": 2.796644027167399, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7000 + }, + { + "epoch": 2.797043547742709, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7001 + }, + { + "epoch": 2.7974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7002 + }, + { + "epoch": 2.797842588893328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7003 + }, + { + "epoch": 2.798242109468638, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7004 + }, + { + "epoch": 2.798641630043947, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7005 + }, + { + "epoch": 2.799041150619257, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7006 + }, + { + "epoch": 2.7994406711945663, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7007 + }, + { + "epoch": 2.799840191769876, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7008 + }, + { + "epoch": 2.800239712345186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 7009 + }, + { + "epoch": 2.8006392329204957, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7010 + }, + { + "epoch": 2.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7011 + }, + { + "epoch": 2.801438274071115, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7012 + }, + { + "epoch": 2.801837794646424, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7013 + }, + { + "epoch": 2.802237315221734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7014 + }, + { + "epoch": 2.8026368357970437, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7015 + }, + { + "epoch": 2.803036356372353, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7016 + }, + { + "epoch": 2.803435876947663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7017 + }, + { + "epoch": 2.803835397522972, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7018 + }, + { + "epoch": 2.804234918098282, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7019 + }, + { + "epoch": 2.8046344386735917, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7020 + }, + { + "epoch": 2.8050339592489015, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7021 + }, + { + "epoch": 2.805433479824211, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7022 + }, + { + "epoch": 2.8058330003995207, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7023 + }, + { + "epoch": 2.80623252097483, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 7024 + }, + { + "epoch": 2.80663204155014, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7025 + }, + { + "epoch": 2.8070315621254496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7026 + }, + { + "epoch": 2.8074310827007594, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7027 + }, + { + "epoch": 2.8078306032760687, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7028 + }, + { + "epoch": 2.8082301238513785, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7029 + }, + { + "epoch": 2.808629644426688, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7030 + }, + { + "epoch": 2.8090291650019976, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7031 + }, + { + "epoch": 2.8094286855773074, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7032 + }, + { + "epoch": 2.8098282061526167, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7033 + }, + { + "epoch": 2.8102277267279265, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7034 + }, + { + "epoch": 2.810627247303236, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7035 + }, + { + "epoch": 2.8110267678785457, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7036 + }, + { + "epoch": 2.8114262884538554, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7037 + }, + { + "epoch": 2.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7038 + }, + { + "epoch": 2.8122253296044746, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7039 + }, + { + "epoch": 2.8126248501797844, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7040 + }, + { + "epoch": 2.8130243707550937, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7041 + }, + { + "epoch": 2.8134238913304035, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7042 + }, + { + "epoch": 2.8138234119057133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7043 + }, + { + "epoch": 2.8142229324810226, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7044 + }, + { + "epoch": 2.8146224530563324, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7045 + }, + { + "epoch": 2.815021973631642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 7046 + }, + { + "epoch": 2.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7047 + }, + { + "epoch": 2.8158210147822613, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 7048 + }, + { + "epoch": 2.816220535357571, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7049 + }, + { + "epoch": 2.8166200559328805, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7050 + }, + { + "epoch": 2.8170195765081902, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7051 + }, + { + "epoch": 2.8174190970834996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7052 + }, + { + "epoch": 2.8178186176588094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 7053 + }, + { + "epoch": 2.818218138234119, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7054 + }, + { + "epoch": 2.818617658809429, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7055 + }, + { + "epoch": 2.8190171793847383, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7056 + }, + { + "epoch": 2.819416699960048, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7057 + }, + { + "epoch": 2.8198162205353574, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7058 + }, + { + "epoch": 2.820215741110667, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7059 + }, + { + "epoch": 2.820615261685977, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7060 + }, + { + "epoch": 2.8210147822612863, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7061 + }, + { + "epoch": 2.821414302836596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7062 + }, + { + "epoch": 2.8218138234119055, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7063 + }, + { + "epoch": 2.8222133439872152, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7064 + }, + { + "epoch": 2.822612864562525, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7065 + }, + { + "epoch": 2.823012385137835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7066 + }, + { + "epoch": 2.823411905713144, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7067 + }, + { + "epoch": 2.823811426288454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7068 + }, + { + "epoch": 2.8242109468637633, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7069 + }, + { + "epoch": 2.824610467439073, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7070 + }, + { + "epoch": 2.825009988014383, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7071 + }, + { + "epoch": 2.8254095085896926, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7072 + }, + { + "epoch": 2.825809029165002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7073 + }, + { + "epoch": 2.8262085497403118, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 7074 + }, + { + "epoch": 2.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7075 + }, + { + "epoch": 2.827007590890931, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 7076 + }, + { + "epoch": 2.8274071114662407, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7077 + }, + { + "epoch": 2.82780663204155, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 7078 + }, + { + "epoch": 2.82820615261686, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7079 + }, + { + "epoch": 2.828605673192169, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7080 + }, + { + "epoch": 2.829005193767479, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7081 + }, + { + "epoch": 2.8294047143427887, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7082 + }, + { + "epoch": 2.8298042349180985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7083 + }, + { + "epoch": 2.830203755493408, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7084 + }, + { + "epoch": 2.8306032760687176, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 7085 + }, + { + "epoch": 2.831002796644027, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7086 + }, + { + "epoch": 2.8314023172193368, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7087 + }, + { + "epoch": 2.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7088 + }, + { + "epoch": 2.832201358369956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7089 + }, + { + "epoch": 2.8326008789452657, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7090 + }, + { + "epoch": 2.8330003995205755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 7091 + }, + { + "epoch": 2.833399920095885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7092 + }, + { + "epoch": 2.8337994406711946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7093 + }, + { + "epoch": 2.8341989612465044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7094 + }, + { + "epoch": 2.8345984818218137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 7095 + }, + { + "epoch": 2.8349980023971235, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7096 + }, + { + "epoch": 2.835397522972433, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7097 + }, + { + "epoch": 2.8357970435477426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7098 + }, + { + "epoch": 2.8361965641230524, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7099 + }, + { + "epoch": 2.836596084698362, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7100 + }, + { + "epoch": 2.8369956052736716, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7101 + }, + { + "epoch": 2.8373951258489813, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7102 + }, + { + "epoch": 2.8377946464242907, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7103 + }, + { + "epoch": 2.8381941669996005, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7104 + }, + { + "epoch": 2.8385936875749103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7105 + }, + { + "epoch": 2.8389932081502196, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7106 + }, + { + "epoch": 2.8393927287255294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7107 + }, + { + "epoch": 2.8397922493008387, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7108 + }, + { + "epoch": 2.8401917698761485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7109 + }, + { + "epoch": 2.8405912904514583, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7110 + }, + { + "epoch": 2.840990811026768, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7111 + }, + { + "epoch": 2.8413903316020774, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7112 + }, + { + "epoch": 2.841789852177387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7113 + }, + { + "epoch": 2.8421893727526966, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7114 + }, + { + "epoch": 2.8425888933280063, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7115 + }, + { + "epoch": 2.842988413903316, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 7116 + }, + { + "epoch": 2.843387934478626, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7117 + }, + { + "epoch": 2.8437874550539353, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7118 + }, + { + "epoch": 2.844186975629245, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7119 + }, + { + "epoch": 2.8445864962045544, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7120 + }, + { + "epoch": 2.844986016779864, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7121 + }, + { + "epoch": 2.845385537355174, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7122 + }, + { + "epoch": 2.8457850579304833, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7123 + }, + { + "epoch": 2.846184578505793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 7124 + }, + { + "epoch": 2.8465840990811024, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7125 + }, + { + "epoch": 2.8469836196564122, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7126 + }, + { + "epoch": 2.847383140231722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7127 + }, + { + "epoch": 2.847782660807032, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7128 + }, + { + "epoch": 2.848182181382341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7129 + }, + { + "epoch": 2.848581701957651, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7130 + }, + { + "epoch": 2.8489812225329603, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7131 + }, + { + "epoch": 2.84938074310827, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7132 + }, + { + "epoch": 2.84978026368358, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7133 + }, + { + "epoch": 2.850179784258889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7134 + }, + { + "epoch": 2.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7135 + }, + { + "epoch": 2.8509788254095088, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7136 + }, + { + "epoch": 2.851378345984818, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 7137 + }, + { + "epoch": 2.851777866560128, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7138 + }, + { + "epoch": 2.8521773871354377, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7139 + }, + { + "epoch": 2.852576907710747, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7140 + }, + { + "epoch": 2.852976428286057, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7141 + }, + { + "epoch": 2.853375948861366, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7142 + }, + { + "epoch": 2.853775469436676, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 7143 + }, + { + "epoch": 2.8541749900119857, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7144 + }, + { + "epoch": 2.8545745105872955, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7145 + }, + { + "epoch": 2.854974031162605, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7146 + }, + { + "epoch": 2.8553735517379146, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7147 + }, + { + "epoch": 2.855773072313224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7148 + }, + { + "epoch": 2.8561725928885338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7149 + }, + { + "epoch": 2.8565721134638435, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7150 + }, + { + "epoch": 2.856971634039153, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7151 + }, + { + "epoch": 2.8573711546144627, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7152 + }, + { + "epoch": 2.857770675189772, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7153 + }, + { + "epoch": 2.858170195765082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7154 + }, + { + "epoch": 2.8585697163403916, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7155 + }, + { + "epoch": 2.8589692369157014, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7156 + }, + { + "epoch": 2.8593687574910107, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7157 + }, + { + "epoch": 2.8597682780663205, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7158 + }, + { + "epoch": 2.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7159 + }, + { + "epoch": 2.8605673192169396, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7160 + }, + { + "epoch": 2.8609668397922494, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7161 + }, + { + "epoch": 2.861366360367559, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7162 + }, + { + "epoch": 2.8617658809428685, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7163 + }, + { + "epoch": 2.8621654015181783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7164 + }, + { + "epoch": 2.8625649220934877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 7165 + }, + { + "epoch": 2.8629644426687975, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7166 + }, + { + "epoch": 2.8633639632441072, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7167 + }, + { + "epoch": 2.8637634838194166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7168 + }, + { + "epoch": 2.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7169 + }, + { + "epoch": 2.8645625249700357, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7170 + }, + { + "epoch": 2.8649620455453455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7171 + }, + { + "epoch": 2.8653615661206553, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 7172 + }, + { + "epoch": 2.865761086695965, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7173 + }, + { + "epoch": 2.8661606072712744, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7174 + }, + { + "epoch": 2.866560127846584, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7175 + }, + { + "epoch": 2.8669596484218935, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7176 + }, + { + "epoch": 2.8673591689972033, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7177 + }, + { + "epoch": 2.867758689572513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7178 + }, + { + "epoch": 2.8681582101478225, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7179 + }, + { + "epoch": 2.8685577307231322, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7180 + }, + { + "epoch": 2.868957251298442, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7181 + }, + { + "epoch": 2.8693567718737514, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7182 + }, + { + "epoch": 2.869756292449061, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7183 + }, + { + "epoch": 2.870155813024371, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7184 + }, + { + "epoch": 2.8705553335996803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7185 + }, + { + "epoch": 2.87095485417499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7186 + }, + { + "epoch": 2.8713543747502994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7187 + }, + { + "epoch": 2.871753895325609, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7188 + }, + { + "epoch": 2.872153415900919, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7189 + }, + { + "epoch": 2.872552936476229, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7190 + }, + { + "epoch": 2.872952457051538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7191 + }, + { + "epoch": 2.873351977626848, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7192 + }, + { + "epoch": 2.8737514982021573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7193 + }, + { + "epoch": 2.874151018777467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7194 + }, + { + "epoch": 2.874550539352777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7195 + }, + { + "epoch": 2.874950059928086, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 7196 + }, + { + "epoch": 2.875349580503396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7197 + }, + { + "epoch": 2.8757491010787053, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7198 + }, + { + "epoch": 2.876148621654015, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7199 + }, + { + "epoch": 2.876548142229325, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 7200 + }, + { + "epoch": 2.8769476628046347, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7201 + }, + { + "epoch": 2.877347183379944, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7202 + }, + { + "epoch": 2.877746703955254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7203 + }, + { + "epoch": 2.878146224530563, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7204 + }, + { + "epoch": 2.878545745105873, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7205 + }, + { + "epoch": 2.8789452656811827, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 7206 + }, + { + "epoch": 2.8793447862564925, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7207 + }, + { + "epoch": 2.879744306831802, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7208 + }, + { + "epoch": 2.8801438274071116, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7209 + }, + { + "epoch": 2.880543347982421, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 7210 + }, + { + "epoch": 2.8809428685577307, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7211 + }, + { + "epoch": 2.8813423891330405, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7212 + }, + { + "epoch": 2.88174190970835, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7213 + }, + { + "epoch": 2.8821414302836597, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7214 + }, + { + "epoch": 2.882540950858969, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7215 + }, + { + "epoch": 2.882940471434279, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7216 + }, + { + "epoch": 2.8833399920095886, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7217 + }, + { + "epoch": 2.8837395125848984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 7218 + }, + { + "epoch": 2.8841390331602077, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7219 + }, + { + "epoch": 2.8845385537355175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7220 + }, + { + "epoch": 2.884938074310827, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7221 + }, + { + "epoch": 2.8853375948861366, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7222 + }, + { + "epoch": 2.8857371154614464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7223 + }, + { + "epoch": 2.8861366360367557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7224 + }, + { + "epoch": 2.8865361566120655, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7225 + }, + { + "epoch": 2.8869356771873753, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7226 + }, + { + "epoch": 2.8873351977626847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7227 + }, + { + "epoch": 2.8877347183379944, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7228 + }, + { + "epoch": 2.8881342389133042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7229 + }, + { + "epoch": 2.8885337594886136, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7230 + }, + { + "epoch": 2.8889332800639234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7231 + }, + { + "epoch": 2.8893328006392327, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7232 + }, + { + "epoch": 2.8897323212145425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 7233 + }, + { + "epoch": 2.8901318417898523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7234 + }, + { + "epoch": 2.890531362365162, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7235 + }, + { + "epoch": 2.8909308829404714, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7236 + }, + { + "epoch": 2.891330403515781, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7237 + }, + { + "epoch": 2.8917299240910905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7238 + }, + { + "epoch": 2.8921294446664003, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7239 + }, + { + "epoch": 2.89252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7240 + }, + { + "epoch": 2.8929284858170194, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7241 + }, + { + "epoch": 2.8933280063923292, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7242 + }, + { + "epoch": 2.8937275269676386, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7243 + }, + { + "epoch": 2.8941270475429484, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7244 + }, + { + "epoch": 2.894526568118258, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 7245 + }, + { + "epoch": 2.894926088693568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7246 + }, + { + "epoch": 2.8953256092688773, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7247 + }, + { + "epoch": 2.895725129844187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7248 + }, + { + "epoch": 2.8961246504194964, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7249 + }, + { + "epoch": 2.896524170994806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 7250 + }, + { + "epoch": 2.896923691570116, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 7251 + }, + { + "epoch": 2.8973232121454258, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7252 + }, + { + "epoch": 2.897722732720735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7253 + }, + { + "epoch": 2.898122253296045, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7254 + }, + { + "epoch": 2.8985217738713542, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7255 + }, + { + "epoch": 2.898921294446664, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 7256 + }, + { + "epoch": 2.899320815021974, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7257 + }, + { + "epoch": 2.899720335597283, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7258 + }, + { + "epoch": 2.900119856172593, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7259 + }, + { + "epoch": 2.9005193767479023, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7260 + }, + { + "epoch": 2.900918897323212, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7261 + }, + { + "epoch": 2.901318417898522, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7262 + }, + { + "epoch": 2.9017179384738316, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7263 + }, + { + "epoch": 2.902117459049141, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7264 + }, + { + "epoch": 2.9025169796244508, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7265 + }, + { + "epoch": 2.90291650019976, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 7266 + }, + { + "epoch": 2.90331602077507, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7267 + }, + { + "epoch": 2.9037155413503797, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7268 + }, + { + "epoch": 2.904115061925689, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7269 + }, + { + "epoch": 2.904514582500999, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7270 + }, + { + "epoch": 2.9049141030763086, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7271 + }, + { + "epoch": 2.905313623651618, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7272 + }, + { + "epoch": 2.9057131442269277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7273 + }, + { + "epoch": 2.9061126648022375, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7274 + }, + { + "epoch": 2.906512185377547, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7275 + }, + { + "epoch": 2.9069117059528566, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7276 + }, + { + "epoch": 2.907311226528166, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7277 + }, + { + "epoch": 2.9077107471034758, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7278 + }, + { + "epoch": 2.9081102676787856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7279 + }, + { + "epoch": 2.9085097882540953, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7280 + }, + { + "epoch": 2.9089093088294047, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 7281 + }, + { + "epoch": 2.9093088294047145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7282 + }, + { + "epoch": 2.909708349980024, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7283 + }, + { + "epoch": 2.9101078705553336, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7284 + }, + { + "epoch": 2.9105073911306434, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7285 + }, + { + "epoch": 2.9109069117059527, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7286 + }, + { + "epoch": 2.9113064322812625, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7287 + }, + { + "epoch": 2.911705952856572, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 7288 + }, + { + "epoch": 2.9121054734318816, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7289 + }, + { + "epoch": 2.9125049940071914, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7290 + }, + { + "epoch": 2.912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7291 + }, + { + "epoch": 2.9133040351578106, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7292 + }, + { + "epoch": 2.9137035557331203, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7293 + }, + { + "epoch": 2.9141030763084297, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7294 + }, + { + "epoch": 2.9145025968837395, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7295 + }, + { + "epoch": 2.9149021174590493, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7296 + }, + { + "epoch": 2.915301638034359, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7297 + }, + { + "epoch": 2.9157011586096684, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7298 + }, + { + "epoch": 2.916100679184978, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7299 + }, + { + "epoch": 2.9165001997602875, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7300 + }, + { + "epoch": 2.9168997203355973, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7301 + }, + { + "epoch": 2.917299240910907, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7302 + }, + { + "epoch": 2.9176987614862164, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7303 + }, + { + "epoch": 2.918098282061526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7304 + }, + { + "epoch": 2.9184978026368356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7305 + }, + { + "epoch": 2.9188973232121453, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7306 + }, + { + "epoch": 2.919296843787455, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7307 + }, + { + "epoch": 2.919696364362765, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7308 + }, + { + "epoch": 2.9200958849380743, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7309 + }, + { + "epoch": 2.920495405513384, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7310 + }, + { + "epoch": 2.9208949260886934, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 7311 + }, + { + "epoch": 2.921294446664003, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7312 + }, + { + "epoch": 2.921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7313 + }, + { + "epoch": 2.9220934878146223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 7314 + }, + { + "epoch": 2.922493008389932, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7315 + }, + { + "epoch": 2.922892528965242, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7316 + }, + { + "epoch": 2.923292049540551, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7317 + }, + { + "epoch": 2.923691570115861, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7318 + }, + { + "epoch": 2.924091090691171, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7319 + }, + { + "epoch": 2.92449061126648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7320 + }, + { + "epoch": 2.92489013184179, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7321 + }, + { + "epoch": 2.9252896524170993, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7322 + }, + { + "epoch": 2.925689172992409, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7323 + }, + { + "epoch": 2.926088693567719, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7324 + }, + { + "epoch": 2.9264882141430286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7325 + }, + { + "epoch": 2.926887734718338, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7326 + }, + { + "epoch": 2.9272872552936477, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7327 + }, + { + "epoch": 2.927686775868957, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7328 + }, + { + "epoch": 2.928086296444267, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7329 + }, + { + "epoch": 2.9284858170195767, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7330 + }, + { + "epoch": 2.928885337594886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 7331 + }, + { + "epoch": 2.929284858170196, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7332 + }, + { + "epoch": 2.929684378745505, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7333 + }, + { + "epoch": 2.930083899320815, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7334 + }, + { + "epoch": 2.9304834198961247, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7335 + }, + { + "epoch": 2.9308829404714345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7336 + }, + { + "epoch": 2.931282461046744, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 7337 + }, + { + "epoch": 2.9316819816220536, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7338 + }, + { + "epoch": 2.932081502197363, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7339 + }, + { + "epoch": 2.9324810227726728, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7340 + }, + { + "epoch": 2.9328805433479825, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7341 + }, + { + "epoch": 2.9332800639232923, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7342 + }, + { + "epoch": 2.9336795844986017, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7343 + }, + { + "epoch": 2.9340791050739115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7344 + }, + { + "epoch": 2.934478625649221, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7345 + }, + { + "epoch": 2.9348781462245306, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7346 + }, + { + "epoch": 2.9352776667998404, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7347 + }, + { + "epoch": 2.9356771873751497, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7348 + }, + { + "epoch": 2.9360767079504595, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7349 + }, + { + "epoch": 2.936476228525769, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7350 + }, + { + "epoch": 2.9368757491010786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7351 + }, + { + "epoch": 2.9372752696763884, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7352 + }, + { + "epoch": 2.937674790251698, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7353 + }, + { + "epoch": 2.9380743108270075, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7354 + }, + { + "epoch": 2.9384738314023173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7355 + }, + { + "epoch": 2.9388733519776267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7356 + }, + { + "epoch": 2.9392728725529365, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7357 + }, + { + "epoch": 2.9396723931282462, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7358 + }, + { + "epoch": 2.9400719137035556, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7359 + }, + { + "epoch": 2.9404714342788654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7360 + }, + { + "epoch": 2.9408709548541747, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7361 + }, + { + "epoch": 2.9412704754294845, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7362 + }, + { + "epoch": 2.9416699960047943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7363 + }, + { + "epoch": 2.942069516580104, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7364 + }, + { + "epoch": 2.9424690371554134, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7365 + }, + { + "epoch": 2.942868557730723, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7366 + }, + { + "epoch": 2.9432680783060325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7367 + }, + { + "epoch": 2.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7368 + }, + { + "epoch": 2.944067119456652, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7369 + }, + { + "epoch": 2.944466640031962, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7370 + }, + { + "epoch": 2.9448661606072712, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7371 + }, + { + "epoch": 2.945265681182581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7372 + }, + { + "epoch": 2.9456652017578904, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7373 + }, + { + "epoch": 2.9460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7374 + }, + { + "epoch": 2.94646424290851, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7375 + }, + { + "epoch": 2.9468637634838193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7376 + }, + { + "epoch": 2.947263284059129, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7377 + }, + { + "epoch": 2.9476628046344384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7378 + }, + { + "epoch": 2.948062325209748, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7379 + }, + { + "epoch": 2.948461845785058, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7380 + }, + { + "epoch": 2.9488613663603678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7381 + }, + { + "epoch": 2.949260886935677, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7382 + }, + { + "epoch": 2.949660407510987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 7383 + }, + { + "epoch": 2.9500599280862962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7384 + }, + { + "epoch": 2.950459448661606, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7385 + }, + { + "epoch": 2.950858969236916, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7386 + }, + { + "epoch": 2.9512584898122256, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7387 + }, + { + "epoch": 2.951658010387535, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7388 + }, + { + "epoch": 2.9520575309628447, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7389 + }, + { + "epoch": 2.952457051538154, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7390 + }, + { + "epoch": 2.952856572113464, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7391 + }, + { + "epoch": 2.9532560926887736, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7392 + }, + { + "epoch": 2.953655613264083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7393 + }, + { + "epoch": 2.9540551338393928, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7394 + }, + { + "epoch": 2.954454654414702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7395 + }, + { + "epoch": 2.954854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7396 + }, + { + "epoch": 2.9552536955653217, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7397 + }, + { + "epoch": 2.9556532161406315, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7398 + }, + { + "epoch": 2.956052736715941, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7399 + }, + { + "epoch": 2.9564522572912506, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7400 + }, + { + "epoch": 2.95685177786656, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7401 + }, + { + "epoch": 2.9572512984418697, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7402 + }, + { + "epoch": 2.9576508190171795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7403 + }, + { + "epoch": 2.958050339592489, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7404 + }, + { + "epoch": 2.9584498601677987, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7405 + }, + { + "epoch": 2.958849380743108, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7406 + }, + { + "epoch": 2.959248901318418, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 7407 + }, + { + "epoch": 2.9596484218937276, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7408 + }, + { + "epoch": 2.9600479424690374, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7409 + }, + { + "epoch": 2.9604474630443467, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7410 + }, + { + "epoch": 2.9608469836196565, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 7411 + }, + { + "epoch": 2.961246504194966, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7412 + }, + { + "epoch": 2.9616460247702756, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7413 + }, + { + "epoch": 2.9620455453455854, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7414 + }, + { + "epoch": 2.962445065920895, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7415 + }, + { + "epoch": 2.9628445864962045, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7416 + }, + { + "epoch": 2.9632441070715143, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7417 + }, + { + "epoch": 2.9636436276468237, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7418 + }, + { + "epoch": 2.9640431482221334, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7419 + }, + { + "epoch": 2.9644426687974432, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7420 + }, + { + "epoch": 2.9648421893727526, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7421 + }, + { + "epoch": 2.9652417099480624, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7422 + }, + { + "epoch": 2.9656412305233717, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7423 + }, + { + "epoch": 2.9660407510986815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7424 + }, + { + "epoch": 2.9664402716739913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7425 + }, + { + "epoch": 2.966839792249301, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7426 + }, + { + "epoch": 2.9672393128246104, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7427 + }, + { + "epoch": 2.96763883339992, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7428 + }, + { + "epoch": 2.9680383539752295, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7429 + }, + { + "epoch": 2.9684378745505393, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7430 + }, + { + "epoch": 2.968837395125849, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7431 + }, + { + "epoch": 2.969236915701159, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7432 + }, + { + "epoch": 2.9696364362764682, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7433 + }, + { + "epoch": 2.970035956851778, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7434 + }, + { + "epoch": 2.9704354774270874, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7435 + }, + { + "epoch": 2.970834998002397, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7436 + }, + { + "epoch": 2.971234518577707, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 7437 + }, + { + "epoch": 2.9716340391530163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7438 + }, + { + "epoch": 2.972033559728326, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7439 + }, + { + "epoch": 2.9724330803036354, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7440 + }, + { + "epoch": 2.972832600878945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7441 + }, + { + "epoch": 2.973232121454255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7442 + }, + { + "epoch": 2.9736316420295648, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7443 + }, + { + "epoch": 2.974031162604874, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7444 + }, + { + "epoch": 2.974430683180184, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7445 + }, + { + "epoch": 2.9748302037554932, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7446 + }, + { + "epoch": 2.975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7447 + }, + { + "epoch": 2.975629244906113, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7448 + }, + { + "epoch": 2.976028765481422, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7449 + }, + { + "epoch": 2.976428286056732, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7450 + }, + { + "epoch": 2.9768278066320413, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7451 + }, + { + "epoch": 2.977227327207351, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7452 + }, + { + "epoch": 2.977626847782661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7453 + }, + { + "epoch": 2.9780263683579706, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7454 + }, + { + "epoch": 2.97842588893328, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7455 + }, + { + "epoch": 2.9788254095085898, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7456 + }, + { + "epoch": 2.979224930083899, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7457 + }, + { + "epoch": 2.979624450659209, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7458 + }, + { + "epoch": 2.9800239712345187, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7459 + }, + { + "epoch": 2.9804234918098285, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7460 + }, + { + "epoch": 2.980823012385138, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7461 + }, + { + "epoch": 2.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7462 + }, + { + "epoch": 2.981622053535757, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7463 + }, + { + "epoch": 2.9820215741110667, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7464 + }, + { + "epoch": 2.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7465 + }, + { + "epoch": 2.982820615261686, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7466 + }, + { + "epoch": 2.9832201358369956, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7467 + }, + { + "epoch": 2.983619656412305, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7468 + }, + { + "epoch": 2.9840191769876148, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7469 + }, + { + "epoch": 2.9844186975629245, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7470 + }, + { + "epoch": 2.9848182181382343, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7471 + }, + { + "epoch": 2.9852177387135437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7472 + }, + { + "epoch": 2.9856172592888535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7473 + }, + { + "epoch": 2.986016779864163, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7474 + }, + { + "epoch": 2.9864163004394726, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7475 + }, + { + "epoch": 2.9868158210147824, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7476 + }, + { + "epoch": 2.987215341590092, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7477 + }, + { + "epoch": 2.9876148621654015, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7478 + }, + { + "epoch": 2.9880143827407113, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7479 + }, + { + "epoch": 2.9884139033160206, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 7480 + }, + { + "epoch": 2.9888134238913304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7481 + }, + { + "epoch": 2.98921294446664, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7482 + }, + { + "epoch": 2.9896124650419496, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7483 + }, + { + "epoch": 2.9900119856172593, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7484 + }, + { + "epoch": 2.9904115061925687, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7485 + }, + { + "epoch": 2.9908110267678785, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 7486 + }, + { + "epoch": 2.9912105473431883, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7487 + }, + { + "epoch": 2.991610067918498, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7488 + }, + { + "epoch": 2.9920095884938074, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7489 + }, + { + "epoch": 2.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7490 + }, + { + "epoch": 2.9928086296444265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7491 + }, + { + "epoch": 2.9932081502197363, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7492 + }, + { + "epoch": 2.993607670795046, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7493 + }, + { + "epoch": 2.9940071913703554, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7494 + }, + { + "epoch": 2.994406711945665, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7495 + }, + { + "epoch": 2.9948062325209746, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7496 + }, + { + "epoch": 2.9952057530962843, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7497 + }, + { + "epoch": 2.995605273671594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7498 + }, + { + "epoch": 2.996004794246904, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7499 + }, + { + "epoch": 2.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7500 + }, + { + "epoch": 2.996803835397523, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7501 + }, + { + "epoch": 2.9972033559728324, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7502 + }, + { + "epoch": 2.997602876548142, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7503 + }, + { + "epoch": 2.998002397123452, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7504 + }, + { + "epoch": 2.9984019176987617, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7505 + }, + { + "epoch": 2.998801438274071, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7506 + }, + { + "epoch": 2.999200958849381, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7507 + }, + { + "epoch": 2.99960047942469, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7508 + }, + { + "epoch": 3.0, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7509 + }, + { + "epoch": 3.00039952057531, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7510 + }, + { + "epoch": 3.000799041150619, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7511 + }, + { + "epoch": 3.001198561725929, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7512 + }, + { + "epoch": 3.0015980823012387, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7513 + }, + { + "epoch": 3.001997602876548, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7514 + }, + { + "epoch": 3.002397123451858, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7515 + }, + { + "epoch": 3.0027966440271676, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7516 + }, + { + "epoch": 3.003196164602477, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7517 + }, + { + "epoch": 3.0035956851777867, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7518 + }, + { + "epoch": 3.003995205753096, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7519 + }, + { + "epoch": 3.004394726328406, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7520 + }, + { + "epoch": 3.0047942469037157, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7521 + }, + { + "epoch": 3.005193767479025, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7522 + }, + { + "epoch": 3.005593288054335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7523 + }, + { + "epoch": 3.0059928086296446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7524 + }, + { + "epoch": 3.006392329204954, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 7525 + }, + { + "epoch": 3.0067918497802637, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7526 + }, + { + "epoch": 3.0071913703555735, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7527 + }, + { + "epoch": 3.007590890930883, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7528 + }, + { + "epoch": 3.0079904115061926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7529 + }, + { + "epoch": 3.0083899320815024, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7530 + }, + { + "epoch": 3.0087894526568117, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7531 + }, + { + "epoch": 3.0091889732321215, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7532 + }, + { + "epoch": 3.009588493807431, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7533 + }, + { + "epoch": 3.0099880143827407, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7534 + }, + { + "epoch": 3.0103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7535 + }, + { + "epoch": 3.01078705553336, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7536 + }, + { + "epoch": 3.0111865761086696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7537 + }, + { + "epoch": 3.0115860966839794, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7538 + }, + { + "epoch": 3.0119856172592887, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7539 + }, + { + "epoch": 3.0123851378345985, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7540 + }, + { + "epoch": 3.0127846584099083, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7541 + }, + { + "epoch": 3.0131841789852176, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7542 + }, + { + "epoch": 3.0135836995605274, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7543 + }, + { + "epoch": 3.013983220135837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7544 + }, + { + "epoch": 3.0143827407111465, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7545 + }, + { + "epoch": 3.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 7546 + }, + { + "epoch": 3.0151817818617657, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 7547 + }, + { + "epoch": 3.0155813024370755, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7548 + }, + { + "epoch": 3.0159808230123852, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7549 + }, + { + "epoch": 3.0163803435876946, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7550 + }, + { + "epoch": 3.0167798641630044, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7551 + }, + { + "epoch": 3.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7552 + }, + { + "epoch": 3.0175789053136235, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7553 + }, + { + "epoch": 3.0179784258889333, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7554 + }, + { + "epoch": 3.018377946464243, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7555 + }, + { + "epoch": 3.0187774670395524, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7556 + }, + { + "epoch": 3.019176987614862, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7557 + }, + { + "epoch": 3.019576508190172, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7558 + }, + { + "epoch": 3.0199760287654813, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7559 + }, + { + "epoch": 3.020375549340791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7560 + }, + { + "epoch": 3.020775069916101, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7561 + }, + { + "epoch": 3.0211745904914102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7562 + }, + { + "epoch": 3.02157411106672, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7563 + }, + { + "epoch": 3.0219736316420294, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7564 + }, + { + "epoch": 3.022373152217339, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7565 + }, + { + "epoch": 3.022772672792649, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7566 + }, + { + "epoch": 3.0231721933679583, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7567 + }, + { + "epoch": 3.023571713943268, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7568 + }, + { + "epoch": 3.023971234518578, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7569 + }, + { + "epoch": 3.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7570 + }, + { + "epoch": 3.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7571 + }, + { + "epoch": 3.0251697962445068, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7572 + }, + { + "epoch": 3.025569316819816, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7573 + }, + { + "epoch": 3.025968837395126, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7574 + }, + { + "epoch": 3.0263683579704357, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7575 + }, + { + "epoch": 3.026767878545745, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7576 + }, + { + "epoch": 3.027167399121055, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7577 + }, + { + "epoch": 3.027566919696364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7578 + }, + { + "epoch": 3.027966440271674, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7579 + }, + { + "epoch": 3.0283659608469837, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7580 + }, + { + "epoch": 3.028765481422293, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7581 + }, + { + "epoch": 3.029165001997603, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7582 + }, + { + "epoch": 3.0295645225729126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 7583 + }, + { + "epoch": 3.029964043148222, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7584 + }, + { + "epoch": 3.0303635637235318, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7585 + }, + { + "epoch": 3.0307630842988416, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7586 + }, + { + "epoch": 3.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7587 + }, + { + "epoch": 3.0315621254494607, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7588 + }, + { + "epoch": 3.0319616460247705, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7589 + }, + { + "epoch": 3.03236116660008, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7590 + }, + { + "epoch": 3.0327606871753896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 7591 + }, + { + "epoch": 3.033160207750699, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 7592 + }, + { + "epoch": 3.0335597283260087, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7593 + }, + { + "epoch": 3.0339592489013185, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7594 + }, + { + "epoch": 3.034358769476628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7595 + }, + { + "epoch": 3.0347582900519376, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7596 + }, + { + "epoch": 3.0351578106272474, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7597 + }, + { + "epoch": 3.0355573312025568, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 7598 + }, + { + "epoch": 3.0359568517778666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7599 + }, + { + "epoch": 3.0363563723531763, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7600 + }, + { + "epoch": 3.0367558929284857, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 7601 + }, + { + "epoch": 3.0371554135037955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7602 + }, + { + "epoch": 3.0375549340791053, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7603 + }, + { + "epoch": 3.0379544546544146, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7604 + }, + { + "epoch": 3.0383539752297244, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7605 + }, + { + "epoch": 3.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7606 + }, + { + "epoch": 3.0391530163803435, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7607 + }, + { + "epoch": 3.0395525369556533, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7608 + }, + { + "epoch": 3.0399520575309626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7609 + }, + { + "epoch": 3.0403515781062724, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7610 + }, + { + "epoch": 3.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7611 + }, + { + "epoch": 3.0411506192568916, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7612 + }, + { + "epoch": 3.0415501398322013, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7613 + }, + { + "epoch": 3.041949660407511, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7614 + }, + { + "epoch": 3.0423491809828205, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7615 + }, + { + "epoch": 3.0427487015581303, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7616 + }, + { + "epoch": 3.04314822213344, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7617 + }, + { + "epoch": 3.0435477427087494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7618 + }, + { + "epoch": 3.043947263284059, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7619 + }, + { + "epoch": 3.044346783859369, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7620 + }, + { + "epoch": 3.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7621 + }, + { + "epoch": 3.045145825009988, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7622 + }, + { + "epoch": 3.0455453455852974, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7623 + }, + { + "epoch": 3.0459448661606072, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7624 + }, + { + "epoch": 3.046344386735917, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 7625 + }, + { + "epoch": 3.0467439073112264, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7626 + }, + { + "epoch": 3.047143427886536, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7627 + }, + { + "epoch": 3.047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7628 + }, + { + "epoch": 3.0479424690371553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7629 + }, + { + "epoch": 3.048341989612465, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 7630 + }, + { + "epoch": 3.048741510187775, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7631 + }, + { + "epoch": 3.049141030763084, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7632 + }, + { + "epoch": 3.049540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7633 + }, + { + "epoch": 3.0499400719137038, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 7634 + }, + { + "epoch": 3.050339592489013, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7635 + }, + { + "epoch": 3.050739113064323, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7636 + }, + { + "epoch": 3.0511386336396322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7637 + }, + { + "epoch": 3.051538154214942, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7638 + }, + { + "epoch": 3.051937674790252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7639 + }, + { + "epoch": 3.052337195365561, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7640 + }, + { + "epoch": 3.052736715940871, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7641 + }, + { + "epoch": 3.0531362365161807, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7642 + }, + { + "epoch": 3.05353575709149, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7643 + }, + { + "epoch": 3.0539352776668, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7644 + }, + { + "epoch": 3.0543347982421096, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7645 + }, + { + "epoch": 3.054734318817419, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7646 + }, + { + "epoch": 3.0551338393927288, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7647 + }, + { + "epoch": 3.0555333599680385, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7648 + }, + { + "epoch": 3.055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7649 + }, + { + "epoch": 3.0563324011186577, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 7650 + }, + { + "epoch": 3.0567319216939675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7651 + }, + { + "epoch": 3.057131442269277, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7652 + }, + { + "epoch": 3.0575309628445866, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7653 + }, + { + "epoch": 3.057930483419896, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7654 + }, + { + "epoch": 3.0583300039952057, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7655 + }, + { + "epoch": 3.0587295245705155, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7656 + }, + { + "epoch": 3.059129045145825, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7657 + }, + { + "epoch": 3.0595285657211346, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7658 + }, + { + "epoch": 3.0599280862964444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7659 + }, + { + "epoch": 3.0603276068717538, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7660 + }, + { + "epoch": 3.0607271274470635, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7661 + }, + { + "epoch": 3.0611266480223733, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7662 + }, + { + "epoch": 3.0615261685976827, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7663 + }, + { + "epoch": 3.0619256891729925, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7664 + }, + { + "epoch": 3.0623252097483022, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7665 + }, + { + "epoch": 3.0627247303236116, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7666 + }, + { + "epoch": 3.0631242508989214, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7667 + }, + { + "epoch": 3.0635237714742307, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7668 + }, + { + "epoch": 3.0639232920495405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7669 + }, + { + "epoch": 3.0643228126248503, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7670 + }, + { + "epoch": 3.0647223332001596, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 7671 + }, + { + "epoch": 3.0651218537754694, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 7672 + }, + { + "epoch": 3.065521374350779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7673 + }, + { + "epoch": 3.0659208949260885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7674 + }, + { + "epoch": 3.0663204155013983, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7675 + }, + { + "epoch": 3.066719936076708, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7676 + }, + { + "epoch": 3.0671194566520175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7677 + }, + { + "epoch": 3.0675189772273272, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7678 + }, + { + "epoch": 3.067918497802637, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7679 + }, + { + "epoch": 3.0683180183779464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7680 + }, + { + "epoch": 3.068717538953256, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7681 + }, + { + "epoch": 3.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7682 + }, + { + "epoch": 3.0695165801038753, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7683 + }, + { + "epoch": 3.069916100679185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7684 + }, + { + "epoch": 3.0703156212544944, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7685 + }, + { + "epoch": 3.070715141829804, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7686 + }, + { + "epoch": 3.071114662405114, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7687 + }, + { + "epoch": 3.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7688 + }, + { + "epoch": 3.071913703555733, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 7689 + }, + { + "epoch": 3.072313224131043, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7690 + }, + { + "epoch": 3.0727127447063523, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7691 + }, + { + "epoch": 3.073112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7692 + }, + { + "epoch": 3.073511785856972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7693 + }, + { + "epoch": 3.073911306432281, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7694 + }, + { + "epoch": 3.074310827007591, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 7695 + }, + { + "epoch": 3.0747103475829007, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7696 + }, + { + "epoch": 3.07510986815821, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7697 + }, + { + "epoch": 3.07550938873352, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7698 + }, + { + "epoch": 3.075908909308829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7699 + }, + { + "epoch": 3.076308429884139, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7700 + }, + { + "epoch": 3.076707950459449, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7701 + }, + { + "epoch": 3.077107471034758, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7702 + }, + { + "epoch": 3.077506991610068, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7703 + }, + { + "epoch": 3.0779065121853777, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7704 + }, + { + "epoch": 3.078306032760687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7705 + }, + { + "epoch": 3.078705553335997, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7706 + }, + { + "epoch": 3.0791050739113066, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7707 + }, + { + "epoch": 3.079504594486616, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7708 + }, + { + "epoch": 3.0799041150619257, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7709 + }, + { + "epoch": 3.0803036356372355, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7710 + }, + { + "epoch": 3.080703156212545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7711 + }, + { + "epoch": 3.0811026767878547, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7712 + }, + { + "epoch": 3.081502197363164, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7713 + }, + { + "epoch": 3.081901717938474, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7714 + }, + { + "epoch": 3.0823012385137836, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7715 + }, + { + "epoch": 3.082700759089093, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7716 + }, + { + "epoch": 3.0831002796644027, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7717 + }, + { + "epoch": 3.0834998002397125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7718 + }, + { + "epoch": 3.083899320815022, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7719 + }, + { + "epoch": 3.0842988413903316, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7720 + }, + { + "epoch": 3.0846983619656414, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7721 + }, + { + "epoch": 3.0850978825409507, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7722 + }, + { + "epoch": 3.0854974031162605, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7723 + }, + { + "epoch": 3.0858969236915703, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 7724 + }, + { + "epoch": 3.0862964442668797, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7725 + }, + { + "epoch": 3.0866959648421894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7726 + }, + { + "epoch": 3.087095485417499, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7727 + }, + { + "epoch": 3.0874950059928086, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7728 + }, + { + "epoch": 3.0878945265681184, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7729 + }, + { + "epoch": 3.0882940471434277, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7730 + }, + { + "epoch": 3.0886935677187375, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7731 + }, + { + "epoch": 3.0890930882940473, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7732 + }, + { + "epoch": 3.0894926088693566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7733 + }, + { + "epoch": 3.0898921294446664, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7734 + }, + { + "epoch": 3.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7735 + }, + { + "epoch": 3.0906911705952855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7736 + }, + { + "epoch": 3.0910906911705953, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7737 + }, + { + "epoch": 3.091490211745905, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7738 + }, + { + "epoch": 3.0918897323212144, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7739 + }, + { + "epoch": 3.0922892528965242, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7740 + }, + { + "epoch": 3.092688773471834, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7741 + }, + { + "epoch": 3.0930882940471434, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7742 + }, + { + "epoch": 3.093487814622453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7743 + }, + { + "epoch": 3.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7744 + }, + { + "epoch": 3.0942868557730723, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7745 + }, + { + "epoch": 3.094686376348382, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7746 + }, + { + "epoch": 3.0950858969236914, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7747 + }, + { + "epoch": 3.095485417499001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 7748 + }, + { + "epoch": 3.095884938074311, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7749 + }, + { + "epoch": 3.0962844586496203, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7750 + }, + { + "epoch": 3.09668397922493, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7751 + }, + { + "epoch": 3.09708349980024, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7752 + }, + { + "epoch": 3.0974830203755492, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7753 + }, + { + "epoch": 3.097882540950859, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 7754 + }, + { + "epoch": 3.098282061526169, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7755 + }, + { + "epoch": 3.098681582101478, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7756 + }, + { + "epoch": 3.099081102676788, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7757 + }, + { + "epoch": 3.0994806232520973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7758 + }, + { + "epoch": 3.099880143827407, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7759 + }, + { + "epoch": 3.100279664402717, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7760 + }, + { + "epoch": 3.100679184978026, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7761 + }, + { + "epoch": 3.101078705553336, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7762 + }, + { + "epoch": 3.1014782261286458, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7763 + }, + { + "epoch": 3.101877746703955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7764 + }, + { + "epoch": 3.102277267279265, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7765 + }, + { + "epoch": 3.1026767878545747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 7766 + }, + { + "epoch": 3.103076308429884, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7767 + }, + { + "epoch": 3.103475829005194, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7768 + }, + { + "epoch": 3.1038753495805036, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7769 + }, + { + "epoch": 3.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7770 + }, + { + "epoch": 3.1046743907311227, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7771 + }, + { + "epoch": 3.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7772 + }, + { + "epoch": 3.105473431881742, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7773 + }, + { + "epoch": 3.1058729524570516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7774 + }, + { + "epoch": 3.106272473032361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7775 + }, + { + "epoch": 3.1066719936076708, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7776 + }, + { + "epoch": 3.1070715141829806, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7777 + }, + { + "epoch": 3.10747103475829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7778 + }, + { + "epoch": 3.1078705553335997, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7779 + }, + { + "epoch": 3.1082700759089095, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7780 + }, + { + "epoch": 3.108669596484219, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7781 + }, + { + "epoch": 3.1090691170595286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 7782 + }, + { + "epoch": 3.1094686376348384, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7783 + }, + { + "epoch": 3.1098681582101477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7784 + }, + { + "epoch": 3.1102676787854575, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 7785 + }, + { + "epoch": 3.1106671993607673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7786 + }, + { + "epoch": 3.1110667199360766, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7787 + }, + { + "epoch": 3.1114662405113864, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7788 + }, + { + "epoch": 3.1118657610866958, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7789 + }, + { + "epoch": 3.1122652816620056, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7790 + }, + { + "epoch": 3.1126648022373153, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7791 + }, + { + "epoch": 3.1130643228126247, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7792 + }, + { + "epoch": 3.1134638433879345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 7793 + }, + { + "epoch": 3.1138633639632443, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7794 + }, + { + "epoch": 3.1142628845385536, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7795 + }, + { + "epoch": 3.1146624051138634, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7796 + }, + { + "epoch": 3.115061925689173, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7797 + }, + { + "epoch": 3.1154614462644825, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7798 + }, + { + "epoch": 3.1158609668397923, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7799 + }, + { + "epoch": 3.116260487415102, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7800 + }, + { + "epoch": 3.1166600079904114, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7801 + }, + { + "epoch": 3.117059528565721, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 7802 + }, + { + "epoch": 3.1174590491410306, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7803 + }, + { + "epoch": 3.1178585697163403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7804 + }, + { + "epoch": 3.11825809029165, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7805 + }, + { + "epoch": 3.1186576108669595, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 7806 + }, + { + "epoch": 3.1190571314422693, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7807 + }, + { + "epoch": 3.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7808 + }, + { + "epoch": 3.1198561725928884, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7809 + }, + { + "epoch": 3.120255693168198, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7810 + }, + { + "epoch": 3.120655213743508, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7811 + }, + { + "epoch": 3.1210547343188173, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7812 + }, + { + "epoch": 3.121454254894127, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7813 + }, + { + "epoch": 3.121853775469437, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7814 + }, + { + "epoch": 3.122253296044746, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7815 + }, + { + "epoch": 3.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7816 + }, + { + "epoch": 3.1230523371953653, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7817 + }, + { + "epoch": 3.123451857770675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7818 + }, + { + "epoch": 3.123851378345985, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7819 + }, + { + "epoch": 3.1242508989212943, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7820 + }, + { + "epoch": 3.124650419496604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7821 + }, + { + "epoch": 3.125049940071914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7822 + }, + { + "epoch": 3.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7823 + }, + { + "epoch": 3.125848981222533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7824 + }, + { + "epoch": 3.1262485017978427, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7825 + }, + { + "epoch": 3.126648022373152, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7826 + }, + { + "epoch": 3.127047542948462, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7827 + }, + { + "epoch": 3.1274470635237717, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7828 + }, + { + "epoch": 3.127846584099081, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7829 + }, + { + "epoch": 3.128246104674391, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7830 + }, + { + "epoch": 3.1286456252497006, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7831 + }, + { + "epoch": 3.12904514582501, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7832 + }, + { + "epoch": 3.1294446664003197, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7833 + }, + { + "epoch": 3.129844186975629, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7834 + }, + { + "epoch": 3.130243707550939, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7835 + }, + { + "epoch": 3.1306432281262486, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7836 + }, + { + "epoch": 3.131042748701558, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7837 + }, + { + "epoch": 3.1314422692768678, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7838 + }, + { + "epoch": 3.1318417898521775, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7839 + }, + { + "epoch": 3.132241310427487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7840 + }, + { + "epoch": 3.1326408310027967, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7841 + }, + { + "epoch": 3.1330403515781065, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7842 + }, + { + "epoch": 3.133439872153416, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7843 + }, + { + "epoch": 3.1338393927287256, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7844 + }, + { + "epoch": 3.134238913304035, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7845 + }, + { + "epoch": 3.1346384338793447, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7846 + }, + { + "epoch": 3.1350379544546545, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7847 + }, + { + "epoch": 3.135437475029964, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7848 + }, + { + "epoch": 3.1358369956052736, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7849 + }, + { + "epoch": 3.1362365161805834, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7850 + }, + { + "epoch": 3.1366360367558928, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7851 + }, + { + "epoch": 3.1370355573312025, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7852 + }, + { + "epoch": 3.1374350779065123, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7853 + }, + { + "epoch": 3.1378345984818217, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7854 + }, + { + "epoch": 3.1382341190571315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7855 + }, + { + "epoch": 3.1386336396324412, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7856 + }, + { + "epoch": 3.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7857 + }, + { + "epoch": 3.1394326807830604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7858 + }, + { + "epoch": 3.13983220135837, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7859 + }, + { + "epoch": 3.1402317219336795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7860 + }, + { + "epoch": 3.1406312425089893, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7861 + }, + { + "epoch": 3.1410307630842986, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7862 + }, + { + "epoch": 3.1414302836596084, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7863 + }, + { + "epoch": 3.141829804234918, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7864 + }, + { + "epoch": 3.1422293248102275, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7865 + }, + { + "epoch": 3.1426288453855373, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7866 + }, + { + "epoch": 3.143028365960847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7867 + }, + { + "epoch": 3.1434278865361565, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7868 + }, + { + "epoch": 3.1438274071114662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7869 + }, + { + "epoch": 3.144226927686776, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7870 + }, + { + "epoch": 3.1446264482620854, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7871 + }, + { + "epoch": 3.145025968837395, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7872 + }, + { + "epoch": 3.145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7873 + }, + { + "epoch": 3.1458250099880143, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 7874 + }, + { + "epoch": 3.146224530563324, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7875 + }, + { + "epoch": 3.146624051138634, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7876 + }, + { + "epoch": 3.147023571713943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7877 + }, + { + "epoch": 3.147423092289253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7878 + }, + { + "epoch": 3.1478226128645623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7879 + }, + { + "epoch": 3.148222133439872, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7880 + }, + { + "epoch": 3.148621654015182, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7881 + }, + { + "epoch": 3.1490211745904912, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7882 + }, + { + "epoch": 3.149420695165801, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7883 + }, + { + "epoch": 3.149820215741111, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7884 + }, + { + "epoch": 3.15021973631642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7885 + }, + { + "epoch": 3.15061925689173, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7886 + }, + { + "epoch": 3.1510187774670397, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 7887 + }, + { + "epoch": 3.151418298042349, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7888 + }, + { + "epoch": 3.151817818617659, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7889 + }, + { + "epoch": 3.152217339192968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7890 + }, + { + "epoch": 3.152616859768278, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7891 + }, + { + "epoch": 3.1530163803435878, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7892 + }, + { + "epoch": 3.153415900918897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7893 + }, + { + "epoch": 3.153815421494207, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7894 + }, + { + "epoch": 3.1542149420695167, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7895 + }, + { + "epoch": 3.154614462644826, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7896 + }, + { + "epoch": 3.155013983220136, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7897 + }, + { + "epoch": 3.1554135037954456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7898 + }, + { + "epoch": 3.155813024370755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7899 + }, + { + "epoch": 3.1562125449460647, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7900 + }, + { + "epoch": 3.1566120655213745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7901 + }, + { + "epoch": 3.157011586096684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7902 + }, + { + "epoch": 3.1574111066719937, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7903 + }, + { + "epoch": 3.1578106272473034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7904 + }, + { + "epoch": 3.158210147822613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7905 + }, + { + "epoch": 3.1586096683979226, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7906 + }, + { + "epoch": 3.159009188973232, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7907 + }, + { + "epoch": 3.1594087095485417, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7908 + }, + { + "epoch": 3.1598082301238515, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 7909 + }, + { + "epoch": 3.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7910 + }, + { + "epoch": 3.1606072712744706, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7911 + }, + { + "epoch": 3.1610067918497804, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7912 + }, + { + "epoch": 3.1614063124250897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7913 + }, + { + "epoch": 3.1618058330003995, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 7914 + }, + { + "epoch": 3.1622053535757093, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7915 + }, + { + "epoch": 3.1626048741510187, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7916 + }, + { + "epoch": 3.1630043947263284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 7917 + }, + { + "epoch": 3.1634039153016382, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7918 + }, + { + "epoch": 3.1638034358769476, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7919 + }, + { + "epoch": 3.1642029564522574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7920 + }, + { + "epoch": 3.164602477027567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7921 + }, + { + "epoch": 3.1650019976028765, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7922 + }, + { + "epoch": 3.1654015181781863, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7923 + }, + { + "epoch": 3.1658010387534956, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7924 + }, + { + "epoch": 3.1662005593288054, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 7925 + }, + { + "epoch": 3.166600079904115, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7926 + }, + { + "epoch": 3.1669996004794245, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7927 + }, + { + "epoch": 3.1673991210547343, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7928 + }, + { + "epoch": 3.167798641630044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7929 + }, + { + "epoch": 3.1681981622053534, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 7930 + }, + { + "epoch": 3.1685976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7931 + }, + { + "epoch": 3.168997203355973, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7932 + }, + { + "epoch": 3.1693967239312824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7933 + }, + { + "epoch": 3.169796244506592, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7934 + }, + { + "epoch": 3.1701957650819015, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7935 + }, + { + "epoch": 3.1705952856572113, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7936 + }, + { + "epoch": 3.170994806232521, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7937 + }, + { + "epoch": 3.1713943268078304, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7938 + }, + { + "epoch": 3.17179384738314, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7939 + }, + { + "epoch": 3.17219336795845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 7940 + }, + { + "epoch": 3.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7941 + }, + { + "epoch": 3.172992409109069, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7942 + }, + { + "epoch": 3.173391929684379, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7943 + }, + { + "epoch": 3.1737914502596882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7944 + }, + { + "epoch": 3.174190970834998, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7945 + }, + { + "epoch": 3.174590491410308, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7946 + }, + { + "epoch": 3.174990011985617, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 7947 + }, + { + "epoch": 3.175389532560927, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7948 + }, + { + "epoch": 3.1757890531362367, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7949 + }, + { + "epoch": 3.176188573711546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 7950 + }, + { + "epoch": 3.176588094286856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7951 + }, + { + "epoch": 3.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7952 + }, + { + "epoch": 3.177387135437475, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7953 + }, + { + "epoch": 3.1777866560127848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7954 + }, + { + "epoch": 3.178186176588094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 7955 + }, + { + "epoch": 3.178585697163404, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7956 + }, + { + "epoch": 3.1789852177387137, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7957 + }, + { + "epoch": 3.179384738314023, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7958 + }, + { + "epoch": 3.179784258889333, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7959 + }, + { + "epoch": 3.1801837794646426, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 7960 + }, + { + "epoch": 3.180583300039952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 7961 + }, + { + "epoch": 3.1809828206152617, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7962 + }, + { + "epoch": 3.1813823411905715, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7963 + }, + { + "epoch": 3.181781861765881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7964 + }, + { + "epoch": 3.1821813823411906, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7965 + }, + { + "epoch": 3.1825809029165004, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7966 + }, + { + "epoch": 3.1829804234918098, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7967 + }, + { + "epoch": 3.1833799440671195, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7968 + }, + { + "epoch": 3.183779464642429, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7969 + }, + { + "epoch": 3.1841789852177387, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7970 + }, + { + "epoch": 3.1845785057930485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7971 + }, + { + "epoch": 3.184978026368358, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7972 + }, + { + "epoch": 3.1853775469436676, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7973 + }, + { + "epoch": 3.1857770675189774, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 7974 + }, + { + "epoch": 3.1861765880942867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7975 + }, + { + "epoch": 3.1865761086695965, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7976 + }, + { + "epoch": 3.1869756292449063, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7977 + }, + { + "epoch": 3.1873751498202156, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7978 + }, + { + "epoch": 3.1877746703955254, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7979 + }, + { + "epoch": 3.1881741909708348, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7980 + }, + { + "epoch": 3.1885737115461446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7981 + }, + { + "epoch": 3.1889732321214543, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7982 + }, + { + "epoch": 3.1893727526967637, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7983 + }, + { + "epoch": 3.1897722732720735, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7984 + }, + { + "epoch": 3.1901717938473833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7985 + }, + { + "epoch": 3.1905713144226926, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7986 + }, + { + "epoch": 3.1909708349980024, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7987 + }, + { + "epoch": 3.191370355573312, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7988 + }, + { + "epoch": 3.1917698761486215, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7989 + }, + { + "epoch": 3.1921693967239313, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7990 + }, + { + "epoch": 3.192568917299241, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7991 + }, + { + "epoch": 3.1929684378745504, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 7992 + }, + { + "epoch": 3.19336795844986, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7993 + }, + { + "epoch": 3.19376747902517, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7994 + }, + { + "epoch": 3.1941669996004793, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 7995 + }, + { + "epoch": 3.194566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7996 + }, + { + "epoch": 3.1949660407510985, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7997 + }, + { + "epoch": 3.1953655613264083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7998 + }, + { + "epoch": 3.195765081901718, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7999 + }, + { + "epoch": 3.1961646024770274, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8000 + }, + { + "epoch": 3.196564123052337, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8001 + }, + { + "epoch": 3.196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8002 + }, + { + "epoch": 3.1973631642029563, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8003 + }, + { + "epoch": 3.197762684778266, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8004 + }, + { + "epoch": 3.198162205353576, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8005 + }, + { + "epoch": 3.198561725928885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 8006 + }, + { + "epoch": 3.198961246504195, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8007 + }, + { + "epoch": 3.199360767079505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8008 + }, + { + "epoch": 3.199760287654814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8009 + }, + { + "epoch": 3.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 8010 + }, + { + "epoch": 3.2005593288054337, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8011 + }, + { + "epoch": 3.200958849380743, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8012 + }, + { + "epoch": 3.201358369956053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8013 + }, + { + "epoch": 3.201757890531362, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8014 + }, + { + "epoch": 3.202157411106672, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8015 + }, + { + "epoch": 3.2025569316819817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8016 + }, + { + "epoch": 3.202956452257291, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8017 + }, + { + "epoch": 3.203355972832601, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8018 + }, + { + "epoch": 3.2037554934079107, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8019 + }, + { + "epoch": 3.20415501398322, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 8020 + }, + { + "epoch": 3.20455453455853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8021 + }, + { + "epoch": 3.2049540551338396, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 8022 + }, + { + "epoch": 3.205353575709149, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8023 + }, + { + "epoch": 3.2057530962844587, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8024 + }, + { + "epoch": 3.206152616859768, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8025 + }, + { + "epoch": 3.206552137435078, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8026 + }, + { + "epoch": 3.2069516580103876, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8027 + }, + { + "epoch": 3.207351178585697, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8028 + }, + { + "epoch": 3.2077506991610067, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8029 + }, + { + "epoch": 3.2081502197363165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 8030 + }, + { + "epoch": 3.208549740311626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8031 + }, + { + "epoch": 3.2089492608869357, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8032 + }, + { + "epoch": 3.2093487814622454, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8033 + }, + { + "epoch": 3.209748302037555, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8034 + }, + { + "epoch": 3.2101478226128646, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8035 + }, + { + "epoch": 3.2105473431881744, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8036 + }, + { + "epoch": 3.2109468637634837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 8037 + }, + { + "epoch": 3.2113463843387935, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8038 + }, + { + "epoch": 3.2117459049141033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8039 + }, + { + "epoch": 3.2121454254894126, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8040 + }, + { + "epoch": 3.2125449460647224, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8041 + }, + { + "epoch": 3.2129444666400317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8042 + }, + { + "epoch": 3.2133439872153415, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8043 + }, + { + "epoch": 3.2137435077906513, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8044 + }, + { + "epoch": 3.2141430283659607, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8045 + }, + { + "epoch": 3.2145425489412705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8046 + }, + { + "epoch": 3.2149420695165802, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8047 + }, + { + "epoch": 3.2153415900918896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8048 + }, + { + "epoch": 3.2157411106671994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8049 + }, + { + "epoch": 3.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8050 + }, + { + "epoch": 3.2165401518178185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8051 + }, + { + "epoch": 3.2169396723931283, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8052 + }, + { + "epoch": 3.217339192968438, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8053 + }, + { + "epoch": 3.2177387135437474, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8054 + }, + { + "epoch": 3.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8055 + }, + { + "epoch": 3.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8056 + }, + { + "epoch": 3.2189372752696763, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8057 + }, + { + "epoch": 3.219336795844986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8058 + }, + { + "epoch": 3.2197363164202955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8059 + }, + { + "epoch": 3.2201358369956052, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 8060 + }, + { + "epoch": 3.220535357570915, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8061 + }, + { + "epoch": 3.2209348781462244, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8062 + }, + { + "epoch": 3.221334398721534, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8063 + }, + { + "epoch": 3.221733919296844, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8064 + }, + { + "epoch": 3.2221334398721533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8065 + }, + { + "epoch": 3.222532960447463, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 8066 + }, + { + "epoch": 3.222932481022773, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8067 + }, + { + "epoch": 3.223332001598082, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 8068 + }, + { + "epoch": 3.223731522173392, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8069 + }, + { + "epoch": 3.2241310427487013, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8070 + }, + { + "epoch": 3.224530563324011, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8071 + }, + { + "epoch": 3.224930083899321, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8072 + }, + { + "epoch": 3.2253296044746302, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8073 + }, + { + "epoch": 3.22572912504994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8074 + }, + { + "epoch": 3.22612864562525, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8075 + }, + { + "epoch": 3.226528166200559, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8076 + }, + { + "epoch": 3.226927686775869, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8077 + }, + { + "epoch": 3.2273272073511787, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8078 + }, + { + "epoch": 3.227726727926488, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8079 + }, + { + "epoch": 3.228126248501798, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8080 + }, + { + "epoch": 3.2285257690771076, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8081 + }, + { + "epoch": 3.228925289652417, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8082 + }, + { + "epoch": 3.2293248102277268, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8083 + }, + { + "epoch": 3.2297243308030366, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8084 + }, + { + "epoch": 3.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8085 + }, + { + "epoch": 3.2305233719536557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8086 + }, + { + "epoch": 3.230922892528965, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8087 + }, + { + "epoch": 3.231322413104275, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8088 + }, + { + "epoch": 3.2317219336795846, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8089 + }, + { + "epoch": 3.232121454254894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8090 + }, + { + "epoch": 3.2325209748302037, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8091 + }, + { + "epoch": 3.2329204954055135, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8092 + }, + { + "epoch": 3.233320015980823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8093 + }, + { + "epoch": 3.2337195365561326, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8094 + }, + { + "epoch": 3.2341190571314424, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8095 + }, + { + "epoch": 3.2345185777067518, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8096 + }, + { + "epoch": 3.2349180982820616, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 8097 + }, + { + "epoch": 3.2353176188573713, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8098 + }, + { + "epoch": 3.2357171394326807, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8099 + }, + { + "epoch": 3.2361166600079905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 8100 + }, + { + "epoch": 3.2365161805833003, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8101 + }, + { + "epoch": 3.2369157011586096, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 8102 + }, + { + "epoch": 3.2373152217339194, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8103 + }, + { + "epoch": 3.2377147423092287, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8104 + }, + { + "epoch": 3.2381142628845385, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8105 + }, + { + "epoch": 3.2385137834598483, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8106 + }, + { + "epoch": 3.2389133040351576, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 8107 + }, + { + "epoch": 3.2393128246104674, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8108 + }, + { + "epoch": 3.239712345185777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8109 + }, + { + "epoch": 3.2401118657610866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8110 + }, + { + "epoch": 3.2405113863363963, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 8111 + }, + { + "epoch": 3.240910906911706, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8112 + }, + { + "epoch": 3.2413104274870155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8113 + }, + { + "epoch": 3.2417099480623253, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8114 + }, + { + "epoch": 3.2421094686376346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8115 + }, + { + "epoch": 3.2425089892129444, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8116 + }, + { + "epoch": 3.242908509788254, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8117 + }, + { + "epoch": 3.2433080303635635, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8118 + }, + { + "epoch": 3.2437075509388733, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 8119 + }, + { + "epoch": 3.244107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8120 + }, + { + "epoch": 3.2445065920894924, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8121 + }, + { + "epoch": 3.2449061126648022, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8122 + }, + { + "epoch": 3.245305633240112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8123 + }, + { + "epoch": 3.2457051538154214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8124 + }, + { + "epoch": 3.246104674390731, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8125 + }, + { + "epoch": 3.246504194966041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8126 + }, + { + "epoch": 3.2469037155413503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8127 + }, + { + "epoch": 3.24730323611666, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8128 + }, + { + "epoch": 3.24770275669197, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8129 + }, + { + "epoch": 3.248102277267279, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8130 + }, + { + "epoch": 3.248501797842589, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8131 + }, + { + "epoch": 3.2489013184178983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8132 + }, + { + "epoch": 3.249300838993208, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8133 + }, + { + "epoch": 3.249700359568518, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8134 + }, + { + "epoch": 3.2500998801438272, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8135 + }, + { + "epoch": 3.250499400719137, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8136 + }, + { + "epoch": 3.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8137 + }, + { + "epoch": 3.251298441869756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8138 + }, + { + "epoch": 3.251697962445066, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 8139 + }, + { + "epoch": 3.2520974830203757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8140 + }, + { + "epoch": 3.252497003595685, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8141 + }, + { + "epoch": 3.252896524170995, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 8142 + }, + { + "epoch": 3.2532960447463046, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8143 + }, + { + "epoch": 3.253695565321614, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8144 + }, + { + "epoch": 3.2540950858969238, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8145 + }, + { + "epoch": 3.2544946064722335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8146 + }, + { + "epoch": 3.254894127047543, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8147 + }, + { + "epoch": 3.2552936476228527, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8148 + }, + { + "epoch": 3.255693168198162, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8149 + }, + { + "epoch": 3.256092688773472, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8150 + }, + { + "epoch": 3.2564922093487816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8151 + }, + { + "epoch": 3.256891729924091, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8152 + }, + { + "epoch": 3.2572912504994007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8153 + }, + { + "epoch": 3.2576907710747105, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8154 + }, + { + "epoch": 3.25809029165002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8155 + }, + { + "epoch": 3.2584898122253296, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8156 + }, + { + "epoch": 3.2588893328006394, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8157 + }, + { + "epoch": 3.2592888533759488, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8158 + }, + { + "epoch": 3.2596883739512585, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8159 + }, + { + "epoch": 3.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8160 + }, + { + "epoch": 3.2604874151018777, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8161 + }, + { + "epoch": 3.2608869356771875, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8162 + }, + { + "epoch": 3.2612864562524972, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8163 + }, + { + "epoch": 3.2616859768278066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8164 + }, + { + "epoch": 3.2620854974031164, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 8165 + }, + { + "epoch": 3.2624850179784257, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8166 + }, + { + "epoch": 3.2628845385537355, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8167 + }, + { + "epoch": 3.2632840591290453, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8168 + }, + { + "epoch": 3.2636835797043546, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8169 + }, + { + "epoch": 3.2640831002796644, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8170 + }, + { + "epoch": 3.264482620854974, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8171 + }, + { + "epoch": 3.2648821414302835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8172 + }, + { + "epoch": 3.2652816620055933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8173 + }, + { + "epoch": 3.265681182580903, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8174 + }, + { + "epoch": 3.2660807031562125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8175 + }, + { + "epoch": 3.2664802237315222, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8176 + }, + { + "epoch": 3.2668797443068316, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8177 + }, + { + "epoch": 3.2672792648821414, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8178 + }, + { + "epoch": 3.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8179 + }, + { + "epoch": 3.2680783060327605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8180 + }, + { + "epoch": 3.2684778266080703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8181 + }, + { + "epoch": 3.26887734718338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8182 + }, + { + "epoch": 3.2692768677586894, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8183 + }, + { + "epoch": 3.269676388333999, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8184 + }, + { + "epoch": 3.270075908909309, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8185 + }, + { + "epoch": 3.2704754294846183, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8186 + }, + { + "epoch": 3.270874950059928, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 8187 + }, + { + "epoch": 3.271274470635238, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8188 + }, + { + "epoch": 3.2716739912105472, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8189 + }, + { + "epoch": 3.272073511785857, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8190 + }, + { + "epoch": 3.272473032361167, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8191 + }, + { + "epoch": 3.272872552936476, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 8192 + }, + { + "epoch": 3.273272073511786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8193 + }, + { + "epoch": 3.2736715940870953, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8194 + }, + { + "epoch": 3.274071114662405, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8195 + }, + { + "epoch": 3.274470635237715, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8196 + }, + { + "epoch": 3.274870155813024, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 8197 + }, + { + "epoch": 3.275269676388334, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8198 + }, + { + "epoch": 3.275669196963644, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8199 + }, + { + "epoch": 3.276068717538953, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8200 + }, + { + "epoch": 3.276468238114263, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8201 + }, + { + "epoch": 3.2768677586895727, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8202 + }, + { + "epoch": 3.277267279264882, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8203 + }, + { + "epoch": 3.277666799840192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8204 + }, + { + "epoch": 3.278066320415501, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8205 + }, + { + "epoch": 3.278465840990811, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8206 + }, + { + "epoch": 3.2788653615661207, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8207 + }, + { + "epoch": 3.2792648821414305, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8208 + }, + { + "epoch": 3.27966440271674, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8209 + }, + { + "epoch": 3.2800639232920497, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8210 + }, + { + "epoch": 3.280463443867359, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 8211 + }, + { + "epoch": 3.280862964442669, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8212 + }, + { + "epoch": 3.2812624850179786, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8213 + }, + { + "epoch": 3.281662005593288, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 8214 + }, + { + "epoch": 3.2820615261685977, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8215 + }, + { + "epoch": 3.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8216 + }, + { + "epoch": 3.282860567319217, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8217 + }, + { + "epoch": 3.2832600878945266, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8218 + }, + { + "epoch": 3.2836596084698364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8219 + }, + { + "epoch": 3.2840591290451457, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8220 + }, + { + "epoch": 3.2844586496204555, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8221 + }, + { + "epoch": 3.284858170195765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8222 + }, + { + "epoch": 3.2852576907710747, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8223 + }, + { + "epoch": 3.2856572113463844, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8224 + }, + { + "epoch": 3.286056731921694, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8225 + }, + { + "epoch": 3.2864562524970036, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8226 + }, + { + "epoch": 3.2868557730723134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8227 + }, + { + "epoch": 3.2872552936476227, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 8228 + }, + { + "epoch": 3.2876548142229325, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8229 + }, + { + "epoch": 3.2880543347982423, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8230 + }, + { + "epoch": 3.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8231 + }, + { + "epoch": 3.2888533759488614, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8232 + }, + { + "epoch": 3.289252896524171, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8233 + }, + { + "epoch": 3.2896524170994805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8234 + }, + { + "epoch": 3.2900519376747903, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8235 + }, + { + "epoch": 3.2904514582501, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8236 + }, + { + "epoch": 3.2908509788254094, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8237 + }, + { + "epoch": 3.2912504994007192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8238 + }, + { + "epoch": 3.2916500199760286, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8239 + }, + { + "epoch": 3.2920495405513384, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8240 + }, + { + "epoch": 3.292449061126648, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8241 + }, + { + "epoch": 3.2928485817019575, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8242 + }, + { + "epoch": 3.2932481022772673, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8243 + }, + { + "epoch": 3.293647622852577, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8244 + }, + { + "epoch": 3.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8245 + }, + { + "epoch": 3.294446664003196, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8246 + }, + { + "epoch": 3.294846184578506, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8247 + }, + { + "epoch": 3.2952457051538153, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8248 + }, + { + "epoch": 3.295645225729125, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8249 + }, + { + "epoch": 3.2960447463044344, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8250 + }, + { + "epoch": 3.2964442668797442, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8251 + }, + { + "epoch": 3.296843787455054, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8252 + }, + { + "epoch": 3.297243308030364, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8253 + }, + { + "epoch": 3.297642828605673, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8254 + }, + { + "epoch": 3.298042349180983, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8255 + }, + { + "epoch": 3.2984418697562923, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8256 + }, + { + "epoch": 3.298841390331602, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8257 + }, + { + "epoch": 3.299240910906912, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8258 + }, + { + "epoch": 3.299640431482221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8259 + }, + { + "epoch": 3.300039952057531, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8260 + }, + { + "epoch": 3.3004394726328408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8261 + }, + { + "epoch": 3.30083899320815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8262 + }, + { + "epoch": 3.30123851378346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8263 + }, + { + "epoch": 3.3016380343587697, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8264 + }, + { + "epoch": 3.302037554934079, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8265 + }, + { + "epoch": 3.302437075509389, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8266 + }, + { + "epoch": 3.302836596084698, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8267 + }, + { + "epoch": 3.303236116660008, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8268 + }, + { + "epoch": 3.3036356372353177, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8269 + }, + { + "epoch": 3.304035157810627, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8270 + }, + { + "epoch": 3.304434678385937, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8271 + }, + { + "epoch": 3.3048341989612466, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8272 + }, + { + "epoch": 3.305233719536556, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8273 + }, + { + "epoch": 3.3056332401118658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8274 + }, + { + "epoch": 3.3060327606871756, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8275 + }, + { + "epoch": 3.306432281262485, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8276 + }, + { + "epoch": 3.3068318018377947, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8277 + }, + { + "epoch": 3.3072313224131045, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8278 + }, + { + "epoch": 3.307630842988414, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 8279 + }, + { + "epoch": 3.3080303635637236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8280 + }, + { + "epoch": 3.3084298841390334, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8281 + }, + { + "epoch": 3.3088294047143427, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8282 + }, + { + "epoch": 3.3092289252896525, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8283 + }, + { + "epoch": 3.309628445864962, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8284 + }, + { + "epoch": 3.3100279664402716, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8285 + }, + { + "epoch": 3.3104274870155814, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8286 + }, + { + "epoch": 3.3108270075908908, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8287 + }, + { + "epoch": 3.3112265281662006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8288 + }, + { + "epoch": 3.3116260487415103, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8289 + }, + { + "epoch": 3.3120255693168197, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8290 + }, + { + "epoch": 3.3124250898921295, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8291 + }, + { + "epoch": 3.3128246104674393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8292 + }, + { + "epoch": 3.3132241310427486, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8293 + }, + { + "epoch": 3.3136236516180584, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8294 + }, + { + "epoch": 3.3140231721933677, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8295 + }, + { + "epoch": 3.3144226927686775, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8296 + }, + { + "epoch": 3.3148222133439873, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8297 + }, + { + "epoch": 3.315221733919297, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8298 + }, + { + "epoch": 3.3156212544946064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8299 + }, + { + "epoch": 3.316020775069916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 8300 + }, + { + "epoch": 3.3164202956452256, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8301 + }, + { + "epoch": 3.3168198162205353, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8302 + }, + { + "epoch": 3.317219336795845, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8303 + }, + { + "epoch": 3.3176188573711545, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8304 + }, + { + "epoch": 3.3180183779464643, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8305 + }, + { + "epoch": 3.318417898521774, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8306 + }, + { + "epoch": 3.3188174190970834, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8307 + }, + { + "epoch": 3.319216939672393, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8308 + }, + { + "epoch": 3.319616460247703, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8309 + }, + { + "epoch": 3.3200159808230123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8310 + }, + { + "epoch": 3.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8311 + }, + { + "epoch": 3.3208150219736314, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8312 + }, + { + "epoch": 3.321214542548941, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8313 + }, + { + "epoch": 3.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8314 + }, + { + "epoch": 3.3220135836995603, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8315 + }, + { + "epoch": 3.32241310427487, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8316 + }, + { + "epoch": 3.32281262485018, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8317 + }, + { + "epoch": 3.3232121454254893, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8318 + }, + { + "epoch": 3.323611666000799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8319 + }, + { + "epoch": 3.324011186576109, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8320 + }, + { + "epoch": 3.324410707151418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8321 + }, + { + "epoch": 3.324810227726728, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8322 + }, + { + "epoch": 3.3252097483020377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8323 + }, + { + "epoch": 3.325609268877347, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8324 + }, + { + "epoch": 3.326008789452657, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8325 + }, + { + "epoch": 3.3264083100279667, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 8326 + }, + { + "epoch": 3.326807830603276, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8327 + }, + { + "epoch": 3.327207351178586, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8328 + }, + { + "epoch": 3.327606871753895, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8329 + }, + { + "epoch": 3.328006392329205, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8330 + }, + { + "epoch": 3.3284059129045147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8331 + }, + { + "epoch": 3.328805433479824, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8332 + }, + { + "epoch": 3.329204954055134, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8333 + }, + { + "epoch": 3.3296044746304436, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8334 + }, + { + "epoch": 3.330003995205753, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8335 + }, + { + "epoch": 3.3304035157810628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8336 + }, + { + "epoch": 3.3308030363563725, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8337 + }, + { + "epoch": 3.331202556931682, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8338 + }, + { + "epoch": 3.3316020775069917, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8339 + }, + { + "epoch": 3.332001598082301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8340 + }, + { + "epoch": 3.332401118657611, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 8341 + }, + { + "epoch": 3.3328006392329206, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8342 + }, + { + "epoch": 3.3332001598082304, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8343 + }, + { + "epoch": 3.3335996803835397, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8344 + }, + { + "epoch": 3.3339992009588495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8345 + }, + { + "epoch": 3.334398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8346 + }, + { + "epoch": 3.3347982421094686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8347 + }, + { + "epoch": 3.3351977626847784, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8348 + }, + { + "epoch": 3.3355972832600878, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8349 + }, + { + "epoch": 3.3359968038353975, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8350 + }, + { + "epoch": 3.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8351 + }, + { + "epoch": 3.3367958449860167, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8352 + }, + { + "epoch": 3.3371953655613265, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8353 + }, + { + "epoch": 3.3375948861366362, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8354 + }, + { + "epoch": 3.3379944067119456, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8355 + }, + { + "epoch": 3.3383939272872554, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8356 + }, + { + "epoch": 3.3387934478625647, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8357 + }, + { + "epoch": 3.3391929684378745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8358 + }, + { + "epoch": 3.3395924890131843, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 8359 + }, + { + "epoch": 3.3399920095884936, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8360 + }, + { + "epoch": 3.3403915301638034, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8361 + }, + { + "epoch": 3.340791050739113, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8362 + }, + { + "epoch": 3.3411905713144225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 8363 + }, + { + "epoch": 3.3415900918897323, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8364 + }, + { + "epoch": 3.341989612465042, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8365 + }, + { + "epoch": 3.3423891330403515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8366 + }, + { + "epoch": 3.3427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8367 + }, + { + "epoch": 3.343188174190971, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8368 + }, + { + "epoch": 3.3435876947662804, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8369 + }, + { + "epoch": 3.34398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8370 + }, + { + "epoch": 3.3443867359169, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8371 + }, + { + "epoch": 3.3447862564922093, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8372 + }, + { + "epoch": 3.345185777067519, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8373 + }, + { + "epoch": 3.3455852976428284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8374 + }, + { + "epoch": 3.345984818218138, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8375 + }, + { + "epoch": 3.346384338793448, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8376 + }, + { + "epoch": 3.3467838593687573, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8377 + }, + { + "epoch": 3.347183379944067, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8378 + }, + { + "epoch": 3.347582900519377, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8379 + }, + { + "epoch": 3.3479824210946862, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8380 + }, + { + "epoch": 3.348381941669996, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8381 + }, + { + "epoch": 3.348781462245306, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8382 + }, + { + "epoch": 3.349180982820615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8383 + }, + { + "epoch": 3.349580503395925, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8384 + }, + { + "epoch": 3.3499800239712343, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8385 + }, + { + "epoch": 3.350379544546544, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8386 + }, + { + "epoch": 3.350779065121854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8387 + }, + { + "epoch": 3.3511785856971636, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8388 + }, + { + "epoch": 3.351578106272473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8389 + }, + { + "epoch": 3.3519776268477828, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8390 + }, + { + "epoch": 3.352377147423092, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8391 + }, + { + "epoch": 3.352776667998402, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8392 + }, + { + "epoch": 3.3531761885737117, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8393 + }, + { + "epoch": 3.353575709149021, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8394 + }, + { + "epoch": 3.353975229724331, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8395 + }, + { + "epoch": 3.3543747502996406, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8396 + }, + { + "epoch": 3.35477427087495, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8397 + }, + { + "epoch": 3.3551737914502597, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8398 + }, + { + "epoch": 3.3555733120255695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8399 + }, + { + "epoch": 3.355972832600879, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 8400 + }, + { + "epoch": 3.3563723531761886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8401 + }, + { + "epoch": 3.356771873751498, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8402 + }, + { + "epoch": 3.357171394326808, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8403 + }, + { + "epoch": 3.3575709149021176, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8404 + }, + { + "epoch": 3.357970435477427, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8405 + }, + { + "epoch": 3.3583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 8406 + }, + { + "epoch": 3.3587694766280465, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8407 + }, + { + "epoch": 3.359168997203356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 8408 + }, + { + "epoch": 3.3595685177786656, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8409 + }, + { + "epoch": 3.3599680383539754, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8410 + }, + { + "epoch": 3.3603675589292847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8411 + }, + { + "epoch": 3.3607670795045945, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8412 + }, + { + "epoch": 3.361166600079904, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 8413 + }, + { + "epoch": 3.3615661206552137, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8414 + }, + { + "epoch": 3.3619656412305234, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8415 + }, + { + "epoch": 3.3623651618058332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 8416 + }, + { + "epoch": 3.3627646823811426, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8417 + }, + { + "epoch": 3.3631642029564524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8418 + }, + { + "epoch": 3.3635637235317617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8419 + }, + { + "epoch": 3.3639632441070715, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 8420 + }, + { + "epoch": 3.3643627646823813, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8421 + }, + { + "epoch": 3.3647622852576906, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8422 + }, + { + "epoch": 3.3651618058330004, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8423 + }, + { + "epoch": 3.36556132640831, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8424 + }, + { + "epoch": 3.3659608469836195, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8425 + }, + { + "epoch": 3.3663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8426 + }, + { + "epoch": 3.366759888134239, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8427 + }, + { + "epoch": 3.3671594087095484, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8428 + }, + { + "epoch": 3.3675589292848582, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8429 + }, + { + "epoch": 3.3679584498601676, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8430 + }, + { + "epoch": 3.3683579704354774, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8431 + }, + { + "epoch": 3.368757491010787, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8432 + }, + { + "epoch": 3.369157011586097, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8433 + }, + { + "epoch": 3.3695565321614063, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8434 + }, + { + "epoch": 3.369956052736716, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8435 + }, + { + "epoch": 3.3703555733120254, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8436 + }, + { + "epoch": 3.370755093887335, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8437 + }, + { + "epoch": 3.371154614462645, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8438 + }, + { + "epoch": 3.3715541350379543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8439 + }, + { + "epoch": 3.371953655613264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8440 + }, + { + "epoch": 3.372353176188574, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8441 + }, + { + "epoch": 3.3727526967638832, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8442 + }, + { + "epoch": 3.373152217339193, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8443 + }, + { + "epoch": 3.373551737914503, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8444 + }, + { + "epoch": 3.373951258489812, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8445 + }, + { + "epoch": 3.374350779065122, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8446 + }, + { + "epoch": 3.3747502996404313, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8447 + }, + { + "epoch": 3.375149820215741, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8448 + }, + { + "epoch": 3.375549340791051, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8449 + }, + { + "epoch": 3.37594886136636, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8450 + }, + { + "epoch": 3.37634838194167, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8451 + }, + { + "epoch": 3.3767479025169798, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8452 + }, + { + "epoch": 3.377147423092289, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 8453 + }, + { + "epoch": 3.377546943667599, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8454 + }, + { + "epoch": 3.3779464642429087, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8455 + }, + { + "epoch": 3.378345984818218, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8456 + }, + { + "epoch": 3.378745505393528, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8457 + }, + { + "epoch": 3.379145025968837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8458 + }, + { + "epoch": 3.379544546544147, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8459 + }, + { + "epoch": 3.3799440671194567, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8460 + }, + { + "epoch": 3.3803435876947665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8461 + }, + { + "epoch": 3.380743108270076, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8462 + }, + { + "epoch": 3.3811426288453856, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 8463 + }, + { + "epoch": 3.381542149420695, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8464 + }, + { + "epoch": 3.3819416699960048, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8465 + }, + { + "epoch": 3.3823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 8466 + }, + { + "epoch": 3.382740711146624, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 8467 + }, + { + "epoch": 3.3831402317219337, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8468 + }, + { + "epoch": 3.3835397522972435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8469 + }, + { + "epoch": 3.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8470 + }, + { + "epoch": 3.3843387934478626, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8471 + }, + { + "epoch": 3.3847383140231724, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8472 + }, + { + "epoch": 3.3851378345984817, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8473 + }, + { + "epoch": 3.3855373551737915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 8474 + }, + { + "epoch": 3.385936875749101, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8475 + }, + { + "epoch": 3.3863363963244106, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8476 + }, + { + "epoch": 3.3867359168997204, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8477 + }, + { + "epoch": 3.38713543747503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8478 + }, + { + "epoch": 3.3875349580503396, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8479 + }, + { + "epoch": 3.3879344786256493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8480 + }, + { + "epoch": 3.3883339992009587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8481 + }, + { + "epoch": 3.3887335197762685, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8482 + }, + { + "epoch": 3.3891330403515783, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8483 + }, + { + "epoch": 3.3895325609268876, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8484 + }, + { + "epoch": 3.3899320815021974, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8485 + }, + { + "epoch": 3.390331602077507, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8486 + }, + { + "epoch": 3.3907311226528165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8487 + }, + { + "epoch": 3.3911306432281263, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8488 + }, + { + "epoch": 3.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8489 + }, + { + "epoch": 3.3919296843787454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8490 + }, + { + "epoch": 3.392329204954055, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8491 + }, + { + "epoch": 3.3927287255293646, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8492 + }, + { + "epoch": 3.3931282461046743, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8493 + }, + { + "epoch": 3.393527766679984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8494 + }, + { + "epoch": 3.3939272872552935, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8495 + }, + { + "epoch": 3.3943268078306033, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8496 + }, + { + "epoch": 3.394726328405913, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8497 + }, + { + "epoch": 3.3951258489812224, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8498 + }, + { + "epoch": 3.395525369556532, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8499 + }, + { + "epoch": 3.395924890131842, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8500 + }, + { + "epoch": 3.3963244107071513, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8501 + }, + { + "epoch": 3.396723931282461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 8502 + }, + { + "epoch": 3.3971234518577704, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8503 + }, + { + "epoch": 3.39752297243308, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8504 + }, + { + "epoch": 3.39792249300839, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8505 + }, + { + "epoch": 3.3983220135837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8506 + }, + { + "epoch": 3.398721534159009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8507 + }, + { + "epoch": 3.399121054734319, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8508 + }, + { + "epoch": 3.3995205753096283, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8509 + }, + { + "epoch": 3.399920095884938, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8510 + }, + { + "epoch": 3.400319616460248, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8511 + }, + { + "epoch": 3.400719137035557, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8512 + }, + { + "epoch": 3.401118657610867, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8513 + }, + { + "epoch": 3.4015181781861767, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8514 + }, + { + "epoch": 3.401917698761486, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8515 + }, + { + "epoch": 3.402317219336796, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8516 + }, + { + "epoch": 3.4027167399121057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8517 + }, + { + "epoch": 3.403116260487415, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8518 + }, + { + "epoch": 3.403515781062725, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 8519 + }, + { + "epoch": 3.403915301638034, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8520 + }, + { + "epoch": 3.404314822213344, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8521 + }, + { + "epoch": 3.4047143427886537, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8522 + }, + { + "epoch": 3.4051138633639635, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8523 + }, + { + "epoch": 3.405513383939273, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8524 + }, + { + "epoch": 3.4059129045145826, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8525 + }, + { + "epoch": 3.406312425089892, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8526 + }, + { + "epoch": 3.4067119456652017, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8527 + }, + { + "epoch": 3.4071114662405115, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8528 + }, + { + "epoch": 3.407510986815821, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8529 + }, + { + "epoch": 3.4079105073911307, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8530 + }, + { + "epoch": 3.4083100279664404, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8531 + }, + { + "epoch": 3.40870954854175, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8532 + }, + { + "epoch": 3.4091090691170596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8533 + }, + { + "epoch": 3.4095085896923694, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8534 + }, + { + "epoch": 3.4099081102676787, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8535 + }, + { + "epoch": 3.4103076308429885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8536 + }, + { + "epoch": 3.410707151418298, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8537 + }, + { + "epoch": 3.4111066719936076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 8538 + }, + { + "epoch": 3.4115061925689174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8539 + }, + { + "epoch": 3.4119057131442267, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8540 + }, + { + "epoch": 3.4123052337195365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8541 + }, + { + "epoch": 3.4127047542948463, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8542 + }, + { + "epoch": 3.4131042748701557, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8543 + }, + { + "epoch": 3.4135037954454654, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8544 + }, + { + "epoch": 3.4139033160207752, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8545 + }, + { + "epoch": 3.4143028365960846, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8546 + }, + { + "epoch": 3.4147023571713944, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8547 + }, + { + "epoch": 3.4151018777467037, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8548 + }, + { + "epoch": 3.4155013983220135, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8549 + }, + { + "epoch": 3.4159009188973233, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8550 + }, + { + "epoch": 3.416300439472633, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8551 + }, + { + "epoch": 3.4166999600479424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8552 + }, + { + "epoch": 3.417099480623252, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8553 + }, + { + "epoch": 3.4174990011985615, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8554 + }, + { + "epoch": 3.4178985217738713, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8555 + }, + { + "epoch": 3.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8556 + }, + { + "epoch": 3.4186975629244905, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8557 + }, + { + "epoch": 3.4190970834998002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8558 + }, + { + "epoch": 3.41949660407511, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8559 + }, + { + "epoch": 3.4198961246504194, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8560 + }, + { + "epoch": 3.420295645225729, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8561 + }, + { + "epoch": 3.420695165801039, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8562 + }, + { + "epoch": 3.4210946863763483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8563 + }, + { + "epoch": 3.421494206951658, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8564 + }, + { + "epoch": 3.4218937275269674, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8565 + }, + { + "epoch": 3.422293248102277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8566 + }, + { + "epoch": 3.422692768677587, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8567 + }, + { + "epoch": 3.4230922892528968, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8568 + }, + { + "epoch": 3.423491809828206, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8569 + }, + { + "epoch": 3.423891330403516, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8570 + }, + { + "epoch": 3.4242908509788252, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8571 + }, + { + "epoch": 3.424690371554135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8572 + }, + { + "epoch": 3.425089892129445, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 8573 + }, + { + "epoch": 3.425489412704754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8574 + }, + { + "epoch": 3.425888933280064, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 8575 + }, + { + "epoch": 3.4262884538553737, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8576 + }, + { + "epoch": 3.426687974430683, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 8577 + }, + { + "epoch": 3.427087495005993, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8578 + }, + { + "epoch": 3.4274870155813026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8579 + }, + { + "epoch": 3.427886536156612, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8580 + }, + { + "epoch": 3.4282860567319218, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8581 + }, + { + "epoch": 3.428685577307231, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8582 + }, + { + "epoch": 3.429085097882541, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8583 + }, + { + "epoch": 3.4294846184578507, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8584 + }, + { + "epoch": 3.42988413903316, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8585 + }, + { + "epoch": 3.43028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8586 + }, + { + "epoch": 3.4306831801837796, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8587 + }, + { + "epoch": 3.431082700759089, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8588 + }, + { + "epoch": 3.4314822213343987, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8589 + }, + { + "epoch": 3.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8590 + }, + { + "epoch": 3.432281262485018, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8591 + }, + { + "epoch": 3.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 8592 + }, + { + "epoch": 3.433080303635637, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8593 + }, + { + "epoch": 3.4334798242109468, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 8594 + }, + { + "epoch": 3.4338793447862566, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8595 + }, + { + "epoch": 3.4342788653615663, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8596 + }, + { + "epoch": 3.4346783859368757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8597 + }, + { + "epoch": 3.4350779065121855, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 8598 + }, + { + "epoch": 3.435477427087495, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8599 + }, + { + "epoch": 3.4358769476628046, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8600 + }, + { + "epoch": 3.4362764682381144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8601 + }, + { + "epoch": 3.4366759888134237, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8602 + }, + { + "epoch": 3.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8603 + }, + { + "epoch": 3.4374750299640433, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8604 + }, + { + "epoch": 3.4378745505393526, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8605 + }, + { + "epoch": 3.4382740711146624, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8606 + }, + { + "epoch": 3.438673591689972, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8607 + }, + { + "epoch": 3.4390731122652816, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8608 + }, + { + "epoch": 3.4394726328405913, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8609 + }, + { + "epoch": 3.4398721534159007, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8610 + }, + { + "epoch": 3.4402716739912105, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 8611 + }, + { + "epoch": 3.4406711945665203, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8612 + }, + { + "epoch": 3.44107071514183, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8613 + }, + { + "epoch": 3.4414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8614 + }, + { + "epoch": 3.441869756292449, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8615 + }, + { + "epoch": 3.4422692768677585, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8616 + }, + { + "epoch": 3.4426687974430683, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8617 + }, + { + "epoch": 3.443068318018378, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8618 + }, + { + "epoch": 3.4434678385936874, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8619 + }, + { + "epoch": 3.443867359168997, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8620 + }, + { + "epoch": 3.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8621 + }, + { + "epoch": 3.4446664003196164, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8622 + }, + { + "epoch": 3.445065920894926, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8623 + }, + { + "epoch": 3.445465441470236, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8624 + }, + { + "epoch": 3.4458649620455453, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8625 + }, + { + "epoch": 3.446264482620855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8626 + }, + { + "epoch": 3.4466640031961644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8627 + }, + { + "epoch": 3.447063523771474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 8628 + }, + { + "epoch": 3.447463044346784, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8629 + }, + { + "epoch": 3.4478625649220933, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8630 + }, + { + "epoch": 3.448262085497403, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8631 + }, + { + "epoch": 3.448661606072713, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8632 + }, + { + "epoch": 3.4490611266480222, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8633 + }, + { + "epoch": 3.449460647223332, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8634 + }, + { + "epoch": 3.449860167798642, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8635 + }, + { + "epoch": 3.450259688373951, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8636 + }, + { + "epoch": 3.450659208949261, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8637 + }, + { + "epoch": 3.4510587295245703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8638 + }, + { + "epoch": 3.45145825009988, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8639 + }, + { + "epoch": 3.45185777067519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 8640 + }, + { + "epoch": 3.4522572912504996, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8641 + }, + { + "epoch": 3.452656811825809, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8642 + }, + { + "epoch": 3.4530563324011188, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8643 + }, + { + "epoch": 3.453455852976428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8644 + }, + { + "epoch": 3.453855373551738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8645 + }, + { + "epoch": 3.4542548941270477, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8646 + }, + { + "epoch": 3.454654414702357, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8647 + }, + { + "epoch": 3.455053935277667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8648 + }, + { + "epoch": 3.4554534558529766, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8649 + }, + { + "epoch": 3.455852976428286, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8650 + }, + { + "epoch": 3.4562524970035957, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 8651 + }, + { + "epoch": 3.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8652 + }, + { + "epoch": 3.457051538154215, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8653 + }, + { + "epoch": 3.4574510587295246, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8654 + }, + { + "epoch": 3.457850579304834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8655 + }, + { + "epoch": 3.4582500998801438, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8656 + }, + { + "epoch": 3.4586496204554535, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8657 + }, + { + "epoch": 3.4590491410307633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8658 + }, + { + "epoch": 3.4594486616060727, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8659 + }, + { + "epoch": 3.4598481821813825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 8660 + }, + { + "epoch": 3.460247702756692, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8661 + }, + { + "epoch": 3.4606472233320016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8662 + }, + { + "epoch": 3.4610467439073114, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8663 + }, + { + "epoch": 3.4614462644826207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8664 + }, + { + "epoch": 3.4618457850579305, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8665 + }, + { + "epoch": 3.4622453056332403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8666 + }, + { + "epoch": 3.4626448262085496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8667 + }, + { + "epoch": 3.4630443467838594, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8668 + }, + { + "epoch": 3.463443867359169, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8669 + }, + { + "epoch": 3.4638433879344785, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8670 + }, + { + "epoch": 3.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8671 + }, + { + "epoch": 3.4646424290850977, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8672 + }, + { + "epoch": 3.4650419496604075, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 8673 + }, + { + "epoch": 3.4654414702357172, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8674 + }, + { + "epoch": 3.4658409908110266, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 8675 + }, + { + "epoch": 3.4662405113863364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8676 + }, + { + "epoch": 3.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8677 + }, + { + "epoch": 3.4670395525369555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8678 + }, + { + "epoch": 3.4674390731122653, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8679 + }, + { + "epoch": 3.467838593687575, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8680 + }, + { + "epoch": 3.4682381142628844, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8681 + }, + { + "epoch": 3.468637634838194, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8682 + }, + { + "epoch": 3.4690371554135035, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8683 + }, + { + "epoch": 3.4694366759888133, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8684 + }, + { + "epoch": 3.469836196564123, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 8685 + }, + { + "epoch": 3.470235717139433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8686 + }, + { + "epoch": 3.4706352377147422, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8687 + }, + { + "epoch": 3.471034758290052, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8688 + }, + { + "epoch": 3.4714342788653614, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8689 + }, + { + "epoch": 3.471833799440671, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8690 + }, + { + "epoch": 3.472233320015981, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8691 + }, + { + "epoch": 3.4726328405912903, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8692 + }, + { + "epoch": 3.4730323611666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8693 + }, + { + "epoch": 3.47343188174191, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8694 + }, + { + "epoch": 3.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8695 + }, + { + "epoch": 3.474230922892529, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8696 + }, + { + "epoch": 3.474630443467839, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8697 + }, + { + "epoch": 3.475029964043148, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8698 + }, + { + "epoch": 3.475429484618458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 8699 + }, + { + "epoch": 3.4758290051937673, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8700 + }, + { + "epoch": 3.476228525769077, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8701 + }, + { + "epoch": 3.476628046344387, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8702 + }, + { + "epoch": 3.4770275669196966, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8703 + }, + { + "epoch": 3.477427087495006, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8704 + }, + { + "epoch": 3.4778266080703157, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8705 + }, + { + "epoch": 3.478226128645625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8706 + }, + { + "epoch": 3.478625649220935, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8707 + }, + { + "epoch": 3.4790251697962447, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8708 + }, + { + "epoch": 3.479424690371554, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8709 + }, + { + "epoch": 3.479824210946864, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8710 + }, + { + "epoch": 3.4802237315221736, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8711 + }, + { + "epoch": 3.480623252097483, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8712 + }, + { + "epoch": 3.4810227726727927, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8713 + }, + { + "epoch": 3.4814222932481025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8714 + }, + { + "epoch": 3.481821813823412, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8715 + }, + { + "epoch": 3.4822213343987216, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8716 + }, + { + "epoch": 3.482620854974031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8717 + }, + { + "epoch": 3.4830203755493407, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 8718 + }, + { + "epoch": 3.4834198961246505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8719 + }, + { + "epoch": 3.48381941669996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8720 + }, + { + "epoch": 3.4842189372752697, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8721 + }, + { + "epoch": 3.4846184578505794, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8722 + }, + { + "epoch": 3.485017978425889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8723 + }, + { + "epoch": 3.4854174990011986, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8724 + }, + { + "epoch": 3.4858170195765084, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8725 + }, + { + "epoch": 3.4862165401518177, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8726 + }, + { + "epoch": 3.4866160607271275, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 8727 + }, + { + "epoch": 3.487015581302437, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8728 + }, + { + "epoch": 3.4874151018777466, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8729 + }, + { + "epoch": 3.4878146224530564, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8730 + }, + { + "epoch": 3.488214143028366, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8731 + }, + { + "epoch": 3.4886136636036755, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8732 + }, + { + "epoch": 3.4890131841789853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8733 + }, + { + "epoch": 3.4894127047542947, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8734 + }, + { + "epoch": 3.4898122253296044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8735 + }, + { + "epoch": 3.4902117459049142, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8736 + }, + { + "epoch": 3.4906112664802236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8737 + }, + { + "epoch": 3.4910107870555334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8738 + }, + { + "epoch": 3.491410307630843, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8739 + }, + { + "epoch": 3.4918098282061525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8740 + }, + { + "epoch": 3.4922093487814623, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8741 + }, + { + "epoch": 3.492608869356772, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8742 + }, + { + "epoch": 3.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8743 + }, + { + "epoch": 3.493407910507391, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8744 + }, + { + "epoch": 3.4938074310827005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8745 + }, + { + "epoch": 3.4942069516580103, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8746 + }, + { + "epoch": 3.49460647223332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 8747 + }, + { + "epoch": 3.49500599280863, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8748 + }, + { + "epoch": 3.4954055133839392, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8749 + }, + { + "epoch": 3.495805033959249, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8750 + }, + { + "epoch": 3.4962045545345584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8751 + }, + { + "epoch": 3.496604075109868, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8752 + }, + { + "epoch": 3.497003595685178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8753 + }, + { + "epoch": 3.4974031162604873, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 8754 + }, + { + "epoch": 3.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8755 + }, + { + "epoch": 3.498202157411107, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8756 + }, + { + "epoch": 3.498601677986416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8757 + }, + { + "epoch": 3.499001198561726, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8758 + }, + { + "epoch": 3.4994007191370358, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8759 + }, + { + "epoch": 3.499800239712345, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8760 + }, + { + "epoch": 3.500199760287655, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8761 + }, + { + "epoch": 3.5005992808629642, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8762 + }, + { + "epoch": 3.500998801438274, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8763 + }, + { + "epoch": 3.501398322013584, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8764 + }, + { + "epoch": 3.5017978425888936, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8765 + }, + { + "epoch": 3.502197363164203, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8766 + }, + { + "epoch": 3.5025968837395127, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8767 + }, + { + "epoch": 3.502996404314822, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8768 + }, + { + "epoch": 3.503395924890132, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8769 + }, + { + "epoch": 3.5037954454654416, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 8770 + }, + { + "epoch": 3.504194966040751, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8771 + }, + { + "epoch": 3.5045944866160608, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8772 + }, + { + "epoch": 3.50499400719137, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8773 + }, + { + "epoch": 3.50539352776668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8774 + }, + { + "epoch": 3.5057930483419897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8775 + }, + { + "epoch": 3.5061925689172995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8776 + }, + { + "epoch": 3.506592089492609, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8777 + }, + { + "epoch": 3.5069916100679186, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8778 + }, + { + "epoch": 3.507391130643228, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 8779 + }, + { + "epoch": 3.5077906512185377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8780 + }, + { + "epoch": 3.5081901717938475, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8781 + }, + { + "epoch": 3.508589692369157, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8782 + }, + { + "epoch": 3.5089892129444666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8783 + }, + { + "epoch": 3.5093887335197764, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8784 + }, + { + "epoch": 3.5097882540950858, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8785 + }, + { + "epoch": 3.5101877746703956, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8786 + }, + { + "epoch": 3.5105872952457053, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8787 + }, + { + "epoch": 3.5109868158210147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8788 + }, + { + "epoch": 3.5113863363963245, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8789 + }, + { + "epoch": 3.511785856971634, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8790 + }, + { + "epoch": 3.5121853775469436, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8791 + }, + { + "epoch": 3.5125848981222534, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8792 + }, + { + "epoch": 3.512984418697563, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8793 + }, + { + "epoch": 3.5133839392728725, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8794 + }, + { + "epoch": 3.5137834598481823, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8795 + }, + { + "epoch": 3.5141829804234916, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8796 + }, + { + "epoch": 3.5145825009988014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8797 + }, + { + "epoch": 3.514982021574111, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 8798 + }, + { + "epoch": 3.5153815421494206, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8799 + }, + { + "epoch": 3.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8800 + }, + { + "epoch": 3.5161805833000397, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8801 + }, + { + "epoch": 3.5165801038753495, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8802 + }, + { + "epoch": 3.5169796244506593, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8803 + }, + { + "epoch": 3.517379145025969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8804 + }, + { + "epoch": 3.5177786656012784, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8805 + }, + { + "epoch": 3.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8806 + }, + { + "epoch": 3.5185777067518975, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8807 + }, + { + "epoch": 3.5189772273272073, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8808 + }, + { + "epoch": 3.519376747902517, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8809 + }, + { + "epoch": 3.519776268477827, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8810 + }, + { + "epoch": 3.520175789053136, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8811 + }, + { + "epoch": 3.520575309628446, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8812 + }, + { + "epoch": 3.5209748302037553, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 8813 + }, + { + "epoch": 3.521374350779065, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8814 + }, + { + "epoch": 3.521773871354375, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8815 + }, + { + "epoch": 3.5221733919296843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8816 + }, + { + "epoch": 3.522572912504994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8817 + }, + { + "epoch": 3.5229724330803034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8818 + }, + { + "epoch": 3.523371953655613, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8819 + }, + { + "epoch": 3.523771474230923, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8820 + }, + { + "epoch": 3.5241709948062327, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8821 + }, + { + "epoch": 3.524570515381542, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8822 + }, + { + "epoch": 3.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8823 + }, + { + "epoch": 3.525369556532161, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8824 + }, + { + "epoch": 3.525769077107471, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8825 + }, + { + "epoch": 3.526168597682781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8826 + }, + { + "epoch": 3.52656811825809, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8827 + }, + { + "epoch": 3.5269676388334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8828 + }, + { + "epoch": 3.5273671594087097, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8829 + }, + { + "epoch": 3.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8830 + }, + { + "epoch": 3.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8831 + }, + { + "epoch": 3.5285657211346386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8832 + }, + { + "epoch": 3.528965241709948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8833 + }, + { + "epoch": 3.5293647622852578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8834 + }, + { + "epoch": 3.529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8835 + }, + { + "epoch": 3.530163803435877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8836 + }, + { + "epoch": 3.5305633240111867, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8837 + }, + { + "epoch": 3.5309628445864965, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8838 + }, + { + "epoch": 3.531362365161806, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8839 + }, + { + "epoch": 3.5317618857371156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8840 + }, + { + "epoch": 3.532161406312425, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8841 + }, + { + "epoch": 3.5325609268877347, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8842 + }, + { + "epoch": 3.5329604474630445, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8843 + }, + { + "epoch": 3.533359968038354, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8844 + }, + { + "epoch": 3.5337594886136636, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8845 + }, + { + "epoch": 3.534159009188973, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8846 + }, + { + "epoch": 3.5345585297642828, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8847 + }, + { + "epoch": 3.5349580503395925, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8848 + }, + { + "epoch": 3.5353575709149023, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8849 + }, + { + "epoch": 3.5357570914902117, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8850 + }, + { + "epoch": 3.5361566120655215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8851 + }, + { + "epoch": 3.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8852 + }, + { + "epoch": 3.5369556532161406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8853 + }, + { + "epoch": 3.5373551737914504, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8854 + }, + { + "epoch": 3.53775469436676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8855 + }, + { + "epoch": 3.5381542149420695, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8856 + }, + { + "epoch": 3.5385537355173793, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8857 + }, + { + "epoch": 3.5389532560926886, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8858 + }, + { + "epoch": 3.5393527766679984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8859 + }, + { + "epoch": 3.539752297243308, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8860 + }, + { + "epoch": 3.5401518178186175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8861 + }, + { + "epoch": 3.5405513383939273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8862 + }, + { + "epoch": 3.5409508589692367, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8863 + }, + { + "epoch": 3.5413503795445465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8864 + }, + { + "epoch": 3.5417499001198562, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 8865 + }, + { + "epoch": 3.542149420695166, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8866 + }, + { + "epoch": 3.5425489412704754, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8867 + }, + { + "epoch": 3.542948461845785, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8868 + }, + { + "epoch": 3.5433479824210945, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 8869 + }, + { + "epoch": 3.5437475029964043, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8870 + }, + { + "epoch": 3.544147023571714, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8871 + }, + { + "epoch": 3.5445465441470234, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8872 + }, + { + "epoch": 3.544946064722333, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8873 + }, + { + "epoch": 3.545345585297643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8874 + }, + { + "epoch": 3.5457451058729523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8875 + }, + { + "epoch": 3.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8876 + }, + { + "epoch": 3.546544147023572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8877 + }, + { + "epoch": 3.5469436675988812, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8878 + }, + { + "epoch": 3.547343188174191, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8879 + }, + { + "epoch": 3.5477427087495004, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8880 + }, + { + "epoch": 3.54814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8881 + }, + { + "epoch": 3.54854174990012, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8882 + }, + { + "epoch": 3.5489412704754297, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8883 + }, + { + "epoch": 3.549340791050739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8884 + }, + { + "epoch": 3.549740311626049, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8885 + }, + { + "epoch": 3.550139832201358, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8886 + }, + { + "epoch": 3.550539352776668, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8887 + }, + { + "epoch": 3.5509388733519778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8888 + }, + { + "epoch": 3.551338393927287, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8889 + }, + { + "epoch": 3.551737914502597, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8890 + }, + { + "epoch": 3.5521374350779062, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8891 + }, + { + "epoch": 3.552536955653216, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8892 + }, + { + "epoch": 3.552936476228526, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8893 + }, + { + "epoch": 3.5533359968038356, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8894 + }, + { + "epoch": 3.553735517379145, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8895 + }, + { + "epoch": 3.5541350379544547, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8896 + }, + { + "epoch": 3.554534558529764, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8897 + }, + { + "epoch": 3.554934079105074, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8898 + }, + { + "epoch": 3.5553335996803836, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8899 + }, + { + "epoch": 3.5557331202556934, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8900 + }, + { + "epoch": 3.556132640831003, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8901 + }, + { + "epoch": 3.5565321614063126, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8902 + }, + { + "epoch": 3.556931681981622, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8903 + }, + { + "epoch": 3.5573312025569317, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8904 + }, + { + "epoch": 3.5577307231322415, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8905 + }, + { + "epoch": 3.558130243707551, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8906 + }, + { + "epoch": 3.5585297642828606, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8907 + }, + { + "epoch": 3.55892928485817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8908 + }, + { + "epoch": 3.5593288054334797, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8909 + }, + { + "epoch": 3.5597283260087895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8910 + }, + { + "epoch": 3.5601278465840993, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8911 + }, + { + "epoch": 3.5605273671594087, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8912 + }, + { + "epoch": 3.5609268877347184, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8913 + }, + { + "epoch": 3.561326408310028, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8914 + }, + { + "epoch": 3.5617259288853376, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8915 + }, + { + "epoch": 3.5621254494606474, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8916 + }, + { + "epoch": 3.5625249700359567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8917 + }, + { + "epoch": 3.5629244906112665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8918 + }, + { + "epoch": 3.5633240111865763, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8919 + }, + { + "epoch": 3.5637235317618856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8920 + }, + { + "epoch": 3.5641230523371954, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8921 + }, + { + "epoch": 3.564522572912505, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8922 + }, + { + "epoch": 3.5649220934878145, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8923 + }, + { + "epoch": 3.5653216140631243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8924 + }, + { + "epoch": 3.5657211346384337, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 8925 + }, + { + "epoch": 3.5661206552137434, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8926 + }, + { + "epoch": 3.5665201757890532, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8927 + }, + { + "epoch": 3.566919696364363, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8928 + }, + { + "epoch": 3.5673192169396724, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8929 + }, + { + "epoch": 3.567718737514982, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8930 + }, + { + "epoch": 3.5681182580902915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8931 + }, + { + "epoch": 3.5685177786656013, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8932 + }, + { + "epoch": 3.568917299240911, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 8933 + }, + { + "epoch": 3.5693168198162204, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8934 + }, + { + "epoch": 3.56971634039153, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8935 + }, + { + "epoch": 3.5701158609668395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8936 + }, + { + "epoch": 3.5705153815421493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8937 + }, + { + "epoch": 3.570914902117459, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8938 + }, + { + "epoch": 3.571314422692769, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8939 + }, + { + "epoch": 3.5717139432680782, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8940 + }, + { + "epoch": 3.572113463843388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8941 + }, + { + "epoch": 3.5725129844186974, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8942 + }, + { + "epoch": 3.572912504994007, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 8943 + }, + { + "epoch": 3.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8944 + }, + { + "epoch": 3.5737115461446267, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8945 + }, + { + "epoch": 3.574111066719936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8946 + }, + { + "epoch": 3.574510587295246, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8947 + }, + { + "epoch": 3.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8948 + }, + { + "epoch": 3.575309628445865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 8949 + }, + { + "epoch": 3.5757091490211748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8950 + }, + { + "epoch": 3.576108669596484, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8951 + }, + { + "epoch": 3.576508190171794, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8952 + }, + { + "epoch": 3.5769077107471032, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8953 + }, + { + "epoch": 3.577307231322413, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8954 + }, + { + "epoch": 3.577706751897723, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8955 + }, + { + "epoch": 3.5781062724730326, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8956 + }, + { + "epoch": 3.578505793048342, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8957 + }, + { + "epoch": 3.5789053136236517, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8958 + }, + { + "epoch": 3.579304834198961, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8959 + }, + { + "epoch": 3.579704354774271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8960 + }, + { + "epoch": 3.5801038753495806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8961 + }, + { + "epoch": 3.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8962 + }, + { + "epoch": 3.5809029165001998, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8963 + }, + { + "epoch": 3.5813024370755095, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8964 + }, + { + "epoch": 3.581701957650819, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8965 + }, + { + "epoch": 3.5821014782261287, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8966 + }, + { + "epoch": 3.5825009988014385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8967 + }, + { + "epoch": 3.582900519376748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8968 + }, + { + "epoch": 3.5833000399520576, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8969 + }, + { + "epoch": 3.583699560527367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8970 + }, + { + "epoch": 3.5840990811026767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8971 + }, + { + "epoch": 3.5844986016779865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8972 + }, + { + "epoch": 3.5848981222532963, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8973 + }, + { + "epoch": 3.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8974 + }, + { + "epoch": 3.5856971634039154, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 8975 + }, + { + "epoch": 3.5860966839792248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8976 + }, + { + "epoch": 3.5864962045545346, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8977 + }, + { + "epoch": 3.5868957251298443, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8978 + }, + { + "epoch": 3.5872952457051537, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8979 + }, + { + "epoch": 3.5876947662804635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8980 + }, + { + "epoch": 3.588094286855773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8981 + }, + { + "epoch": 3.5884938074310826, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 8982 + }, + { + "epoch": 3.5888933280063924, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8983 + }, + { + "epoch": 3.589292848581702, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8984 + }, + { + "epoch": 3.5896923691570115, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 8985 + }, + { + "epoch": 3.5900918897323213, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8986 + }, + { + "epoch": 3.5904914103076306, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 8987 + }, + { + "epoch": 3.5908909308829404, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8988 + }, + { + "epoch": 3.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8989 + }, + { + "epoch": 3.59168997203356, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8990 + }, + { + "epoch": 3.5920894926088693, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8991 + }, + { + "epoch": 3.592489013184179, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8992 + }, + { + "epoch": 3.5928885337594885, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8993 + }, + { + "epoch": 3.5932880543347983, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8994 + }, + { + "epoch": 3.593687574910108, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8995 + }, + { + "epoch": 3.5940870954854174, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 8996 + }, + { + "epoch": 3.594486616060727, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8997 + }, + { + "epoch": 3.5948861366360365, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8998 + }, + { + "epoch": 3.5952856572113463, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8999 + }, + { + "epoch": 3.595685177786656, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9000 + }, + { + "epoch": 3.596084698361966, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9001 + }, + { + "epoch": 3.596484218937275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9002 + }, + { + "epoch": 3.596883739512585, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 9003 + }, + { + "epoch": 3.5972832600878943, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9004 + }, + { + "epoch": 3.597682780663204, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9005 + }, + { + "epoch": 3.598082301238514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9006 + }, + { + "epoch": 3.5984818218138233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9007 + }, + { + "epoch": 3.598881342389133, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9008 + }, + { + "epoch": 3.599280862964443, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 9009 + }, + { + "epoch": 3.599680383539752, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9010 + }, + { + "epoch": 3.600079904115062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9011 + }, + { + "epoch": 3.6004794246903717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9012 + }, + { + "epoch": 3.600878945265681, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9013 + }, + { + "epoch": 3.601278465840991, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9014 + }, + { + "epoch": 3.6016779864163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9015 + }, + { + "epoch": 3.60207750699161, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9016 + }, + { + "epoch": 3.60247702756692, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9017 + }, + { + "epoch": 3.6028765481422296, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9018 + }, + { + "epoch": 3.603276068717539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9019 + }, + { + "epoch": 3.6036755892928487, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9020 + }, + { + "epoch": 3.604075109868158, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9021 + }, + { + "epoch": 3.604474630443468, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9022 + }, + { + "epoch": 3.6048741510187776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9023 + }, + { + "epoch": 3.605273671594087, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9024 + }, + { + "epoch": 3.6056731921693967, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9025 + }, + { + "epoch": 3.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9026 + }, + { + "epoch": 3.606472233320016, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9027 + }, + { + "epoch": 3.6068717538953257, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9028 + }, + { + "epoch": 3.6072712744706354, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9029 + }, + { + "epoch": 3.607670795045945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9030 + }, + { + "epoch": 3.6080703156212546, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9031 + }, + { + "epoch": 3.608469836196564, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9032 + }, + { + "epoch": 3.6088693567718737, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9033 + }, + { + "epoch": 3.6092688773471835, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 9034 + }, + { + "epoch": 3.6096683979224933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9035 + }, + { + "epoch": 3.6100679184978026, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9036 + }, + { + "epoch": 3.6104674390731124, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9037 + }, + { + "epoch": 3.6108669596484217, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9038 + }, + { + "epoch": 3.6112664802237315, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 9039 + }, + { + "epoch": 3.6116660007990413, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9040 + }, + { + "epoch": 3.6120655213743507, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9041 + }, + { + "epoch": 3.6124650419496604, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9042 + }, + { + "epoch": 3.61286456252497, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9043 + }, + { + "epoch": 3.6132640831002796, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9044 + }, + { + "epoch": 3.6136636036755894, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9045 + }, + { + "epoch": 3.614063124250899, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9046 + }, + { + "epoch": 3.6144626448262085, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9047 + }, + { + "epoch": 3.6148621654015183, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9048 + }, + { + "epoch": 3.6152616859768276, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9049 + }, + { + "epoch": 3.6156612065521374, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9050 + }, + { + "epoch": 3.616060727127447, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9051 + }, + { + "epoch": 3.6164602477027565, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9052 + }, + { + "epoch": 3.6168597682780663, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9053 + }, + { + "epoch": 3.617259288853376, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9054 + }, + { + "epoch": 3.6176588094286855, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9055 + }, + { + "epoch": 3.6180583300039952, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9056 + }, + { + "epoch": 3.618457850579305, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9057 + }, + { + "epoch": 3.6188573711546144, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9058 + }, + { + "epoch": 3.619256891729924, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9059 + }, + { + "epoch": 3.6196564123052335, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9060 + }, + { + "epoch": 3.6200559328805433, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9061 + }, + { + "epoch": 3.620455453455853, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9062 + }, + { + "epoch": 3.620854974031163, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 9063 + }, + { + "epoch": 3.621254494606472, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9064 + }, + { + "epoch": 3.621654015181782, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9065 + }, + { + "epoch": 3.6220535357570913, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9066 + }, + { + "epoch": 3.622453056332401, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9067 + }, + { + "epoch": 3.622852576907711, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9068 + }, + { + "epoch": 3.6232520974830202, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9069 + }, + { + "epoch": 3.62365161805833, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9070 + }, + { + "epoch": 3.6240511386336394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9071 + }, + { + "epoch": 3.624450659208949, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9072 + }, + { + "epoch": 3.624850179784259, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9073 + }, + { + "epoch": 3.6252497003595687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9074 + }, + { + "epoch": 3.625649220934878, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 9075 + }, + { + "epoch": 3.626048741510188, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9076 + }, + { + "epoch": 3.626448262085497, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9077 + }, + { + "epoch": 3.626847782660807, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9078 + }, + { + "epoch": 3.6272473032361168, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9079 + }, + { + "epoch": 3.6276468238114266, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9080 + }, + { + "epoch": 3.628046344386736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9081 + }, + { + "epoch": 3.6284458649620457, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9082 + }, + { + "epoch": 3.628845385537355, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9083 + }, + { + "epoch": 3.629244906112665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9084 + }, + { + "epoch": 3.6296444266879746, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9085 + }, + { + "epoch": 3.630043947263284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9086 + }, + { + "epoch": 3.6304434678385937, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9087 + }, + { + "epoch": 3.630842988413903, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9088 + }, + { + "epoch": 3.631242508989213, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9089 + }, + { + "epoch": 3.6316420295645226, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9090 + }, + { + "epoch": 3.6320415501398324, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9091 + }, + { + "epoch": 3.6324410707151418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 9092 + }, + { + "epoch": 3.6328405912904516, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9093 + }, + { + "epoch": 3.633240111865761, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9094 + }, + { + "epoch": 3.6336396324410707, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9095 + }, + { + "epoch": 3.6340391530163805, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9096 + }, + { + "epoch": 3.63443867359169, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9097 + }, + { + "epoch": 3.6348381941669996, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9098 + }, + { + "epoch": 3.6352377147423094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9099 + }, + { + "epoch": 3.6356372353176187, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9100 + }, + { + "epoch": 3.6360367558929285, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9101 + }, + { + "epoch": 3.6364362764682383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9102 + }, + { + "epoch": 3.6368357970435476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9103 + }, + { + "epoch": 3.6372353176188574, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9104 + }, + { + "epoch": 3.6376348381941668, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9105 + }, + { + "epoch": 3.6380343587694766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9106 + }, + { + "epoch": 3.6384338793447863, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9107 + }, + { + "epoch": 3.638833399920096, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9108 + }, + { + "epoch": 3.6392329204954055, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9109 + }, + { + "epoch": 3.6396324410707153, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9110 + }, + { + "epoch": 3.6400319616460246, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9111 + }, + { + "epoch": 3.6404314822213344, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9112 + }, + { + "epoch": 3.640831002796644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9113 + }, + { + "epoch": 3.6412305233719535, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9114 + }, + { + "epoch": 3.6416300439472633, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9115 + }, + { + "epoch": 3.6420295645225726, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 9116 + }, + { + "epoch": 3.6424290850978824, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9117 + }, + { + "epoch": 3.642828605673192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9118 + }, + { + "epoch": 3.643228126248502, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9119 + }, + { + "epoch": 3.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9120 + }, + { + "epoch": 3.644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9121 + }, + { + "epoch": 3.6444266879744305, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9122 + }, + { + "epoch": 3.6448262085497403, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9123 + }, + { + "epoch": 3.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9124 + }, + { + "epoch": 3.64562524970036, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9125 + }, + { + "epoch": 3.646024770275669, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9126 + }, + { + "epoch": 3.646424290850979, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9127 + }, + { + "epoch": 3.6468238114262883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9128 + }, + { + "epoch": 3.647223332001598, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9129 + }, + { + "epoch": 3.647622852576908, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9130 + }, + { + "epoch": 3.6480223731522172, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9131 + }, + { + "epoch": 3.648421893727527, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9132 + }, + { + "epoch": 3.6488214143028364, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 9133 + }, + { + "epoch": 3.649220934878146, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9134 + }, + { + "epoch": 3.649620455453456, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9135 + }, + { + "epoch": 3.6500199760287657, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9136 + }, + { + "epoch": 3.650419496604075, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9137 + }, + { + "epoch": 3.650819017179385, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9138 + }, + { + "epoch": 3.651218537754694, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9139 + }, + { + "epoch": 3.651618058330004, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 9140 + }, + { + "epoch": 3.6520175789053138, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9141 + }, + { + "epoch": 3.652417099480623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9142 + }, + { + "epoch": 3.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9143 + }, + { + "epoch": 3.6532161406312427, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9144 + }, + { + "epoch": 3.653615661206552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 9145 + }, + { + "epoch": 3.654015181781862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9146 + }, + { + "epoch": 3.6544147023571716, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9147 + }, + { + "epoch": 3.654814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9148 + }, + { + "epoch": 3.6552137435077907, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9149 + }, + { + "epoch": 3.6556132640831, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9150 + }, + { + "epoch": 3.65601278465841, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9151 + }, + { + "epoch": 3.6564123052337196, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9152 + }, + { + "epoch": 3.6568118258090294, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9153 + }, + { + "epoch": 3.6572113463843388, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9154 + }, + { + "epoch": 3.6576108669596485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 9155 + }, + { + "epoch": 3.658010387534958, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9156 + }, + { + "epoch": 3.6584099081102677, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9157 + }, + { + "epoch": 3.6588094286855775, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9158 + }, + { + "epoch": 3.659208949260887, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9159 + }, + { + "epoch": 3.6596084698361966, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9160 + }, + { + "epoch": 3.660007990411506, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9161 + }, + { + "epoch": 3.6604075109868157, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9162 + }, + { + "epoch": 3.6608070315621255, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9163 + }, + { + "epoch": 3.6612065521374353, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9164 + }, + { + "epoch": 3.6616060727127446, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9165 + }, + { + "epoch": 3.6620055932880544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9166 + }, + { + "epoch": 3.6624051138633638, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9167 + }, + { + "epoch": 3.6628046344386735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9168 + }, + { + "epoch": 3.6632041550139833, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9169 + }, + { + "epoch": 3.663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9170 + }, + { + "epoch": 3.6640031961646025, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9171 + }, + { + "epoch": 3.6644027167399122, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9172 + }, + { + "epoch": 3.6648022373152216, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9173 + }, + { + "epoch": 3.6652017578905314, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9174 + }, + { + "epoch": 3.665601278465841, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9175 + }, + { + "epoch": 3.6660007990411505, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9176 + }, + { + "epoch": 3.6664003196164603, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9177 + }, + { + "epoch": 3.6667998401917696, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9178 + }, + { + "epoch": 3.6671993607670794, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9179 + }, + { + "epoch": 3.667598881342389, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9180 + }, + { + "epoch": 3.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9181 + }, + { + "epoch": 3.6683979224930083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9182 + }, + { + "epoch": 3.668797443068318, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9183 + }, + { + "epoch": 3.6691969636436275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9184 + }, + { + "epoch": 3.6695964842189372, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9185 + }, + { + "epoch": 3.669996004794247, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9186 + }, + { + "epoch": 3.6703955253695564, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9187 + }, + { + "epoch": 3.670795045944866, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9188 + }, + { + "epoch": 3.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9189 + }, + { + "epoch": 3.6715940870954853, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9190 + }, + { + "epoch": 3.671993607670795, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9191 + }, + { + "epoch": 3.672393128246105, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9192 + }, + { + "epoch": 3.672792648821414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9193 + }, + { + "epoch": 3.673192169396724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9194 + }, + { + "epoch": 3.6735916899720333, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9195 + }, + { + "epoch": 3.673991210547343, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9196 + }, + { + "epoch": 3.674390731122653, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9197 + }, + { + "epoch": 3.6747902516979627, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9198 + }, + { + "epoch": 3.675189772273272, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9199 + }, + { + "epoch": 3.675589292848582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9200 + }, + { + "epoch": 3.675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 9201 + }, + { + "epoch": 3.676388333999201, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9202 + }, + { + "epoch": 3.6767878545745107, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9203 + }, + { + "epoch": 3.67718737514982, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9204 + }, + { + "epoch": 3.67758689572513, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9205 + }, + { + "epoch": 3.677986416300439, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9206 + }, + { + "epoch": 3.678385936875749, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9207 + }, + { + "epoch": 3.678785457451059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9208 + }, + { + "epoch": 3.6791849780263686, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9209 + }, + { + "epoch": 3.679584498601678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9210 + }, + { + "epoch": 3.6799840191769877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9211 + }, + { + "epoch": 3.680383539752297, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9212 + }, + { + "epoch": 3.680783060327607, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9213 + }, + { + "epoch": 3.6811825809029166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9214 + }, + { + "epoch": 3.6815821014782264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9215 + }, + { + "epoch": 3.6819816220535357, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9216 + }, + { + "epoch": 3.6823811426288455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9217 + }, + { + "epoch": 3.682780663204155, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9218 + }, + { + "epoch": 3.6831801837794647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9219 + }, + { + "epoch": 3.6835797043547744, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9220 + }, + { + "epoch": 3.683979224930084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9221 + }, + { + "epoch": 3.6843787455053936, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9222 + }, + { + "epoch": 3.684778266080703, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9223 + }, + { + "epoch": 3.6851777866560127, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9224 + }, + { + "epoch": 3.6855773072313225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9225 + }, + { + "epoch": 3.6859768278066323, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9226 + }, + { + "epoch": 3.6863763483819416, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9227 + }, + { + "epoch": 3.6867758689572514, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9228 + }, + { + "epoch": 3.6871753895325607, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9229 + }, + { + "epoch": 3.6875749101078705, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9230 + }, + { + "epoch": 3.6879744306831803, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9231 + }, + { + "epoch": 3.6883739512584897, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9232 + }, + { + "epoch": 3.6887734718337994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9233 + }, + { + "epoch": 3.6891729924091092, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9234 + }, + { + "epoch": 3.6895725129844186, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9235 + }, + { + "epoch": 3.6899720335597284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9236 + }, + { + "epoch": 3.690371554135038, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9237 + }, + { + "epoch": 3.6907710747103475, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9238 + }, + { + "epoch": 3.6911705952856573, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9239 + }, + { + "epoch": 3.6915701158609666, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9240 + }, + { + "epoch": 3.6919696364362764, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9241 + }, + { + "epoch": 3.692369157011586, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9242 + }, + { + "epoch": 3.692768677586896, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9243 + }, + { + "epoch": 3.6931681981622053, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9244 + }, + { + "epoch": 3.693567718737515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9245 + }, + { + "epoch": 3.6939672393128244, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9246 + }, + { + "epoch": 3.6943667598881342, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9247 + }, + { + "epoch": 3.694766280463444, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9248 + }, + { + "epoch": 3.6951658010387534, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 9249 + }, + { + "epoch": 3.695565321614063, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9250 + }, + { + "epoch": 3.6959648421893725, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 9251 + }, + { + "epoch": 3.6963643627646823, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 9252 + }, + { + "epoch": 3.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9253 + }, + { + "epoch": 3.697163403915302, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9254 + }, + { + "epoch": 3.697562924490611, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9255 + }, + { + "epoch": 3.697962445065921, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9256 + }, + { + "epoch": 3.6983619656412303, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9257 + }, + { + "epoch": 3.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9258 + }, + { + "epoch": 3.69916100679185, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9259 + }, + { + "epoch": 3.6995605273671597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9260 + }, + { + "epoch": 3.699960047942469, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9261 + }, + { + "epoch": 3.700359568517779, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 9262 + }, + { + "epoch": 3.700759089093088, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9263 + }, + { + "epoch": 3.701158609668398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9264 + }, + { + "epoch": 3.7015581302437077, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9265 + }, + { + "epoch": 3.701957650819017, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9266 + }, + { + "epoch": 3.702357171394327, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9267 + }, + { + "epoch": 3.702756691969636, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9268 + }, + { + "epoch": 3.703156212544946, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9269 + }, + { + "epoch": 3.7035557331202558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9270 + }, + { + "epoch": 3.7039552536955656, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9271 + }, + { + "epoch": 3.704354774270875, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9272 + }, + { + "epoch": 3.7047542948461847, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9273 + }, + { + "epoch": 3.705153815421494, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9274 + }, + { + "epoch": 3.705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9275 + }, + { + "epoch": 3.7059528565721136, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9276 + }, + { + "epoch": 3.706352377147423, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9277 + }, + { + "epoch": 3.7067518977227327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9278 + }, + { + "epoch": 3.7071514182980425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9279 + }, + { + "epoch": 3.707550938873352, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9280 + }, + { + "epoch": 3.7079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9281 + }, + { + "epoch": 3.7083499800239714, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9282 + }, + { + "epoch": 3.7087495005992808, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9283 + }, + { + "epoch": 3.7091490211745906, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9284 + }, + { + "epoch": 3.7095485417499, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9285 + }, + { + "epoch": 3.7099480623252097, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9286 + }, + { + "epoch": 3.7103475829005195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9287 + }, + { + "epoch": 3.7107471034758293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9288 + }, + { + "epoch": 3.7111466240511386, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9289 + }, + { + "epoch": 3.7115461446264484, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9290 + }, + { + "epoch": 3.7119456652017577, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9291 + }, + { + "epoch": 3.7123451857770675, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9292 + }, + { + "epoch": 3.7127447063523773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9293 + }, + { + "epoch": 3.7131442269276866, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9294 + }, + { + "epoch": 3.7135437475029964, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9295 + }, + { + "epoch": 3.7139432680783058, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 9296 + }, + { + "epoch": 3.7143427886536156, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9297 + }, + { + "epoch": 3.7147423092289253, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9298 + }, + { + "epoch": 3.715141829804235, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9299 + }, + { + "epoch": 3.7155413503795445, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9300 + }, + { + "epoch": 3.7159408709548543, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 9301 + }, + { + "epoch": 3.7163403915301636, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9302 + }, + { + "epoch": 3.7167399121054734, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 9303 + }, + { + "epoch": 3.717139432680783, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9304 + }, + { + "epoch": 3.717538953256093, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9305 + }, + { + "epoch": 3.7179384738314023, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9306 + }, + { + "epoch": 3.718337994406712, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 9307 + }, + { + "epoch": 3.7187375149820214, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9308 + }, + { + "epoch": 3.719137035557331, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9309 + }, + { + "epoch": 3.719536556132641, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9310 + }, + { + "epoch": 3.7199360767079503, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9311 + }, + { + "epoch": 3.72033559728326, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 9312 + }, + { + "epoch": 3.7207351178585695, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9313 + }, + { + "epoch": 3.7211346384338793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9314 + }, + { + "epoch": 3.721534159009189, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9315 + }, + { + "epoch": 3.721933679584499, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9316 + }, + { + "epoch": 3.722333200159808, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9317 + }, + { + "epoch": 3.722732720735118, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9318 + }, + { + "epoch": 3.7231322413104273, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9319 + }, + { + "epoch": 3.723531761885737, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 9320 + }, + { + "epoch": 3.723931282461047, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9321 + }, + { + "epoch": 3.724330803036356, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9322 + }, + { + "epoch": 3.724730323611666, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9323 + }, + { + "epoch": 3.725129844186976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9324 + }, + { + "epoch": 3.725529364762285, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9325 + }, + { + "epoch": 3.725928885337595, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9326 + }, + { + "epoch": 3.7263284059129047, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9327 + }, + { + "epoch": 3.726727926488214, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9328 + }, + { + "epoch": 3.727127447063524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9329 + }, + { + "epoch": 3.727526967638833, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 9330 + }, + { + "epoch": 3.727926488214143, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9331 + }, + { + "epoch": 3.7283260087894528, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9332 + }, + { + "epoch": 3.7287255293647625, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9333 + }, + { + "epoch": 3.729125049940072, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9334 + }, + { + "epoch": 3.7295245705153817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 9335 + }, + { + "epoch": 3.729924091090691, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9336 + }, + { + "epoch": 3.730323611666001, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9337 + }, + { + "epoch": 3.7307231322413106, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9338 + }, + { + "epoch": 3.73112265281662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9339 + }, + { + "epoch": 3.7315221733919297, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9340 + }, + { + "epoch": 3.731921693967239, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9341 + }, + { + "epoch": 3.732321214542549, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9342 + }, + { + "epoch": 3.7327207351178586, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9343 + }, + { + "epoch": 3.7331202556931684, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9344 + }, + { + "epoch": 3.7335197762684778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9345 + }, + { + "epoch": 3.7339192968437875, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9346 + }, + { + "epoch": 3.734318817419097, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9347 + }, + { + "epoch": 3.7347183379944067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9348 + }, + { + "epoch": 3.7351178585697165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9349 + }, + { + "epoch": 3.7355173791450262, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9350 + }, + { + "epoch": 3.7359168997203356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9351 + }, + { + "epoch": 3.7363164202956454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9352 + }, + { + "epoch": 3.7367159408709547, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9353 + }, + { + "epoch": 3.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 9354 + }, + { + "epoch": 3.7375149820215743, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9355 + }, + { + "epoch": 3.7379145025968836, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9356 + }, + { + "epoch": 3.7383140231721934, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9357 + }, + { + "epoch": 3.7387135437475028, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 9358 + }, + { + "epoch": 3.7391130643228125, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 9359 + }, + { + "epoch": 3.7395125848981223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9360 + }, + { + "epoch": 3.739912105473432, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9361 + }, + { + "epoch": 3.7403116260487415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9362 + }, + { + "epoch": 3.7407111466240512, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 9363 + }, + { + "epoch": 3.7411106671993606, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9364 + }, + { + "epoch": 3.7415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9365 + }, + { + "epoch": 3.74190970834998, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9366 + }, + { + "epoch": 3.7423092289252895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9367 + }, + { + "epoch": 3.7427087495005993, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9368 + }, + { + "epoch": 3.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 9369 + }, + { + "epoch": 3.7435077906512184, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 9370 + }, + { + "epoch": 3.743907311226528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9371 + }, + { + "epoch": 3.744306831801838, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9372 + }, + { + "epoch": 3.7447063523771473, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 9373 + }, + { + "epoch": 3.745105872952457, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9374 + }, + { + "epoch": 3.7455053935277665, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9375 + }, + { + "epoch": 3.7459049141030762, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9376 + }, + { + "epoch": 3.746304434678386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9377 + }, + { + "epoch": 3.746703955253696, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9378 + }, + { + "epoch": 3.747103475829005, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9379 + }, + { + "epoch": 3.747502996404315, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9380 + }, + { + "epoch": 3.7479025169796243, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9381 + }, + { + "epoch": 3.748302037554934, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9382 + }, + { + "epoch": 3.748701558130244, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9383 + }, + { + "epoch": 3.749101078705553, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9384 + }, + { + "epoch": 3.749500599280863, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9385 + }, + { + "epoch": 3.7499001198561723, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9386 + }, + { + "epoch": 3.750299640431482, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9387 + }, + { + "epoch": 3.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9388 + }, + { + "epoch": 3.7510986815821017, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9389 + }, + { + "epoch": 3.751498202157411, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9390 + }, + { + "epoch": 3.751897722732721, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9391 + }, + { + "epoch": 3.75229724330803, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9392 + }, + { + "epoch": 3.75269676388334, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9393 + }, + { + "epoch": 3.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9394 + }, + { + "epoch": 3.7534958050339595, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9395 + }, + { + "epoch": 3.753895325609269, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9396 + }, + { + "epoch": 3.7542948461845786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9397 + }, + { + "epoch": 3.754694366759888, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9398 + }, + { + "epoch": 3.7550938873351978, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9399 + }, + { + "epoch": 3.7554934079105076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9400 + }, + { + "epoch": 3.755892928485817, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9401 + }, + { + "epoch": 3.7562924490611267, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9402 + }, + { + "epoch": 3.756691969636436, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9403 + }, + { + "epoch": 3.757091490211746, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9404 + }, + { + "epoch": 3.7574910107870556, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9405 + }, + { + "epoch": 3.7578905313623654, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9406 + }, + { + "epoch": 3.7582900519376747, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9407 + }, + { + "epoch": 3.7586895725129845, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9408 + }, + { + "epoch": 3.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9409 + }, + { + "epoch": 3.7594886136636037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9410 + }, + { + "epoch": 3.7598881342389134, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9411 + }, + { + "epoch": 3.760287654814223, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9412 + }, + { + "epoch": 3.7606871753895326, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9413 + }, + { + "epoch": 3.7610866959648424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9414 + }, + { + "epoch": 3.7614862165401517, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9415 + }, + { + "epoch": 3.7618857371154615, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9416 + }, + { + "epoch": 3.7622852576907713, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9417 + }, + { + "epoch": 3.7626847782660806, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9418 + }, + { + "epoch": 3.7630842988413904, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9419 + }, + { + "epoch": 3.7634838194166997, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9420 + }, + { + "epoch": 3.7638833399920095, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9421 + }, + { + "epoch": 3.7642828605673193, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9422 + }, + { + "epoch": 3.764682381142629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9423 + }, + { + "epoch": 3.7650819017179384, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9424 + }, + { + "epoch": 3.7654814222932482, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9425 + }, + { + "epoch": 3.7658809428685576, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9426 + }, + { + "epoch": 3.7662804634438674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9427 + }, + { + "epoch": 3.766679984019177, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9428 + }, + { + "epoch": 3.7670795045944865, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9429 + }, + { + "epoch": 3.7674790251697963, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9430 + }, + { + "epoch": 3.7678785457451056, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9431 + }, + { + "epoch": 3.7682780663204154, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9432 + }, + { + "epoch": 3.768677586895725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9433 + }, + { + "epoch": 3.769077107471035, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9434 + }, + { + "epoch": 3.7694766280463443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9435 + }, + { + "epoch": 3.769876148621654, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9436 + }, + { + "epoch": 3.7702756691969634, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9437 + }, + { + "epoch": 3.7706751897722732, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9438 + }, + { + "epoch": 3.771074710347583, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9439 + }, + { + "epoch": 3.771474230922893, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 9440 + }, + { + "epoch": 3.771873751498202, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9441 + }, + { + "epoch": 3.772273272073512, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 9442 + }, + { + "epoch": 3.7726727926488213, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9443 + }, + { + "epoch": 3.773072313224131, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9444 + }, + { + "epoch": 3.773471833799441, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9445 + }, + { + "epoch": 3.77387135437475, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9446 + }, + { + "epoch": 3.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9447 + }, + { + "epoch": 3.7746703955253693, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9448 + }, + { + "epoch": 3.775069916100679, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 9449 + }, + { + "epoch": 3.775469436675989, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9450 + }, + { + "epoch": 3.7758689572512987, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9451 + }, + { + "epoch": 3.776268477826608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9452 + }, + { + "epoch": 3.776667998401918, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9453 + }, + { + "epoch": 3.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9454 + }, + { + "epoch": 3.777467039552537, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9455 + }, + { + "epoch": 3.7778665601278467, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9456 + }, + { + "epoch": 3.778266080703156, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9457 + }, + { + "epoch": 3.778665601278466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9458 + }, + { + "epoch": 3.7790651218537756, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9459 + }, + { + "epoch": 3.779464642429085, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9460 + }, + { + "epoch": 3.7798641630043948, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9461 + }, + { + "epoch": 3.7802636835797045, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9462 + }, + { + "epoch": 3.780663204155014, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9463 + }, + { + "epoch": 3.7810627247303237, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9464 + }, + { + "epoch": 3.781462245305633, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9465 + }, + { + "epoch": 3.781861765880943, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9466 + }, + { + "epoch": 3.7822612864562526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9467 + }, + { + "epoch": 3.7826608070315624, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9468 + }, + { + "epoch": 3.7830603276068717, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9469 + }, + { + "epoch": 3.7834598481821815, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9470 + }, + { + "epoch": 3.783859368757491, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 9471 + }, + { + "epoch": 3.7842588893328006, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9472 + }, + { + "epoch": 3.7846584099081104, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9473 + }, + { + "epoch": 3.7850579304834198, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9474 + }, + { + "epoch": 3.7854574510587295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9475 + }, + { + "epoch": 3.785856971634039, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9476 + }, + { + "epoch": 3.7862564922093487, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9477 + }, + { + "epoch": 3.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9478 + }, + { + "epoch": 3.7870555333599683, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9479 + }, + { + "epoch": 3.7874550539352776, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9480 + }, + { + "epoch": 3.7878545745105874, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9481 + }, + { + "epoch": 3.7882540950858967, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 9482 + }, + { + "epoch": 3.7886536156612065, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9483 + }, + { + "epoch": 3.7890531362365163, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9484 + }, + { + "epoch": 3.789452656811826, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9485 + }, + { + "epoch": 3.7898521773871354, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 9486 + }, + { + "epoch": 3.790251697962445, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9487 + }, + { + "epoch": 3.7906512185377546, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9488 + }, + { + "epoch": 3.7910507391130643, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9489 + }, + { + "epoch": 3.791450259688374, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 9490 + }, + { + "epoch": 3.7918497802636835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9491 + }, + { + "epoch": 3.7922493008389933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9492 + }, + { + "epoch": 3.7926488214143026, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9493 + }, + { + "epoch": 3.7930483419896124, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9494 + }, + { + "epoch": 3.793447862564922, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9495 + }, + { + "epoch": 3.793847383140232, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9496 + }, + { + "epoch": 3.7942469037155413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9497 + }, + { + "epoch": 3.794646424290851, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9498 + }, + { + "epoch": 3.7950459448661604, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9499 + }, + { + "epoch": 3.79544546544147, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9500 + }, + { + "epoch": 3.79584498601678, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9501 + }, + { + "epoch": 3.7962445065920893, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9502 + }, + { + "epoch": 3.796644027167399, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9503 + }, + { + "epoch": 3.797043547742709, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9504 + }, + { + "epoch": 3.7974430683180183, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9505 + }, + { + "epoch": 3.797842588893328, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9506 + }, + { + "epoch": 3.798242109468638, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9507 + }, + { + "epoch": 3.798641630043947, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9508 + }, + { + "epoch": 3.799041150619257, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 9509 + }, + { + "epoch": 3.7994406711945663, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 9510 + }, + { + "epoch": 3.799840191769876, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9511 + }, + { + "epoch": 3.800239712345186, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9512 + }, + { + "epoch": 3.8006392329204957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9513 + }, + { + "epoch": 3.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9514 + }, + { + "epoch": 3.801438274071115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9515 + }, + { + "epoch": 3.801837794646424, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9516 + }, + { + "epoch": 3.802237315221734, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9517 + }, + { + "epoch": 3.8026368357970437, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9518 + }, + { + "epoch": 3.803036356372353, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9519 + }, + { + "epoch": 3.803435876947663, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9520 + }, + { + "epoch": 3.803835397522972, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9521 + }, + { + "epoch": 3.804234918098282, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9522 + }, + { + "epoch": 3.8046344386735917, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9523 + }, + { + "epoch": 3.8050339592489015, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9524 + }, + { + "epoch": 3.805433479824211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9525 + }, + { + "epoch": 3.8058330003995207, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9526 + }, + { + "epoch": 3.80623252097483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9527 + }, + { + "epoch": 3.80663204155014, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9528 + }, + { + "epoch": 3.8070315621254496, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9529 + }, + { + "epoch": 3.8074310827007594, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9530 + }, + { + "epoch": 3.8078306032760687, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9531 + }, + { + "epoch": 3.8082301238513785, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9532 + }, + { + "epoch": 3.808629644426688, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9533 + }, + { + "epoch": 3.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9534 + }, + { + "epoch": 3.8094286855773074, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9535 + }, + { + "epoch": 3.8098282061526167, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9536 + }, + { + "epoch": 3.8102277267279265, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9537 + }, + { + "epoch": 3.810627247303236, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9538 + }, + { + "epoch": 3.8110267678785457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9539 + }, + { + "epoch": 3.8114262884538554, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9540 + }, + { + "epoch": 3.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9541 + }, + { + "epoch": 3.8122253296044746, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9542 + }, + { + "epoch": 3.8126248501797844, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 9543 + }, + { + "epoch": 3.8130243707550937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9544 + }, + { + "epoch": 3.8134238913304035, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9545 + }, + { + "epoch": 3.8138234119057133, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9546 + }, + { + "epoch": 3.8142229324810226, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9547 + }, + { + "epoch": 3.8146224530563324, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9548 + }, + { + "epoch": 3.815021973631642, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9549 + }, + { + "epoch": 3.8154214942069515, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9550 + }, + { + "epoch": 3.8158210147822613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9551 + }, + { + "epoch": 3.816220535357571, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9552 + }, + { + "epoch": 3.8166200559328805, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9553 + }, + { + "epoch": 3.8170195765081902, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 9554 + }, + { + "epoch": 3.8174190970834996, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9555 + }, + { + "epoch": 3.8178186176588094, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9556 + }, + { + "epoch": 3.818218138234119, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9557 + }, + { + "epoch": 3.818617658809429, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9558 + }, + { + "epoch": 3.8190171793847383, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9559 + }, + { + "epoch": 3.819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9560 + }, + { + "epoch": 3.8198162205353574, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9561 + }, + { + "epoch": 3.820215741110667, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9562 + }, + { + "epoch": 3.820615261685977, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9563 + }, + { + "epoch": 3.8210147822612863, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9564 + }, + { + "epoch": 3.821414302836596, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9565 + }, + { + "epoch": 3.8218138234119055, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9566 + }, + { + "epoch": 3.8222133439872152, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9567 + }, + { + "epoch": 3.822612864562525, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9568 + }, + { + "epoch": 3.823012385137835, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 9569 + }, + { + "epoch": 3.823411905713144, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9570 + }, + { + "epoch": 3.823811426288454, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9571 + }, + { + "epoch": 3.8242109468637633, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9572 + }, + { + "epoch": 3.824610467439073, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9573 + }, + { + "epoch": 3.825009988014383, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9574 + }, + { + "epoch": 3.8254095085896926, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9575 + }, + { + "epoch": 3.825809029165002, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9576 + }, + { + "epoch": 3.8262085497403118, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 9577 + }, + { + "epoch": 3.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9578 + }, + { + "epoch": 3.827007590890931, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9579 + }, + { + "epoch": 3.8274071114662407, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9580 + }, + { + "epoch": 3.82780663204155, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9581 + }, + { + "epoch": 3.82820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9582 + }, + { + "epoch": 3.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9583 + }, + { + "epoch": 3.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 9584 + }, + { + "epoch": 3.8294047143427887, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9585 + }, + { + "epoch": 3.8298042349180985, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9586 + }, + { + "epoch": 3.830203755493408, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9587 + }, + { + "epoch": 3.8306032760687176, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9588 + }, + { + "epoch": 3.831002796644027, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9589 + }, + { + "epoch": 3.8314023172193368, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9590 + }, + { + "epoch": 3.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9591 + }, + { + "epoch": 3.832201358369956, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9592 + }, + { + "epoch": 3.8326008789452657, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9593 + }, + { + "epoch": 3.8330003995205755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9594 + }, + { + "epoch": 3.833399920095885, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9595 + }, + { + "epoch": 3.8337994406711946, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9596 + }, + { + "epoch": 3.8341989612465044, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9597 + }, + { + "epoch": 3.8345984818218137, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 9598 + }, + { + "epoch": 3.8349980023971235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9599 + }, + { + "epoch": 3.835397522972433, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9600 + }, + { + "epoch": 3.8357970435477426, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9601 + }, + { + "epoch": 3.8361965641230524, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9602 + }, + { + "epoch": 3.836596084698362, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9603 + }, + { + "epoch": 3.8369956052736716, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9604 + }, + { + "epoch": 3.8373951258489813, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9605 + }, + { + "epoch": 3.8377946464242907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9606 + }, + { + "epoch": 3.8381941669996005, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9607 + }, + { + "epoch": 3.8385936875749103, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9608 + }, + { + "epoch": 3.8389932081502196, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9609 + }, + { + "epoch": 3.8393927287255294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9610 + }, + { + "epoch": 3.8397922493008387, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9611 + }, + { + "epoch": 3.8401917698761485, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9612 + }, + { + "epoch": 3.8405912904514583, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9613 + }, + { + "epoch": 3.840990811026768, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9614 + }, + { + "epoch": 3.8413903316020774, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9615 + }, + { + "epoch": 3.841789852177387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9616 + }, + { + "epoch": 3.8421893727526966, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9617 + }, + { + "epoch": 3.8425888933280063, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9618 + }, + { + "epoch": 3.842988413903316, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 9619 + }, + { + "epoch": 3.843387934478626, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9620 + }, + { + "epoch": 3.8437874550539353, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9621 + }, + { + "epoch": 3.844186975629245, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9622 + }, + { + "epoch": 3.8445864962045544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9623 + }, + { + "epoch": 3.844986016779864, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 9624 + }, + { + "epoch": 3.845385537355174, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9625 + }, + { + "epoch": 3.8457850579304833, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 9626 + }, + { + "epoch": 3.846184578505793, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9627 + }, + { + "epoch": 3.8465840990811024, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9628 + }, + { + "epoch": 3.8469836196564122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9629 + }, + { + "epoch": 3.847383140231722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9630 + }, + { + "epoch": 3.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9631 + }, + { + "epoch": 3.848182181382341, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9632 + }, + { + "epoch": 3.848581701957651, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9633 + }, + { + "epoch": 3.8489812225329603, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 9634 + }, + { + "epoch": 3.84938074310827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9635 + }, + { + "epoch": 3.84978026368358, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 9636 + }, + { + "epoch": 3.850179784258889, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9637 + }, + { + "epoch": 3.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9638 + }, + { + "epoch": 3.8509788254095088, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9639 + }, + { + "epoch": 3.851378345984818, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9640 + }, + { + "epoch": 3.851777866560128, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9641 + }, + { + "epoch": 3.8521773871354377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 9642 + }, + { + "epoch": 3.852576907710747, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9643 + }, + { + "epoch": 3.852976428286057, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9644 + }, + { + "epoch": 3.853375948861366, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9645 + }, + { + "epoch": 3.853775469436676, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9646 + }, + { + "epoch": 3.8541749900119857, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9647 + }, + { + "epoch": 3.8545745105872955, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9648 + }, + { + "epoch": 3.854974031162605, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9649 + }, + { + "epoch": 3.8553735517379146, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9650 + }, + { + "epoch": 3.855773072313224, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9651 + }, + { + "epoch": 3.8561725928885338, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9652 + }, + { + "epoch": 3.8565721134638435, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9653 + }, + { + "epoch": 3.856971634039153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9654 + }, + { + "epoch": 3.8573711546144627, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9655 + }, + { + "epoch": 3.857770675189772, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9656 + }, + { + "epoch": 3.858170195765082, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9657 + }, + { + "epoch": 3.8585697163403916, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 9658 + }, + { + "epoch": 3.8589692369157014, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9659 + }, + { + "epoch": 3.8593687574910107, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9660 + }, + { + "epoch": 3.8597682780663205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9661 + }, + { + "epoch": 3.86016779864163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9662 + }, + { + "epoch": 3.8605673192169396, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9663 + }, + { + "epoch": 3.8609668397922494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9664 + }, + { + "epoch": 3.861366360367559, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9665 + }, + { + "epoch": 3.8617658809428685, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9666 + }, + { + "epoch": 3.8621654015181783, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9667 + }, + { + "epoch": 3.8625649220934877, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9668 + }, + { + "epoch": 3.8629644426687975, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9669 + }, + { + "epoch": 3.8633639632441072, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9670 + }, + { + "epoch": 3.8637634838194166, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9671 + }, + { + "epoch": 3.8641630043947264, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9672 + }, + { + "epoch": 3.8645625249700357, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9673 + }, + { + "epoch": 3.8649620455453455, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9674 + }, + { + "epoch": 3.8653615661206553, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9675 + }, + { + "epoch": 3.865761086695965, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 9676 + }, + { + "epoch": 3.8661606072712744, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 9677 + }, + { + "epoch": 3.866560127846584, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9678 + }, + { + "epoch": 3.8669596484218935, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9679 + }, + { + "epoch": 3.8673591689972033, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9680 + }, + { + "epoch": 3.867758689572513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 9681 + }, + { + "epoch": 3.8681582101478225, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9682 + }, + { + "epoch": 3.8685577307231322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9683 + }, + { + "epoch": 3.868957251298442, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9684 + }, + { + "epoch": 3.8693567718737514, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9685 + }, + { + "epoch": 3.869756292449061, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9686 + }, + { + "epoch": 3.870155813024371, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9687 + }, + { + "epoch": 3.8705553335996803, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 9688 + }, + { + "epoch": 3.87095485417499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9689 + }, + { + "epoch": 3.8713543747502994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9690 + }, + { + "epoch": 3.871753895325609, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9691 + }, + { + "epoch": 3.872153415900919, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9692 + }, + { + "epoch": 3.872552936476229, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9693 + }, + { + "epoch": 3.872952457051538, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9694 + }, + { + "epoch": 3.873351977626848, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9695 + }, + { + "epoch": 3.8737514982021573, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9696 + }, + { + "epoch": 3.874151018777467, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9697 + }, + { + "epoch": 3.874550539352777, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 9698 + }, + { + "epoch": 3.874950059928086, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9699 + }, + { + "epoch": 3.875349580503396, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9700 + }, + { + "epoch": 3.8757491010787053, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9701 + }, + { + "epoch": 3.876148621654015, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9702 + }, + { + "epoch": 3.876548142229325, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9703 + }, + { + "epoch": 3.8769476628046347, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9704 + }, + { + "epoch": 3.877347183379944, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9705 + }, + { + "epoch": 3.877746703955254, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9706 + }, + { + "epoch": 3.878146224530563, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9707 + }, + { + "epoch": 3.878545745105873, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9708 + }, + { + "epoch": 3.8789452656811827, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9709 + }, + { + "epoch": 3.8793447862564925, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9710 + }, + { + "epoch": 3.879744306831802, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9711 + }, + { + "epoch": 3.8801438274071116, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9712 + }, + { + "epoch": 3.880543347982421, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9713 + }, + { + "epoch": 3.8809428685577307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9714 + }, + { + "epoch": 3.8813423891330405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9715 + }, + { + "epoch": 3.88174190970835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9716 + }, + { + "epoch": 3.8821414302836597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9717 + }, + { + "epoch": 3.882540950858969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9718 + }, + { + "epoch": 3.882940471434279, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9719 + }, + { + "epoch": 3.8833399920095886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9720 + }, + { + "epoch": 3.8837395125848984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9721 + }, + { + "epoch": 3.8841390331602077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9722 + }, + { + "epoch": 3.8845385537355175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9723 + }, + { + "epoch": 3.884938074310827, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9724 + }, + { + "epoch": 3.8853375948861366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9725 + }, + { + "epoch": 3.8857371154614464, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9726 + }, + { + "epoch": 3.8861366360367557, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9727 + }, + { + "epoch": 3.8865361566120655, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9728 + }, + { + "epoch": 3.8869356771873753, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9729 + }, + { + "epoch": 3.8873351977626847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9730 + }, + { + "epoch": 3.8877347183379944, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9731 + }, + { + "epoch": 3.8881342389133042, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9732 + }, + { + "epoch": 3.8885337594886136, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9733 + }, + { + "epoch": 3.8889332800639234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9734 + }, + { + "epoch": 3.8893328006392327, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9735 + }, + { + "epoch": 3.8897323212145425, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9736 + }, + { + "epoch": 3.8901318417898523, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9737 + }, + { + "epoch": 3.890531362365162, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9738 + }, + { + "epoch": 3.8909308829404714, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9739 + }, + { + "epoch": 3.891330403515781, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9740 + }, + { + "epoch": 3.8917299240910905, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9741 + }, + { + "epoch": 3.8921294446664003, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9742 + }, + { + "epoch": 3.89252896524171, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9743 + }, + { + "epoch": 3.8929284858170194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9744 + }, + { + "epoch": 3.8933280063923292, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 9745 + }, + { + "epoch": 3.8937275269676386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9746 + }, + { + "epoch": 3.8941270475429484, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9747 + }, + { + "epoch": 3.894526568118258, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9748 + }, + { + "epoch": 3.894926088693568, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9749 + }, + { + "epoch": 3.8953256092688773, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9750 + }, + { + "epoch": 3.895725129844187, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9751 + }, + { + "epoch": 3.8961246504194964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9752 + }, + { + "epoch": 3.896524170994806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9753 + }, + { + "epoch": 3.896923691570116, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9754 + }, + { + "epoch": 3.8973232121454258, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9755 + }, + { + "epoch": 3.897722732720735, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9756 + }, + { + "epoch": 3.898122253296045, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9757 + }, + { + "epoch": 3.8985217738713542, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9758 + }, + { + "epoch": 3.898921294446664, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9759 + }, + { + "epoch": 3.899320815021974, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9760 + }, + { + "epoch": 3.899720335597283, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 9761 + }, + { + "epoch": 3.900119856172593, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9762 + }, + { + "epoch": 3.9005193767479023, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9763 + }, + { + "epoch": 3.900918897323212, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9764 + }, + { + "epoch": 3.901318417898522, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9765 + }, + { + "epoch": 3.9017179384738316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9766 + }, + { + "epoch": 3.902117459049141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9767 + }, + { + "epoch": 3.9025169796244508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9768 + }, + { + "epoch": 3.90291650019976, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9769 + }, + { + "epoch": 3.90331602077507, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9770 + }, + { + "epoch": 3.9037155413503797, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9771 + }, + { + "epoch": 3.904115061925689, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9772 + }, + { + "epoch": 3.904514582500999, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9773 + }, + { + "epoch": 3.9049141030763086, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9774 + }, + { + "epoch": 3.905313623651618, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9775 + }, + { + "epoch": 3.9057131442269277, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9776 + }, + { + "epoch": 3.9061126648022375, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9777 + }, + { + "epoch": 3.906512185377547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9778 + }, + { + "epoch": 3.9069117059528566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9779 + }, + { + "epoch": 3.907311226528166, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9780 + }, + { + "epoch": 3.9077107471034758, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9781 + }, + { + "epoch": 3.9081102676787856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9782 + }, + { + "epoch": 3.9085097882540953, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9783 + }, + { + "epoch": 3.9089093088294047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9784 + }, + { + "epoch": 3.9093088294047145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9785 + }, + { + "epoch": 3.909708349980024, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9786 + }, + { + "epoch": 3.9101078705553336, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9787 + }, + { + "epoch": 3.9105073911306434, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9788 + }, + { + "epoch": 3.9109069117059527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9789 + }, + { + "epoch": 3.9113064322812625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9790 + }, + { + "epoch": 3.911705952856572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9791 + }, + { + "epoch": 3.9121054734318816, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9792 + }, + { + "epoch": 3.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9793 + }, + { + "epoch": 3.912904514582501, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9794 + }, + { + "epoch": 3.9133040351578106, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9795 + }, + { + "epoch": 3.9137035557331203, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9796 + }, + { + "epoch": 3.9141030763084297, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 9797 + }, + { + "epoch": 3.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9798 + }, + { + "epoch": 3.9149021174590493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9799 + }, + { + "epoch": 3.915301638034359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9800 + }, + { + "epoch": 3.9157011586096684, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9801 + }, + { + "epoch": 3.916100679184978, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9802 + }, + { + "epoch": 3.9165001997602875, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9803 + }, + { + "epoch": 3.9168997203355973, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9804 + }, + { + "epoch": 3.917299240910907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9805 + }, + { + "epoch": 3.9176987614862164, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9806 + }, + { + "epoch": 3.918098282061526, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9807 + }, + { + "epoch": 3.9184978026368356, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 9808 + }, + { + "epoch": 3.9188973232121453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 9809 + }, + { + "epoch": 3.919296843787455, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9810 + }, + { + "epoch": 3.919696364362765, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9811 + }, + { + "epoch": 3.9200958849380743, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9812 + }, + { + "epoch": 3.920495405513384, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9813 + }, + { + "epoch": 3.9208949260886934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9814 + }, + { + "epoch": 3.921294446664003, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9815 + }, + { + "epoch": 3.921693967239313, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9816 + }, + { + "epoch": 3.9220934878146223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9817 + }, + { + "epoch": 3.922493008389932, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9818 + }, + { + "epoch": 3.922892528965242, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9819 + }, + { + "epoch": 3.923292049540551, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9820 + }, + { + "epoch": 3.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9821 + }, + { + "epoch": 3.924091090691171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9822 + }, + { + "epoch": 3.92449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9823 + }, + { + "epoch": 3.92489013184179, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9824 + }, + { + "epoch": 3.9252896524170993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9825 + }, + { + "epoch": 3.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9826 + }, + { + "epoch": 3.926088693567719, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9827 + }, + { + "epoch": 3.9264882141430286, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9828 + }, + { + "epoch": 3.926887734718338, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9829 + }, + { + "epoch": 3.9272872552936477, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9830 + }, + { + "epoch": 3.927686775868957, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9831 + }, + { + "epoch": 3.928086296444267, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 9832 + }, + { + "epoch": 3.9284858170195767, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9833 + }, + { + "epoch": 3.928885337594886, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9834 + }, + { + "epoch": 3.929284858170196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9835 + }, + { + "epoch": 3.929684378745505, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9836 + }, + { + "epoch": 3.930083899320815, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9837 + }, + { + "epoch": 3.9304834198961247, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9838 + }, + { + "epoch": 3.9308829404714345, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9839 + }, + { + "epoch": 3.931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9840 + }, + { + "epoch": 3.9316819816220536, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9841 + }, + { + "epoch": 3.932081502197363, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9842 + }, + { + "epoch": 3.9324810227726728, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9843 + }, + { + "epoch": 3.9328805433479825, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9844 + }, + { + "epoch": 3.9332800639232923, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9845 + }, + { + "epoch": 3.9336795844986017, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9846 + }, + { + "epoch": 3.9340791050739115, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9847 + }, + { + "epoch": 3.934478625649221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9848 + }, + { + "epoch": 3.9348781462245306, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9849 + }, + { + "epoch": 3.9352776667998404, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9850 + }, + { + "epoch": 3.9356771873751497, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9851 + }, + { + "epoch": 3.9360767079504595, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9852 + }, + { + "epoch": 3.936476228525769, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9853 + }, + { + "epoch": 3.9368757491010786, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9854 + }, + { + "epoch": 3.9372752696763884, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9855 + }, + { + "epoch": 3.937674790251698, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9856 + }, + { + "epoch": 3.9380743108270075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9857 + }, + { + "epoch": 3.9384738314023173, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9858 + }, + { + "epoch": 3.9388733519776267, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9859 + }, + { + "epoch": 3.9392728725529365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9860 + }, + { + "epoch": 3.9396723931282462, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9861 + }, + { + "epoch": 3.9400719137035556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9862 + }, + { + "epoch": 3.9404714342788654, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9863 + }, + { + "epoch": 3.9408709548541747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9864 + }, + { + "epoch": 3.9412704754294845, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9865 + }, + { + "epoch": 3.9416699960047943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9866 + }, + { + "epoch": 3.942069516580104, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9867 + }, + { + "epoch": 3.9424690371554134, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9868 + }, + { + "epoch": 3.942868557730723, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9869 + }, + { + "epoch": 3.9432680783060325, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9870 + }, + { + "epoch": 3.9436675988813423, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9871 + }, + { + "epoch": 3.944067119456652, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9872 + }, + { + "epoch": 3.944466640031962, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9873 + }, + { + "epoch": 3.9448661606072712, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9874 + }, + { + "epoch": 3.945265681182581, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9875 + }, + { + "epoch": 3.9456652017578904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9876 + }, + { + "epoch": 3.9460647223332, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 9877 + }, + { + "epoch": 3.94646424290851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9878 + }, + { + "epoch": 3.9468637634838193, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9879 + }, + { + "epoch": 3.947263284059129, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9880 + }, + { + "epoch": 3.9476628046344384, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9881 + }, + { + "epoch": 3.948062325209748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9882 + }, + { + "epoch": 3.948461845785058, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9883 + }, + { + "epoch": 3.9488613663603678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9884 + }, + { + "epoch": 3.949260886935677, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9885 + }, + { + "epoch": 3.949660407510987, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9886 + }, + { + "epoch": 3.9500599280862962, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9887 + }, + { + "epoch": 3.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9888 + }, + { + "epoch": 3.950858969236916, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9889 + }, + { + "epoch": 3.9512584898122256, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9890 + }, + { + "epoch": 3.951658010387535, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9891 + }, + { + "epoch": 3.9520575309628447, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9892 + }, + { + "epoch": 3.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9893 + }, + { + "epoch": 3.952856572113464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9894 + }, + { + "epoch": 3.9532560926887736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9895 + }, + { + "epoch": 3.953655613264083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9896 + }, + { + "epoch": 3.9540551338393928, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9897 + }, + { + "epoch": 3.954454654414702, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9898 + }, + { + "epoch": 3.954854174990012, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9899 + }, + { + "epoch": 3.9552536955653217, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9900 + }, + { + "epoch": 3.9556532161406315, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9901 + }, + { + "epoch": 3.956052736715941, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9902 + }, + { + "epoch": 3.9564522572912506, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9903 + }, + { + "epoch": 3.95685177786656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9904 + }, + { + "epoch": 3.9572512984418697, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9905 + }, + { + "epoch": 3.9576508190171795, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9906 + }, + { + "epoch": 3.958050339592489, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9907 + }, + { + "epoch": 3.9584498601677987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9908 + }, + { + "epoch": 3.958849380743108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9909 + }, + { + "epoch": 3.959248901318418, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9910 + }, + { + "epoch": 3.9596484218937276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 9911 + }, + { + "epoch": 3.9600479424690374, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9912 + }, + { + "epoch": 3.9604474630443467, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9913 + }, + { + "epoch": 3.9608469836196565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9914 + }, + { + "epoch": 3.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9915 + }, + { + "epoch": 3.9616460247702756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9916 + }, + { + "epoch": 3.9620455453455854, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9917 + }, + { + "epoch": 3.962445065920895, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9918 + }, + { + "epoch": 3.9628445864962045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9919 + }, + { + "epoch": 3.9632441070715143, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9920 + }, + { + "epoch": 3.9636436276468237, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9921 + }, + { + "epoch": 3.9640431482221334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9922 + }, + { + "epoch": 3.9644426687974432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9923 + }, + { + "epoch": 3.9648421893727526, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9924 + }, + { + "epoch": 3.9652417099480624, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9925 + }, + { + "epoch": 3.9656412305233717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 9926 + }, + { + "epoch": 3.9660407510986815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9927 + }, + { + "epoch": 3.9664402716739913, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9928 + }, + { + "epoch": 3.966839792249301, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9929 + }, + { + "epoch": 3.9672393128246104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9930 + }, + { + "epoch": 3.96763883339992, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9931 + }, + { + "epoch": 3.9680383539752295, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9932 + }, + { + "epoch": 3.9684378745505393, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9933 + }, + { + "epoch": 3.968837395125849, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 9934 + }, + { + "epoch": 3.969236915701159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 9935 + }, + { + "epoch": 3.9696364362764682, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9936 + }, + { + "epoch": 3.970035956851778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 9937 + }, + { + "epoch": 3.9704354774270874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9938 + }, + { + "epoch": 3.970834998002397, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9939 + }, + { + "epoch": 3.971234518577707, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9940 + }, + { + "epoch": 3.9716340391530163, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9941 + }, + { + "epoch": 3.972033559728326, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9942 + }, + { + "epoch": 3.9724330803036354, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9943 + }, + { + "epoch": 3.972832600878945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 9944 + }, + { + "epoch": 3.973232121454255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9945 + }, + { + "epoch": 3.9736316420295648, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9946 + }, + { + "epoch": 3.974031162604874, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9947 + }, + { + "epoch": 3.974430683180184, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9948 + }, + { + "epoch": 3.9748302037554932, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9949 + }, + { + "epoch": 3.975229724330803, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9950 + }, + { + "epoch": 3.975629244906113, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9951 + }, + { + "epoch": 3.976028765481422, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9952 + }, + { + "epoch": 3.976428286056732, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9953 + }, + { + "epoch": 3.9768278066320413, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9954 + }, + { + "epoch": 3.977227327207351, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9955 + }, + { + "epoch": 3.977626847782661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 9956 + }, + { + "epoch": 3.9780263683579706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9957 + }, + { + "epoch": 3.97842588893328, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9958 + }, + { + "epoch": 3.9788254095085898, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9959 + }, + { + "epoch": 3.979224930083899, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9960 + }, + { + "epoch": 3.979624450659209, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9961 + }, + { + "epoch": 3.9800239712345187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9962 + }, + { + "epoch": 3.9804234918098285, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9963 + }, + { + "epoch": 3.980823012385138, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 9964 + }, + { + "epoch": 3.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9965 + }, + { + "epoch": 3.981622053535757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9966 + }, + { + "epoch": 3.9820215741110667, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9967 + }, + { + "epoch": 3.9824210946863765, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9968 + }, + { + "epoch": 3.982820615261686, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9969 + }, + { + "epoch": 3.9832201358369956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9970 + }, + { + "epoch": 3.983619656412305, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9971 + }, + { + "epoch": 3.9840191769876148, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9972 + }, + { + "epoch": 3.9844186975629245, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9973 + }, + { + "epoch": 3.9848182181382343, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9974 + }, + { + "epoch": 3.9852177387135437, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9975 + }, + { + "epoch": 3.9856172592888535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9976 + }, + { + "epoch": 3.986016779864163, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9977 + }, + { + "epoch": 3.9864163004394726, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9978 + }, + { + "epoch": 3.9868158210147824, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9979 + }, + { + "epoch": 3.987215341590092, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9980 + }, + { + "epoch": 3.9876148621654015, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9981 + }, + { + "epoch": 3.9880143827407113, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9982 + }, + { + "epoch": 3.9884139033160206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9983 + }, + { + "epoch": 3.9888134238913304, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9984 + }, + { + "epoch": 3.98921294446664, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9985 + }, + { + "epoch": 3.9896124650419496, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9986 + }, + { + "epoch": 3.9900119856172593, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 9987 + }, + { + "epoch": 3.9904115061925687, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9988 + }, + { + "epoch": 3.9908110267678785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9989 + }, + { + "epoch": 3.9912105473431883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9990 + }, + { + "epoch": 3.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9991 + }, + { + "epoch": 3.9920095884938074, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9992 + }, + { + "epoch": 3.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9993 + }, + { + "epoch": 3.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9994 + }, + { + "epoch": 3.9932081502197363, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9995 + }, + { + "epoch": 3.993607670795046, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9996 + }, + { + "epoch": 3.9940071913703554, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9997 + }, + { + "epoch": 3.994406711945665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9998 + }, + { + "epoch": 3.9948062325209746, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9999 + }, + { + "epoch": 3.9952057530962843, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10000 + }, + { + "epoch": 3.995605273671594, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10001 + }, + { + "epoch": 3.996004794246904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10002 + }, + { + "epoch": 3.9964043148222133, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10003 + }, + { + "epoch": 3.996803835397523, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10004 + }, + { + "epoch": 3.9972033559728324, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10005 + }, + { + "epoch": 3.997602876548142, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10006 + }, + { + "epoch": 3.998002397123452, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10007 + }, + { + "epoch": 3.9984019176987617, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10008 + }, + { + "epoch": 3.998801438274071, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10009 + }, + { + "epoch": 3.999200958849381, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 10010 + }, + { + "epoch": 3.99960047942469, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 10011 + }, + { + "epoch": 4.0, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10012 + }, + { + "epoch": 4.00039952057531, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10013 + }, + { + "epoch": 4.00079904115062, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10014 + }, + { + "epoch": 4.0011985617259285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10015 + }, + { + "epoch": 4.001598082301238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10016 + }, + { + "epoch": 4.001997602876548, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10017 + }, + { + "epoch": 4.002397123451858, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10018 + }, + { + "epoch": 4.002796644027168, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10019 + }, + { + "epoch": 4.003196164602477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10020 + }, + { + "epoch": 4.003595685177786, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10021 + }, + { + "epoch": 4.003995205753096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10022 + }, + { + "epoch": 4.004394726328406, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10023 + }, + { + "epoch": 4.004794246903716, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10024 + }, + { + "epoch": 4.0051937674790254, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10025 + }, + { + "epoch": 4.005593288054335, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10026 + }, + { + "epoch": 4.005992808629644, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10027 + }, + { + "epoch": 4.006392329204954, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10028 + }, + { + "epoch": 4.006791849780264, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10029 + }, + { + "epoch": 4.0071913703555735, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10030 + }, + { + "epoch": 4.007590890930883, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10031 + }, + { + "epoch": 4.007990411506192, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10032 + }, + { + "epoch": 4.008389932081502, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10033 + }, + { + "epoch": 4.008789452656812, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10034 + }, + { + "epoch": 4.0091889732321215, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10035 + }, + { + "epoch": 4.009588493807431, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 10036 + }, + { + "epoch": 4.009988014382741, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10037 + }, + { + "epoch": 4.01038753495805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 10038 + }, + { + "epoch": 4.01078705553336, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10039 + }, + { + "epoch": 4.01118657610867, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10040 + }, + { + "epoch": 4.011586096683979, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10041 + }, + { + "epoch": 4.011985617259289, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10042 + }, + { + "epoch": 4.012385137834598, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10043 + }, + { + "epoch": 4.012784658409908, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10044 + }, + { + "epoch": 4.013184178985218, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10045 + }, + { + "epoch": 4.013583699560527, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10046 + }, + { + "epoch": 4.013983220135837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10047 + }, + { + "epoch": 4.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10048 + }, + { + "epoch": 4.014782261286456, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 10049 + }, + { + "epoch": 4.015181781861766, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10050 + }, + { + "epoch": 4.0155813024370755, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10051 + }, + { + "epoch": 4.015980823012385, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10052 + }, + { + "epoch": 4.016380343587695, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10053 + }, + { + "epoch": 4.016779864163005, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10054 + }, + { + "epoch": 4.017179384738314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10055 + }, + { + "epoch": 4.0175789053136235, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10056 + }, + { + "epoch": 4.017978425888933, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10057 + }, + { + "epoch": 4.018377946464243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10058 + }, + { + "epoch": 4.018777467039553, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10059 + }, + { + "epoch": 4.019176987614862, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 10060 + }, + { + "epoch": 4.0195765081901715, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10061 + }, + { + "epoch": 4.019976028765481, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10062 + }, + { + "epoch": 4.020375549340791, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10063 + }, + { + "epoch": 4.020775069916101, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10064 + }, + { + "epoch": 4.021174590491411, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10065 + }, + { + "epoch": 4.02157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10066 + }, + { + "epoch": 4.021973631642029, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10067 + }, + { + "epoch": 4.022373152217339, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10068 + }, + { + "epoch": 4.022772672792649, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10069 + }, + { + "epoch": 4.023172193367959, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10070 + }, + { + "epoch": 4.0235717139432685, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10071 + }, + { + "epoch": 4.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10072 + }, + { + "epoch": 4.024370755093887, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10073 + }, + { + "epoch": 4.024770275669197, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10074 + }, + { + "epoch": 4.025169796244507, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10075 + }, + { + "epoch": 4.025569316819817, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10076 + }, + { + "epoch": 4.0259688373951255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10077 + }, + { + "epoch": 4.026368357970435, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10078 + }, + { + "epoch": 4.026767878545745, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 10079 + }, + { + "epoch": 4.027167399121055, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10080 + }, + { + "epoch": 4.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10081 + }, + { + "epoch": 4.027966440271674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10082 + }, + { + "epoch": 4.028365960846983, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10083 + }, + { + "epoch": 4.028765481422293, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10084 + }, + { + "epoch": 4.029165001997603, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10085 + }, + { + "epoch": 4.029564522572913, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10086 + }, + { + "epoch": 4.029964043148222, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10087 + }, + { + "epoch": 4.030363563723531, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10088 + }, + { + "epoch": 4.030763084298841, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10089 + }, + { + "epoch": 4.031162604874151, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10090 + }, + { + "epoch": 4.031562125449461, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 10091 + }, + { + "epoch": 4.0319616460247705, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10092 + }, + { + "epoch": 4.03236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10093 + }, + { + "epoch": 4.032760687175389, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10094 + }, + { + "epoch": 4.033160207750699, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10095 + }, + { + "epoch": 4.033559728326009, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10096 + }, + { + "epoch": 4.0339592489013185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10097 + }, + { + "epoch": 4.034358769476628, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10098 + }, + { + "epoch": 4.034758290051938, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10099 + }, + { + "epoch": 4.035157810627247, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10100 + }, + { + "epoch": 4.035557331202557, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10101 + }, + { + "epoch": 4.035956851777867, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10102 + }, + { + "epoch": 4.036356372353176, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10103 + }, + { + "epoch": 4.036755892928486, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10104 + }, + { + "epoch": 4.037155413503795, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10105 + }, + { + "epoch": 4.037554934079105, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10106 + }, + { + "epoch": 4.037954454654415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 10107 + }, + { + "epoch": 4.038353975229724, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10108 + }, + { + "epoch": 4.038753495805034, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10109 + }, + { + "epoch": 4.039153016380344, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10110 + }, + { + "epoch": 4.039552536955653, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10111 + }, + { + "epoch": 4.039952057530963, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10112 + }, + { + "epoch": 4.040351578106272, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10113 + }, + { + "epoch": 4.040751098681582, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10114 + }, + { + "epoch": 4.041150619256892, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10115 + }, + { + "epoch": 4.041550139832202, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10116 + }, + { + "epoch": 4.041949660407511, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10117 + }, + { + "epoch": 4.0423491809828205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10118 + }, + { + "epoch": 4.04274870155813, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10119 + }, + { + "epoch": 4.04314822213344, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10120 + }, + { + "epoch": 4.04354774270875, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10121 + }, + { + "epoch": 4.043947263284059, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10122 + }, + { + "epoch": 4.0443467838593685, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10123 + }, + { + "epoch": 4.044746304434678, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10124 + }, + { + "epoch": 4.045145825009988, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10125 + }, + { + "epoch": 4.045545345585298, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10126 + }, + { + "epoch": 4.045944866160608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10127 + }, + { + "epoch": 4.046344386735917, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10128 + }, + { + "epoch": 4.046743907311226, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10129 + }, + { + "epoch": 4.047143427886536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10130 + }, + { + "epoch": 4.047542948461846, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10131 + }, + { + "epoch": 4.047942469037156, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10132 + }, + { + "epoch": 4.048341989612465, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10133 + }, + { + "epoch": 4.048741510187774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10134 + }, + { + "epoch": 4.049141030763084, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10135 + }, + { + "epoch": 4.049540551338394, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10136 + }, + { + "epoch": 4.049940071913704, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10137 + }, + { + "epoch": 4.0503395924890135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10138 + }, + { + "epoch": 4.050739113064322, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10139 + }, + { + "epoch": 4.051138633639632, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 10140 + }, + { + "epoch": 4.051538154214942, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10141 + }, + { + "epoch": 4.051937674790252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10142 + }, + { + "epoch": 4.052337195365562, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10143 + }, + { + "epoch": 4.052736715940871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10144 + }, + { + "epoch": 4.05313623651618, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10145 + }, + { + "epoch": 4.05353575709149, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10146 + }, + { + "epoch": 4.0539352776668, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10147 + }, + { + "epoch": 4.05433479824211, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10148 + }, + { + "epoch": 4.054734318817419, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10149 + }, + { + "epoch": 4.055133839392728, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10150 + }, + { + "epoch": 4.055533359968038, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10151 + }, + { + "epoch": 4.055932880543348, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10152 + }, + { + "epoch": 4.056332401118658, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10153 + }, + { + "epoch": 4.0567319216939675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10154 + }, + { + "epoch": 4.057131442269277, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10155 + }, + { + "epoch": 4.057530962844586, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10156 + }, + { + "epoch": 4.057930483419896, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10157 + }, + { + "epoch": 4.058330003995206, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10158 + }, + { + "epoch": 4.0587295245705155, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10159 + }, + { + "epoch": 4.059129045145825, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10160 + }, + { + "epoch": 4.059528565721135, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10161 + }, + { + "epoch": 4.059928086296444, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10162 + }, + { + "epoch": 4.060327606871754, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10163 + }, + { + "epoch": 4.0607271274470635, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10164 + }, + { + "epoch": 4.061126648022373, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10165 + }, + { + "epoch": 4.061526168597683, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10166 + }, + { + "epoch": 4.061925689172992, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10167 + }, + { + "epoch": 4.062325209748302, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10168 + }, + { + "epoch": 4.062724730323612, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10169 + }, + { + "epoch": 4.063124250898921, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10170 + }, + { + "epoch": 4.063523771474231, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10171 + }, + { + "epoch": 4.063923292049541, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10172 + }, + { + "epoch": 4.06432281262485, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10173 + }, + { + "epoch": 4.06472233320016, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10174 + }, + { + "epoch": 4.065121853775469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10175 + }, + { + "epoch": 4.065521374350779, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10176 + }, + { + "epoch": 4.065920894926089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10177 + }, + { + "epoch": 4.066320415501398, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10178 + }, + { + "epoch": 4.066719936076708, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10179 + }, + { + "epoch": 4.0671194566520175, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10180 + }, + { + "epoch": 4.067518977227327, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10181 + }, + { + "epoch": 4.067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10182 + }, + { + "epoch": 4.068318018377947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10183 + }, + { + "epoch": 4.068717538953256, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10184 + }, + { + "epoch": 4.0691170595285655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10185 + }, + { + "epoch": 4.069516580103875, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10186 + }, + { + "epoch": 4.069916100679185, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10187 + }, + { + "epoch": 4.070315621254495, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10188 + }, + { + "epoch": 4.070715141829805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10189 + }, + { + "epoch": 4.0711146624051135, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10190 + }, + { + "epoch": 4.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10191 + }, + { + "epoch": 4.071913703555733, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10192 + }, + { + "epoch": 4.072313224131043, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10193 + }, + { + "epoch": 4.072712744706353, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10194 + }, + { + "epoch": 4.073112265281662, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10195 + }, + { + "epoch": 4.073511785856971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10196 + }, + { + "epoch": 4.073911306432281, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10197 + }, + { + "epoch": 4.074310827007591, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10198 + }, + { + "epoch": 4.074710347582901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10199 + }, + { + "epoch": 4.0751098681582105, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10200 + }, + { + "epoch": 4.075509388733519, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10201 + }, + { + "epoch": 4.075908909308829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 10202 + }, + { + "epoch": 4.076308429884139, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10203 + }, + { + "epoch": 4.076707950459449, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 10204 + }, + { + "epoch": 4.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10205 + }, + { + "epoch": 4.077506991610068, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10206 + }, + { + "epoch": 4.077906512185377, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10207 + }, + { + "epoch": 4.078306032760687, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10208 + }, + { + "epoch": 4.078705553335997, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10209 + }, + { + "epoch": 4.079105073911307, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10210 + }, + { + "epoch": 4.079504594486616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10211 + }, + { + "epoch": 4.079904115061925, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10212 + }, + { + "epoch": 4.080303635637235, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10213 + }, + { + "epoch": 4.080703156212545, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10214 + }, + { + "epoch": 4.081102676787855, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10215 + }, + { + "epoch": 4.081502197363164, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10216 + }, + { + "epoch": 4.081901717938474, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10217 + }, + { + "epoch": 4.082301238513783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10218 + }, + { + "epoch": 4.082700759089093, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10219 + }, + { + "epoch": 4.083100279664403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10220 + }, + { + "epoch": 4.0834998002397125, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10221 + }, + { + "epoch": 4.083899320815022, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10222 + }, + { + "epoch": 4.084298841390331, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10223 + }, + { + "epoch": 4.084698361965641, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10224 + }, + { + "epoch": 4.085097882540951, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 10225 + }, + { + "epoch": 4.0854974031162605, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10226 + }, + { + "epoch": 4.08589692369157, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10227 + }, + { + "epoch": 4.08629644426688, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10228 + }, + { + "epoch": 4.086695964842189, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10229 + }, + { + "epoch": 4.087095485417499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10230 + }, + { + "epoch": 4.087495005992809, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10231 + }, + { + "epoch": 4.087894526568118, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10232 + }, + { + "epoch": 4.088294047143428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10233 + }, + { + "epoch": 4.088693567718738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10234 + }, + { + "epoch": 4.089093088294047, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10235 + }, + { + "epoch": 4.089492608869357, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10236 + }, + { + "epoch": 4.089892129444666, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10237 + }, + { + "epoch": 4.090291650019976, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10238 + }, + { + "epoch": 4.090691170595286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10239 + }, + { + "epoch": 4.091090691170595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10240 + }, + { + "epoch": 4.091490211745905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10241 + }, + { + "epoch": 4.0918897323212144, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10242 + }, + { + "epoch": 4.092289252896524, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10243 + }, + { + "epoch": 4.092688773471834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10244 + }, + { + "epoch": 4.093088294047144, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10245 + }, + { + "epoch": 4.093487814622453, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10246 + }, + { + "epoch": 4.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10247 + }, + { + "epoch": 4.094286855773072, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10248 + }, + { + "epoch": 4.094686376348382, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 10249 + }, + { + "epoch": 4.095085896923692, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10250 + }, + { + "epoch": 4.095485417499002, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10251 + }, + { + "epoch": 4.0958849380743105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10252 + }, + { + "epoch": 4.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10253 + }, + { + "epoch": 4.09668397922493, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10254 + }, + { + "epoch": 4.09708349980024, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10255 + }, + { + "epoch": 4.09748302037555, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10256 + }, + { + "epoch": 4.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10257 + }, + { + "epoch": 4.098282061526168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10258 + }, + { + "epoch": 4.098681582101478, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10259 + }, + { + "epoch": 4.099081102676788, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10260 + }, + { + "epoch": 4.099480623252098, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10261 + }, + { + "epoch": 4.0998801438274075, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10262 + }, + { + "epoch": 4.100279664402716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10263 + }, + { + "epoch": 4.100679184978026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10264 + }, + { + "epoch": 4.101078705553336, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10265 + }, + { + "epoch": 4.101478226128646, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10266 + }, + { + "epoch": 4.1018777467039556, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 10267 + }, + { + "epoch": 4.1022772672792644, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10268 + }, + { + "epoch": 4.102676787854574, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10269 + }, + { + "epoch": 4.103076308429884, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10270 + }, + { + "epoch": 4.103475829005194, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10271 + }, + { + "epoch": 4.103875349580504, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10272 + }, + { + "epoch": 4.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 10273 + }, + { + "epoch": 4.104674390731122, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10274 + }, + { + "epoch": 4.105073911306432, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10275 + }, + { + "epoch": 4.105473431881742, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10276 + }, + { + "epoch": 4.105872952457052, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10277 + }, + { + "epoch": 4.106272473032361, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 10278 + }, + { + "epoch": 4.106671993607671, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10279 + }, + { + "epoch": 4.10707151418298, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10280 + }, + { + "epoch": 4.10747103475829, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10281 + }, + { + "epoch": 4.1078705553336, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10282 + }, + { + "epoch": 4.1082700759089095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10283 + }, + { + "epoch": 4.108669596484219, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10284 + }, + { + "epoch": 4.109069117059528, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10285 + }, + { + "epoch": 4.109468637634838, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10286 + }, + { + "epoch": 4.109868158210148, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10287 + }, + { + "epoch": 4.1102676787854575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10288 + }, + { + "epoch": 4.110667199360767, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10289 + }, + { + "epoch": 4.111066719936077, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10290 + }, + { + "epoch": 4.111466240511386, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10291 + }, + { + "epoch": 4.111865761086696, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10292 + }, + { + "epoch": 4.112265281662006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10293 + }, + { + "epoch": 4.112664802237315, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10294 + }, + { + "epoch": 4.113064322812625, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 10295 + }, + { + "epoch": 4.113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10296 + }, + { + "epoch": 4.113863363963244, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10297 + }, + { + "epoch": 4.114262884538554, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10298 + }, + { + "epoch": 4.114662405113863, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10299 + }, + { + "epoch": 4.115061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10300 + }, + { + "epoch": 4.115461446264483, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10301 + }, + { + "epoch": 4.115860966839792, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10302 + }, + { + "epoch": 4.116260487415102, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10303 + }, + { + "epoch": 4.116660007990411, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10304 + }, + { + "epoch": 4.117059528565721, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10305 + }, + { + "epoch": 4.117459049141031, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10306 + }, + { + "epoch": 4.117858569716341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10307 + }, + { + "epoch": 4.11825809029165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10308 + }, + { + "epoch": 4.1186576108669595, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10309 + }, + { + "epoch": 4.119057131442269, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10310 + }, + { + "epoch": 4.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 10311 + }, + { + "epoch": 4.119856172592889, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10312 + }, + { + "epoch": 4.120255693168198, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10313 + }, + { + "epoch": 4.1206552137435075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 10314 + }, + { + "epoch": 4.121054734318817, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10315 + }, + { + "epoch": 4.121454254894127, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10316 + }, + { + "epoch": 4.121853775469437, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10317 + }, + { + "epoch": 4.122253296044747, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10318 + }, + { + "epoch": 4.122652816620056, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10319 + }, + { + "epoch": 4.123052337195365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10320 + }, + { + "epoch": 4.123451857770675, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10321 + }, + { + "epoch": 4.123851378345985, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10322 + }, + { + "epoch": 4.124250898921295, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10323 + }, + { + "epoch": 4.1246504194966045, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10324 + }, + { + "epoch": 4.125049940071913, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10325 + }, + { + "epoch": 4.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10326 + }, + { + "epoch": 4.125848981222533, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10327 + }, + { + "epoch": 4.126248501797843, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 10328 + }, + { + "epoch": 4.1266480223731525, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10329 + }, + { + "epoch": 4.127047542948461, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10330 + }, + { + "epoch": 4.127447063523771, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10331 + }, + { + "epoch": 4.127846584099081, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10332 + }, + { + "epoch": 4.128246104674391, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10333 + }, + { + "epoch": 4.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10334 + }, + { + "epoch": 4.12904514582501, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10335 + }, + { + "epoch": 4.129444666400319, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10336 + }, + { + "epoch": 4.129844186975629, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 10337 + }, + { + "epoch": 4.130243707550939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10338 + }, + { + "epoch": 4.130643228126249, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10339 + }, + { + "epoch": 4.131042748701558, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10340 + }, + { + "epoch": 4.131442269276867, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10341 + }, + { + "epoch": 4.131841789852177, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10342 + }, + { + "epoch": 4.132241310427487, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10343 + }, + { + "epoch": 4.132640831002797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10344 + }, + { + "epoch": 4.1330403515781065, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10345 + }, + { + "epoch": 4.133439872153416, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10346 + }, + { + "epoch": 4.133839392728725, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10347 + }, + { + "epoch": 4.134238913304035, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10348 + }, + { + "epoch": 4.134638433879345, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10349 + }, + { + "epoch": 4.1350379544546545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10350 + }, + { + "epoch": 4.135437475029964, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 10351 + }, + { + "epoch": 4.135836995605274, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10352 + }, + { + "epoch": 4.136236516180583, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10353 + }, + { + "epoch": 4.136636036755893, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10354 + }, + { + "epoch": 4.1370355573312025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10355 + }, + { + "epoch": 4.137435077906512, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10356 + }, + { + "epoch": 4.137834598481822, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10357 + }, + { + "epoch": 4.138234119057131, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10358 + }, + { + "epoch": 4.138633639632441, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10359 + }, + { + "epoch": 4.139033160207751, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10360 + }, + { + "epoch": 4.13943268078306, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10361 + }, + { + "epoch": 4.13983220135837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10362 + }, + { + "epoch": 4.14023172193368, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10363 + }, + { + "epoch": 4.140631242508989, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10364 + }, + { + "epoch": 4.141030763084299, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10365 + }, + { + "epoch": 4.141430283659608, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10366 + }, + { + "epoch": 4.141829804234918, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10367 + }, + { + "epoch": 4.142229324810228, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10368 + }, + { + "epoch": 4.142628845385538, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10369 + }, + { + "epoch": 4.143028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10370 + }, + { + "epoch": 4.1434278865361565, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10371 + }, + { + "epoch": 4.143827407111466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10372 + }, + { + "epoch": 4.144226927686776, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10373 + }, + { + "epoch": 4.144626448262086, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10374 + }, + { + "epoch": 4.145025968837395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10375 + }, + { + "epoch": 4.1454254894127045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10376 + }, + { + "epoch": 4.145825009988014, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10377 + }, + { + "epoch": 4.146224530563324, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10378 + }, + { + "epoch": 4.146624051138634, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10379 + }, + { + "epoch": 4.147023571713944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10380 + }, + { + "epoch": 4.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10381 + }, + { + "epoch": 4.147822612864562, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10382 + }, + { + "epoch": 4.148222133439872, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10383 + }, + { + "epoch": 4.148621654015182, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10384 + }, + { + "epoch": 4.149021174590492, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10385 + }, + { + "epoch": 4.1494206951658015, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 10386 + }, + { + "epoch": 4.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10387 + }, + { + "epoch": 4.15021973631642, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10388 + }, + { + "epoch": 4.15061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10389 + }, + { + "epoch": 4.15101877746704, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10390 + }, + { + "epoch": 4.1514182980423495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10391 + }, + { + "epoch": 4.151817818617658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10392 + }, + { + "epoch": 4.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10393 + }, + { + "epoch": 4.152616859768278, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 10394 + }, + { + "epoch": 4.153016380343588, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10395 + }, + { + "epoch": 4.153415900918898, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 10396 + }, + { + "epoch": 4.153815421494207, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10397 + }, + { + "epoch": 4.154214942069516, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10398 + }, + { + "epoch": 4.154614462644826, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10399 + }, + { + "epoch": 4.155013983220136, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10400 + }, + { + "epoch": 4.155413503795446, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10401 + }, + { + "epoch": 4.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10402 + }, + { + "epoch": 4.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10403 + }, + { + "epoch": 4.156612065521374, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10404 + }, + { + "epoch": 4.157011586096684, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10405 + }, + { + "epoch": 4.157411106671994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10406 + }, + { + "epoch": 4.157810627247303, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10407 + }, + { + "epoch": 4.158210147822613, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10408 + }, + { + "epoch": 4.158609668397922, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10409 + }, + { + "epoch": 4.159009188973232, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10410 + }, + { + "epoch": 4.159408709548542, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10411 + }, + { + "epoch": 4.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10412 + }, + { + "epoch": 4.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10413 + }, + { + "epoch": 4.160607271274471, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10414 + }, + { + "epoch": 4.16100679184978, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10415 + }, + { + "epoch": 4.16140631242509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10416 + }, + { + "epoch": 4.1618058330003995, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10417 + }, + { + "epoch": 4.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10418 + }, + { + "epoch": 4.162604874151019, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10419 + }, + { + "epoch": 4.163004394726328, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10420 + }, + { + "epoch": 4.163403915301638, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10421 + }, + { + "epoch": 4.163803435876948, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10422 + }, + { + "epoch": 4.164202956452257, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10423 + }, + { + "epoch": 4.164602477027567, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10424 + }, + { + "epoch": 4.165001997602877, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10425 + }, + { + "epoch": 4.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10426 + }, + { + "epoch": 4.165801038753496, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10427 + }, + { + "epoch": 4.166200559328805, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10428 + }, + { + "epoch": 4.166600079904115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10429 + }, + { + "epoch": 4.166999600479425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10430 + }, + { + "epoch": 4.167399121054734, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10431 + }, + { + "epoch": 4.167798641630044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10432 + }, + { + "epoch": 4.168198162205353, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 10433 + }, + { + "epoch": 4.168597682780663, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10434 + }, + { + "epoch": 4.168997203355973, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10435 + }, + { + "epoch": 4.169396723931283, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10436 + }, + { + "epoch": 4.169796244506592, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10437 + }, + { + "epoch": 4.1701957650819015, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10438 + }, + { + "epoch": 4.170595285657211, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10439 + }, + { + "epoch": 4.170994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10440 + }, + { + "epoch": 4.171394326807831, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10441 + }, + { + "epoch": 4.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10442 + }, + { + "epoch": 4.1721933679584495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10443 + }, + { + "epoch": 4.172592888533759, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10444 + }, + { + "epoch": 4.172992409109069, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10445 + }, + { + "epoch": 4.173391929684379, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 10446 + }, + { + "epoch": 4.173791450259689, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10447 + }, + { + "epoch": 4.174190970834998, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10448 + }, + { + "epoch": 4.174590491410307, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10449 + }, + { + "epoch": 4.174990011985617, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10450 + }, + { + "epoch": 4.175389532560927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10451 + }, + { + "epoch": 4.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10452 + }, + { + "epoch": 4.1761885737115465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10453 + }, + { + "epoch": 4.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 10454 + }, + { + "epoch": 4.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10455 + }, + { + "epoch": 4.177387135437475, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10456 + }, + { + "epoch": 4.177786656012785, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10457 + }, + { + "epoch": 4.1781861765880945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10458 + }, + { + "epoch": 4.178585697163404, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10459 + }, + { + "epoch": 4.178985217738713, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10460 + }, + { + "epoch": 4.179384738314023, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10461 + }, + { + "epoch": 4.179784258889333, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10462 + }, + { + "epoch": 4.180183779464643, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10463 + }, + { + "epoch": 4.180583300039952, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10464 + }, + { + "epoch": 4.180982820615261, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10465 + }, + { + "epoch": 4.181382341190571, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10466 + }, + { + "epoch": 4.181781861765881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10467 + }, + { + "epoch": 4.182181382341191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10468 + }, + { + "epoch": 4.1825809029165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10469 + }, + { + "epoch": 4.18298042349181, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10470 + }, + { + "epoch": 4.183379944067119, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10471 + }, + { + "epoch": 4.183779464642429, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10472 + }, + { + "epoch": 4.184178985217739, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10473 + }, + { + "epoch": 4.1845785057930485, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10474 + }, + { + "epoch": 4.184978026368358, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10475 + }, + { + "epoch": 4.185377546943668, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10476 + }, + { + "epoch": 4.185777067518977, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10477 + }, + { + "epoch": 4.186176588094287, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 10478 + }, + { + "epoch": 4.1865761086695965, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10479 + }, + { + "epoch": 4.186975629244906, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10480 + }, + { + "epoch": 4.187375149820216, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10481 + }, + { + "epoch": 4.187774670395525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10482 + }, + { + "epoch": 4.188174190970835, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10483 + }, + { + "epoch": 4.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10484 + }, + { + "epoch": 4.188973232121454, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10485 + }, + { + "epoch": 4.189372752696764, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10486 + }, + { + "epoch": 4.189772273272074, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10487 + }, + { + "epoch": 4.190171793847383, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10488 + }, + { + "epoch": 4.190571314422693, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10489 + }, + { + "epoch": 4.190970834998002, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10490 + }, + { + "epoch": 4.191370355573312, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10491 + }, + { + "epoch": 4.191769876148622, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 10492 + }, + { + "epoch": 4.192169396723932, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10493 + }, + { + "epoch": 4.192568917299241, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10494 + }, + { + "epoch": 4.19296843787455, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10495 + }, + { + "epoch": 4.19336795844986, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10496 + }, + { + "epoch": 4.19376747902517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 10497 + }, + { + "epoch": 4.19416699960048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10498 + }, + { + "epoch": 4.194566520175789, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10499 + }, + { + "epoch": 4.1949660407510985, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10500 + }, + { + "epoch": 4.195365561326408, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10501 + }, + { + "epoch": 4.195765081901718, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10502 + }, + { + "epoch": 4.196164602477028, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10503 + }, + { + "epoch": 4.196564123052338, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10504 + }, + { + "epoch": 4.1969636436276465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10505 + }, + { + "epoch": 4.197363164202956, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10506 + }, + { + "epoch": 4.197762684778266, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10507 + }, + { + "epoch": 4.198162205353576, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10508 + }, + { + "epoch": 4.198561725928886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10509 + }, + { + "epoch": 4.1989612465041946, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10510 + }, + { + "epoch": 4.199360767079504, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10511 + }, + { + "epoch": 4.199760287654814, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10512 + }, + { + "epoch": 4.200159808230124, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10513 + }, + { + "epoch": 4.200559328805434, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10514 + }, + { + "epoch": 4.2009588493807435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10515 + }, + { + "epoch": 4.201358369956052, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10516 + }, + { + "epoch": 4.201757890531362, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10517 + }, + { + "epoch": 4.202157411106672, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 10518 + }, + { + "epoch": 4.202556931681982, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10519 + }, + { + "epoch": 4.2029564522572915, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10520 + }, + { + "epoch": 4.2033559728326, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10521 + }, + { + "epoch": 4.20375549340791, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10522 + }, + { + "epoch": 4.20415501398322, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10523 + }, + { + "epoch": 4.20455453455853, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10524 + }, + { + "epoch": 4.20495405513384, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10525 + }, + { + "epoch": 4.205353575709149, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10526 + }, + { + "epoch": 4.205753096284458, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10527 + }, + { + "epoch": 4.206152616859768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10528 + }, + { + "epoch": 4.206552137435078, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10529 + }, + { + "epoch": 4.206951658010388, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10530 + }, + { + "epoch": 4.207351178585697, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10531 + }, + { + "epoch": 4.207750699161007, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 10532 + }, + { + "epoch": 4.208150219736316, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10533 + }, + { + "epoch": 4.208549740311626, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10534 + }, + { + "epoch": 4.208949260886936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10535 + }, + { + "epoch": 4.2093487814622454, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10536 + }, + { + "epoch": 4.209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 10537 + }, + { + "epoch": 4.210147822612864, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10538 + }, + { + "epoch": 4.210547343188174, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10539 + }, + { + "epoch": 4.210946863763484, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10540 + }, + { + "epoch": 4.2113463843387935, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10541 + }, + { + "epoch": 4.211745904914103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10542 + }, + { + "epoch": 4.212145425489413, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10543 + }, + { + "epoch": 4.212544946064722, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10544 + }, + { + "epoch": 4.212944466640032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10545 + }, + { + "epoch": 4.2133439872153415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10546 + }, + { + "epoch": 4.213743507790651, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10547 + }, + { + "epoch": 4.214143028365961, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10548 + }, + { + "epoch": 4.214542548941271, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10549 + }, + { + "epoch": 4.21494206951658, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10550 + }, + { + "epoch": 4.21534159009189, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10551 + }, + { + "epoch": 4.215741110667199, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10552 + }, + { + "epoch": 4.216140631242509, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10553 + }, + { + "epoch": 4.216540151817819, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10554 + }, + { + "epoch": 4.216939672393128, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10555 + }, + { + "epoch": 4.217339192968438, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10556 + }, + { + "epoch": 4.217738713543747, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10557 + }, + { + "epoch": 4.218138234119057, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10558 + }, + { + "epoch": 4.218537754694367, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10559 + }, + { + "epoch": 4.218937275269677, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10560 + }, + { + "epoch": 4.219336795844986, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10561 + }, + { + "epoch": 4.2197363164202955, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10562 + }, + { + "epoch": 4.220135836995605, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10563 + }, + { + "epoch": 4.220535357570915, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10564 + }, + { + "epoch": 4.220934878146225, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10565 + }, + { + "epoch": 4.221334398721535, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10566 + }, + { + "epoch": 4.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10567 + }, + { + "epoch": 4.222133439872153, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10568 + }, + { + "epoch": 4.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10569 + }, + { + "epoch": 4.222932481022773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10570 + }, + { + "epoch": 4.223332001598083, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10571 + }, + { + "epoch": 4.2237315221733915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10572 + }, + { + "epoch": 4.224131042748701, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10573 + }, + { + "epoch": 4.224530563324011, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10574 + }, + { + "epoch": 4.224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10575 + }, + { + "epoch": 4.225329604474631, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10576 + }, + { + "epoch": 4.2257291250499405, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10577 + }, + { + "epoch": 4.226128645625249, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 10578 + }, + { + "epoch": 4.226528166200559, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10579 + }, + { + "epoch": 4.226927686775869, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10580 + }, + { + "epoch": 4.227327207351179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10581 + }, + { + "epoch": 4.2277267279264885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10582 + }, + { + "epoch": 4.228126248501797, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10583 + }, + { + "epoch": 4.228525769077107, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10584 + }, + { + "epoch": 4.228925289652417, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10585 + }, + { + "epoch": 4.229324810227727, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10586 + }, + { + "epoch": 4.229724330803037, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 10587 + }, + { + "epoch": 4.230123851378346, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10588 + }, + { + "epoch": 4.230523371953655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10589 + }, + { + "epoch": 4.230922892528965, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10590 + }, + { + "epoch": 4.231322413104275, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10591 + }, + { + "epoch": 4.231721933679585, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 10592 + }, + { + "epoch": 4.232121454254894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10593 + }, + { + "epoch": 4.232520974830204, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10594 + }, + { + "epoch": 4.232920495405513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10595 + }, + { + "epoch": 4.233320015980823, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10596 + }, + { + "epoch": 4.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10597 + }, + { + "epoch": 4.234119057131442, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10598 + }, + { + "epoch": 4.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10599 + }, + { + "epoch": 4.234918098282061, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10600 + }, + { + "epoch": 4.235317618857371, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10601 + }, + { + "epoch": 4.235717139432681, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10602 + }, + { + "epoch": 4.2361166600079905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10603 + }, + { + "epoch": 4.2365161805833, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10604 + }, + { + "epoch": 4.23691570115861, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 10605 + }, + { + "epoch": 4.237315221733919, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10606 + }, + { + "epoch": 4.237714742309229, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10607 + }, + { + "epoch": 4.2381142628845385, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10608 + }, + { + "epoch": 4.238513783459848, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10609 + }, + { + "epoch": 4.238913304035158, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10610 + }, + { + "epoch": 4.239312824610467, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10611 + }, + { + "epoch": 4.239712345185777, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 10612 + }, + { + "epoch": 4.240111865761087, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10613 + }, + { + "epoch": 4.240511386336396, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10614 + }, + { + "epoch": 4.240910906911706, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 10615 + }, + { + "epoch": 4.241310427487016, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10616 + }, + { + "epoch": 4.241709948062325, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10617 + }, + { + "epoch": 4.242109468637635, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10618 + }, + { + "epoch": 4.242508989212944, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10619 + }, + { + "epoch": 4.242908509788254, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10620 + }, + { + "epoch": 4.243308030363564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10621 + }, + { + "epoch": 4.243707550938874, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 10622 + }, + { + "epoch": 4.244107071514183, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10623 + }, + { + "epoch": 4.244506592089492, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10624 + }, + { + "epoch": 4.244906112664802, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10625 + }, + { + "epoch": 4.245305633240112, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10626 + }, + { + "epoch": 4.245705153815422, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10627 + }, + { + "epoch": 4.246104674390731, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 10628 + }, + { + "epoch": 4.2465041949660405, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10629 + }, + { + "epoch": 4.24690371554135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10630 + }, + { + "epoch": 4.24730323611666, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10631 + }, + { + "epoch": 4.24770275669197, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10632 + }, + { + "epoch": 4.24810227726728, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10633 + }, + { + "epoch": 4.2485017978425885, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10634 + }, + { + "epoch": 4.248901318417898, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10635 + }, + { + "epoch": 4.249300838993208, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10636 + }, + { + "epoch": 4.249700359568518, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10637 + }, + { + "epoch": 4.250099880143828, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10638 + }, + { + "epoch": 4.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10639 + }, + { + "epoch": 4.250898921294446, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10640 + }, + { + "epoch": 4.251298441869756, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10641 + }, + { + "epoch": 4.251697962445066, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10642 + }, + { + "epoch": 4.252097483020376, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10643 + }, + { + "epoch": 4.2524970035956855, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10644 + }, + { + "epoch": 4.252896524170994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10645 + }, + { + "epoch": 4.253296044746304, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10646 + }, + { + "epoch": 4.253695565321614, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10647 + }, + { + "epoch": 4.254095085896924, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10648 + }, + { + "epoch": 4.2544946064722335, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10649 + }, + { + "epoch": 4.254894127047543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 10650 + }, + { + "epoch": 4.255293647622852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10651 + }, + { + "epoch": 4.255693168198162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10652 + }, + { + "epoch": 4.256092688773472, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10653 + }, + { + "epoch": 4.256492209348782, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10654 + }, + { + "epoch": 4.256891729924091, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10655 + }, + { + "epoch": 4.257291250499401, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10656 + }, + { + "epoch": 4.25769077107471, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10657 + }, + { + "epoch": 4.25809029165002, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10658 + }, + { + "epoch": 4.25848981222533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10659 + }, + { + "epoch": 4.258889332800639, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10660 + }, + { + "epoch": 4.259288853375949, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10661 + }, + { + "epoch": 4.259688373951258, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10662 + }, + { + "epoch": 4.260087894526568, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10663 + }, + { + "epoch": 4.260487415101878, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10664 + }, + { + "epoch": 4.2608869356771875, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10665 + }, + { + "epoch": 4.261286456252497, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10666 + }, + { + "epoch": 4.261685976827807, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10667 + }, + { + "epoch": 4.262085497403116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10668 + }, + { + "epoch": 4.262485017978426, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10669 + }, + { + "epoch": 4.2628845385537355, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 10670 + }, + { + "epoch": 4.263284059129045, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10671 + }, + { + "epoch": 4.263683579704355, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10672 + }, + { + "epoch": 4.264083100279665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10673 + }, + { + "epoch": 4.264482620854974, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10674 + }, + { + "epoch": 4.2648821414302835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10675 + }, + { + "epoch": 4.265281662005593, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10676 + }, + { + "epoch": 4.265681182580903, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10677 + }, + { + "epoch": 4.266080703156213, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10678 + }, + { + "epoch": 4.266480223731522, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10679 + }, + { + "epoch": 4.266879744306832, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10680 + }, + { + "epoch": 4.267279264882141, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 10681 + }, + { + "epoch": 4.267678785457451, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10682 + }, + { + "epoch": 4.268078306032761, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 10683 + }, + { + "epoch": 4.26847782660807, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10684 + }, + { + "epoch": 4.26887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 10685 + }, + { + "epoch": 4.269276867758689, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 10686 + }, + { + "epoch": 4.269676388333999, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 10687 + }, + { + "epoch": 4.270075908909309, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10688 + }, + { + "epoch": 4.270475429484619, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10689 + }, + { + "epoch": 4.270874950059928, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10690 + }, + { + "epoch": 4.2712744706352375, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10691 + }, + { + "epoch": 4.271673991210547, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 10692 + }, + { + "epoch": 4.272073511785857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10693 + }, + { + "epoch": 4.272473032361167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10694 + }, + { + "epoch": 4.272872552936477, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10695 + }, + { + "epoch": 4.2732720735117855, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10696 + }, + { + "epoch": 4.273671594087095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10697 + }, + { + "epoch": 4.274071114662405, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10698 + }, + { + "epoch": 4.274470635237715, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10699 + }, + { + "epoch": 4.274870155813025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10700 + }, + { + "epoch": 4.2752696763883336, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10701 + }, + { + "epoch": 4.275669196963643, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10702 + }, + { + "epoch": 4.276068717538953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10703 + }, + { + "epoch": 4.276468238114263, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10704 + }, + { + "epoch": 4.276867758689573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 10705 + }, + { + "epoch": 4.2772672792648825, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10706 + }, + { + "epoch": 4.277666799840191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10707 + }, + { + "epoch": 4.278066320415501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10708 + }, + { + "epoch": 4.278465840990811, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10709 + }, + { + "epoch": 4.278865361566121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10710 + }, + { + "epoch": 4.2792648821414305, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10711 + }, + { + "epoch": 4.27966440271674, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10712 + }, + { + "epoch": 4.280063923292049, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10713 + }, + { + "epoch": 4.280463443867359, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10714 + }, + { + "epoch": 4.280862964442669, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10715 + }, + { + "epoch": 4.281262485017979, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10716 + }, + { + "epoch": 4.281662005593288, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10717 + }, + { + "epoch": 4.282061526168597, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10718 + }, + { + "epoch": 4.282461046743907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10719 + }, + { + "epoch": 4.282860567319217, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10720 + }, + { + "epoch": 4.283260087894527, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10721 + }, + { + "epoch": 4.283659608469836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10722 + }, + { + "epoch": 4.284059129045146, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10723 + }, + { + "epoch": 4.284458649620455, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10724 + }, + { + "epoch": 4.284858170195765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 10725 + }, + { + "epoch": 4.285257690771075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10726 + }, + { + "epoch": 4.285657211346384, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 10727 + }, + { + "epoch": 4.286056731921694, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10728 + }, + { + "epoch": 4.286456252497004, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10729 + }, + { + "epoch": 4.286855773072313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10730 + }, + { + "epoch": 4.287255293647623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10731 + }, + { + "epoch": 4.2876548142229325, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10732 + }, + { + "epoch": 4.288054334798242, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10733 + }, + { + "epoch": 4.288453855373552, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10734 + }, + { + "epoch": 4.288853375948861, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10735 + }, + { + "epoch": 4.289252896524171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 10736 + }, + { + "epoch": 4.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10737 + }, + { + "epoch": 4.29005193767479, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10738 + }, + { + "epoch": 4.2904514582501, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10739 + }, + { + "epoch": 4.29085097882541, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10740 + }, + { + "epoch": 4.291250499400719, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10741 + }, + { + "epoch": 4.291650019976029, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10742 + }, + { + "epoch": 4.292049540551338, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10743 + }, + { + "epoch": 4.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10744 + }, + { + "epoch": 4.292848581701958, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10745 + }, + { + "epoch": 4.293248102277268, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10746 + }, + { + "epoch": 4.293647622852577, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10747 + }, + { + "epoch": 4.294047143427886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10748 + }, + { + "epoch": 4.294446664003196, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10749 + }, + { + "epoch": 4.294846184578506, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10750 + }, + { + "epoch": 4.295245705153816, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10751 + }, + { + "epoch": 4.295645225729125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 10752 + }, + { + "epoch": 4.2960447463044344, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10753 + }, + { + "epoch": 4.296444266879744, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10754 + }, + { + "epoch": 4.296843787455054, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10755 + }, + { + "epoch": 4.297243308030364, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 10756 + }, + { + "epoch": 4.297642828605674, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10757 + }, + { + "epoch": 4.2980423491809825, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10758 + }, + { + "epoch": 4.298441869756292, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10759 + }, + { + "epoch": 4.298841390331602, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10760 + }, + { + "epoch": 4.299240910906912, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10761 + }, + { + "epoch": 4.299640431482222, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 10762 + }, + { + "epoch": 4.300039952057531, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10763 + }, + { + "epoch": 4.30043947263284, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10764 + }, + { + "epoch": 4.30083899320815, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10765 + }, + { + "epoch": 4.30123851378346, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10766 + }, + { + "epoch": 4.30163803435877, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 10767 + }, + { + "epoch": 4.3020375549340795, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10768 + }, + { + "epoch": 4.302437075509388, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10769 + }, + { + "epoch": 4.302836596084698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10770 + }, + { + "epoch": 4.303236116660008, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10771 + }, + { + "epoch": 4.303635637235318, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10772 + }, + { + "epoch": 4.3040351578106275, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10773 + }, + { + "epoch": 4.304434678385936, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10774 + }, + { + "epoch": 4.304834198961246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10775 + }, + { + "epoch": 4.305233719536556, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10776 + }, + { + "epoch": 4.305633240111866, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10777 + }, + { + "epoch": 4.3060327606871756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10778 + }, + { + "epoch": 4.306432281262485, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10779 + }, + { + "epoch": 4.306831801837794, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10780 + }, + { + "epoch": 4.307231322413104, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10781 + }, + { + "epoch": 4.307630842988414, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10782 + }, + { + "epoch": 4.308030363563724, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10783 + }, + { + "epoch": 4.308429884139033, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10784 + }, + { + "epoch": 4.308829404714343, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10785 + }, + { + "epoch": 4.309228925289652, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10786 + }, + { + "epoch": 4.309628445864962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10787 + }, + { + "epoch": 4.310027966440272, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10788 + }, + { + "epoch": 4.310427487015581, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10789 + }, + { + "epoch": 4.310827007590891, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10790 + }, + { + "epoch": 4.3112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10791 + }, + { + "epoch": 4.31162604874151, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10792 + }, + { + "epoch": 4.31202556931682, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10793 + }, + { + "epoch": 4.3124250898921295, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10794 + }, + { + "epoch": 4.312824610467439, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10795 + }, + { + "epoch": 4.313224131042749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10796 + }, + { + "epoch": 4.313623651618058, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10797 + }, + { + "epoch": 4.314023172193368, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10798 + }, + { + "epoch": 4.3144226927686775, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10799 + }, + { + "epoch": 4.314822213343987, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10800 + }, + { + "epoch": 4.315221733919297, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10801 + }, + { + "epoch": 4.315621254494607, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10802 + }, + { + "epoch": 4.316020775069916, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10803 + }, + { + "epoch": 4.316420295645226, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 10804 + }, + { + "epoch": 4.316819816220535, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10805 + }, + { + "epoch": 4.317219336795845, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10806 + }, + { + "epoch": 4.317618857371155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10807 + }, + { + "epoch": 4.318018377946464, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10808 + }, + { + "epoch": 4.318417898521774, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10809 + }, + { + "epoch": 4.318817419097083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10810 + }, + { + "epoch": 4.319216939672393, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10811 + }, + { + "epoch": 4.319616460247703, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10812 + }, + { + "epoch": 4.320015980823013, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10813 + }, + { + "epoch": 4.320415501398322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10814 + }, + { + "epoch": 4.320815021973631, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10815 + }, + { + "epoch": 4.321214542548941, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10816 + }, + { + "epoch": 4.321614063124251, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10817 + }, + { + "epoch": 4.322013583699561, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10818 + }, + { + "epoch": 4.322413104274871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10819 + }, + { + "epoch": 4.3228126248501795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10820 + }, + { + "epoch": 4.323212145425489, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10821 + }, + { + "epoch": 4.323611666000799, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 10822 + }, + { + "epoch": 4.324011186576109, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 10823 + }, + { + "epoch": 4.324410707151419, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10824 + }, + { + "epoch": 4.3248102277267275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10825 + }, + { + "epoch": 4.325209748302037, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10826 + }, + { + "epoch": 4.325609268877347, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10827 + }, + { + "epoch": 4.326008789452657, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10828 + }, + { + "epoch": 4.326408310027967, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10829 + }, + { + "epoch": 4.3268078306032765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10830 + }, + { + "epoch": 4.327207351178585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10831 + }, + { + "epoch": 4.327606871753895, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10832 + }, + { + "epoch": 4.328006392329205, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10833 + }, + { + "epoch": 4.328405912904515, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10834 + }, + { + "epoch": 4.3288054334798245, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10835 + }, + { + "epoch": 4.329204954055134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10836 + }, + { + "epoch": 4.329604474630443, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10837 + }, + { + "epoch": 4.330003995205753, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10838 + }, + { + "epoch": 4.330403515781063, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10839 + }, + { + "epoch": 4.3308030363563725, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10840 + }, + { + "epoch": 4.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10841 + }, + { + "epoch": 4.331602077506991, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10842 + }, + { + "epoch": 4.332001598082301, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10843 + }, + { + "epoch": 4.332401118657611, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10844 + }, + { + "epoch": 4.332800639232921, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10845 + }, + { + "epoch": 4.33320015980823, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10846 + }, + { + "epoch": 4.33359968038354, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10847 + }, + { + "epoch": 4.333999200958849, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10848 + }, + { + "epoch": 4.334398721534159, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10849 + }, + { + "epoch": 4.334798242109469, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10850 + }, + { + "epoch": 4.335197762684778, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10851 + }, + { + "epoch": 4.335597283260088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10852 + }, + { + "epoch": 4.335996803835398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10853 + }, + { + "epoch": 4.336396324410707, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10854 + }, + { + "epoch": 4.336795844986017, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10855 + }, + { + "epoch": 4.3371953655613265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10856 + }, + { + "epoch": 4.337594886136636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10857 + }, + { + "epoch": 4.337994406711946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10858 + }, + { + "epoch": 4.338393927287255, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10859 + }, + { + "epoch": 4.338793447862565, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10860 + }, + { + "epoch": 4.3391929684378745, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10861 + }, + { + "epoch": 4.339592489013184, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10862 + }, + { + "epoch": 4.339992009588494, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 10863 + }, + { + "epoch": 4.340391530163803, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 10864 + }, + { + "epoch": 4.340791050739113, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10865 + }, + { + "epoch": 4.3411905713144225, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10866 + }, + { + "epoch": 4.341590091889732, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10867 + }, + { + "epoch": 4.341989612465042, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10868 + }, + { + "epoch": 4.342389133040352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10869 + }, + { + "epoch": 4.342788653615661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10870 + }, + { + "epoch": 4.343188174190971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10871 + }, + { + "epoch": 4.34358769476628, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 10872 + }, + { + "epoch": 4.34398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10873 + }, + { + "epoch": 4.3443867359169, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10874 + }, + { + "epoch": 4.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10875 + }, + { + "epoch": 4.345185777067519, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10876 + }, + { + "epoch": 4.345585297642828, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10877 + }, + { + "epoch": 4.345984818218138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10878 + }, + { + "epoch": 4.346384338793448, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10879 + }, + { + "epoch": 4.346783859368758, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10880 + }, + { + "epoch": 4.347183379944067, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10881 + }, + { + "epoch": 4.3475829005193765, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10882 + }, + { + "epoch": 4.347982421094686, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10883 + }, + { + "epoch": 4.348381941669996, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10884 + }, + { + "epoch": 4.348781462245306, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10885 + }, + { + "epoch": 4.349180982820616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10886 + }, + { + "epoch": 4.3495805033959245, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10887 + }, + { + "epoch": 4.349980023971234, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10888 + }, + { + "epoch": 4.350379544546544, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10889 + }, + { + "epoch": 4.350779065121854, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10890 + }, + { + "epoch": 4.351178585697164, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10891 + }, + { + "epoch": 4.351578106272473, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10892 + }, + { + "epoch": 4.351977626847782, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 10893 + }, + { + "epoch": 4.352377147423092, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10894 + }, + { + "epoch": 4.352776667998402, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10895 + }, + { + "epoch": 4.353176188573712, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10896 + }, + { + "epoch": 4.3535757091490215, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10897 + }, + { + "epoch": 4.35397522972433, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10898 + }, + { + "epoch": 4.35437475029964, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10899 + }, + { + "epoch": 4.35477427087495, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10900 + }, + { + "epoch": 4.35517379145026, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10901 + }, + { + "epoch": 4.3555733120255695, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10902 + }, + { + "epoch": 4.355972832600879, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10903 + }, + { + "epoch": 4.356372353176188, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10904 + }, + { + "epoch": 4.356771873751498, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 10905 + }, + { + "epoch": 4.357171394326808, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10906 + }, + { + "epoch": 4.357570914902118, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10907 + }, + { + "epoch": 4.357970435477427, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10908 + }, + { + "epoch": 4.358369956052737, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10909 + }, + { + "epoch": 4.358769476628046, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10910 + }, + { + "epoch": 4.359168997203356, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10911 + }, + { + "epoch": 4.359568517778666, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10912 + }, + { + "epoch": 4.359968038353975, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10913 + }, + { + "epoch": 4.360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10914 + }, + { + "epoch": 4.360767079504594, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10915 + }, + { + "epoch": 4.361166600079904, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10916 + }, + { + "epoch": 4.361566120655214, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10917 + }, + { + "epoch": 4.361965641230523, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10918 + }, + { + "epoch": 4.362365161805833, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10919 + }, + { + "epoch": 4.362764682381143, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10920 + }, + { + "epoch": 4.363164202956452, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10921 + }, + { + "epoch": 4.363563723531762, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10922 + }, + { + "epoch": 4.3639632441070715, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10923 + }, + { + "epoch": 4.364362764682381, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10924 + }, + { + "epoch": 4.364762285257691, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10925 + }, + { + "epoch": 4.365161805833001, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10926 + }, + { + "epoch": 4.36556132640831, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10927 + }, + { + "epoch": 4.3659608469836195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10928 + }, + { + "epoch": 4.366360367558929, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10929 + }, + { + "epoch": 4.366759888134239, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10930 + }, + { + "epoch": 4.367159408709549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10931 + }, + { + "epoch": 4.367558929284858, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10932 + }, + { + "epoch": 4.367958449860168, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10933 + }, + { + "epoch": 4.368357970435477, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10934 + }, + { + "epoch": 4.368757491010787, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10935 + }, + { + "epoch": 4.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10936 + }, + { + "epoch": 4.369556532161407, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10937 + }, + { + "epoch": 4.369956052736716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10938 + }, + { + "epoch": 4.370355573312025, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10939 + }, + { + "epoch": 4.370755093887335, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10940 + }, + { + "epoch": 4.371154614462645, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10941 + }, + { + "epoch": 4.371554135037955, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10942 + }, + { + "epoch": 4.3719536556132645, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10943 + }, + { + "epoch": 4.372353176188573, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10944 + }, + { + "epoch": 4.372752696763883, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 10945 + }, + { + "epoch": 4.373152217339193, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10946 + }, + { + "epoch": 4.373551737914503, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10947 + }, + { + "epoch": 4.373951258489813, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10948 + }, + { + "epoch": 4.3743507790651215, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10949 + }, + { + "epoch": 4.374750299640431, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 10950 + }, + { + "epoch": 4.375149820215741, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 10951 + }, + { + "epoch": 4.375549340791051, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10952 + }, + { + "epoch": 4.375948861366361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10953 + }, + { + "epoch": 4.3763483819416695, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 10954 + }, + { + "epoch": 4.376747902516979, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10955 + }, + { + "epoch": 4.377147423092289, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10956 + }, + { + "epoch": 4.377546943667599, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10957 + }, + { + "epoch": 4.377946464242909, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10958 + }, + { + "epoch": 4.3783459848182185, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10959 + }, + { + "epoch": 4.378745505393527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10960 + }, + { + "epoch": 4.379145025968837, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10961 + }, + { + "epoch": 4.379544546544147, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 10962 + }, + { + "epoch": 4.379944067119457, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10963 + }, + { + "epoch": 4.3803435876947665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10964 + }, + { + "epoch": 4.380743108270076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10965 + }, + { + "epoch": 4.381142628845385, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10966 + }, + { + "epoch": 4.381542149420695, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10967 + }, + { + "epoch": 4.381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10968 + }, + { + "epoch": 4.3823411905713145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10969 + }, + { + "epoch": 4.382740711146624, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10970 + }, + { + "epoch": 4.383140231721933, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10971 + }, + { + "epoch": 4.383539752297243, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10972 + }, + { + "epoch": 4.383939272872553, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10973 + }, + { + "epoch": 4.384338793447863, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10974 + }, + { + "epoch": 4.384738314023172, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10975 + }, + { + "epoch": 4.385137834598482, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10976 + }, + { + "epoch": 4.385537355173791, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10977 + }, + { + "epoch": 4.385936875749101, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10978 + }, + { + "epoch": 4.386336396324411, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10979 + }, + { + "epoch": 4.38673591689972, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10980 + }, + { + "epoch": 4.38713543747503, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10981 + }, + { + "epoch": 4.38753495805034, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10982 + }, + { + "epoch": 4.387934478625649, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10983 + }, + { + "epoch": 4.388333999200959, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10984 + }, + { + "epoch": 4.3887335197762685, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10985 + }, + { + "epoch": 4.389133040351578, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 10986 + }, + { + "epoch": 4.389532560926888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 10987 + }, + { + "epoch": 4.389932081502197, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10988 + }, + { + "epoch": 4.390331602077507, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10989 + }, + { + "epoch": 4.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 10990 + }, + { + "epoch": 4.391130643228126, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10991 + }, + { + "epoch": 4.391530163803436, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10992 + }, + { + "epoch": 4.391929684378746, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10993 + }, + { + "epoch": 4.392329204954055, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 10994 + }, + { + "epoch": 4.3927287255293646, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10995 + }, + { + "epoch": 4.393128246104674, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10996 + }, + { + "epoch": 4.393527766679984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10997 + }, + { + "epoch": 4.393927287255294, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10998 + }, + { + "epoch": 4.394326807830604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10999 + }, + { + "epoch": 4.394726328405913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11000 + }, + { + "epoch": 4.395125848981222, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11001 + }, + { + "epoch": 4.395525369556532, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11002 + }, + { + "epoch": 4.395924890131842, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 11003 + }, + { + "epoch": 4.396324410707152, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11004 + }, + { + "epoch": 4.396723931282461, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11005 + }, + { + "epoch": 4.39712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11006 + }, + { + "epoch": 4.39752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11007 + }, + { + "epoch": 4.39792249300839, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11008 + }, + { + "epoch": 4.3983220135837, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11009 + }, + { + "epoch": 4.39872153415901, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11010 + }, + { + "epoch": 4.3991210547343185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11011 + }, + { + "epoch": 4.399520575309628, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11012 + }, + { + "epoch": 4.399920095884938, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11013 + }, + { + "epoch": 4.400319616460248, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11014 + }, + { + "epoch": 4.400719137035558, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11015 + }, + { + "epoch": 4.401118657610867, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11016 + }, + { + "epoch": 4.401518178186176, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11017 + }, + { + "epoch": 4.401917698761486, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11018 + }, + { + "epoch": 4.402317219336796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11019 + }, + { + "epoch": 4.402716739912106, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11020 + }, + { + "epoch": 4.4031162604874154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11021 + }, + { + "epoch": 4.403515781062724, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11022 + }, + { + "epoch": 4.403915301638034, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11023 + }, + { + "epoch": 4.404314822213344, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11024 + }, + { + "epoch": 4.404714342788654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11025 + }, + { + "epoch": 4.4051138633639635, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11026 + }, + { + "epoch": 4.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11027 + }, + { + "epoch": 4.405912904514582, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11028 + }, + { + "epoch": 4.406312425089892, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11029 + }, + { + "epoch": 4.406711945665202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11030 + }, + { + "epoch": 4.4071114662405115, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11031 + }, + { + "epoch": 4.407510986815821, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11032 + }, + { + "epoch": 4.407910507391131, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 11033 + }, + { + "epoch": 4.40831002796644, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11034 + }, + { + "epoch": 4.40870954854175, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11035 + }, + { + "epoch": 4.40910906911706, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11036 + }, + { + "epoch": 4.409508589692369, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11037 + }, + { + "epoch": 4.409908110267679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11038 + }, + { + "epoch": 4.410307630842988, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11039 + }, + { + "epoch": 4.410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11040 + }, + { + "epoch": 4.411106671993608, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11041 + }, + { + "epoch": 4.411506192568917, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11042 + }, + { + "epoch": 4.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11043 + }, + { + "epoch": 4.412305233719536, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11044 + }, + { + "epoch": 4.412704754294846, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11045 + }, + { + "epoch": 4.413104274870156, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11046 + }, + { + "epoch": 4.4135037954454654, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11047 + }, + { + "epoch": 4.413903316020775, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11048 + }, + { + "epoch": 4.414302836596085, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11049 + }, + { + "epoch": 4.414702357171394, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11050 + }, + { + "epoch": 4.415101877746704, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11051 + }, + { + "epoch": 4.4155013983220135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11052 + }, + { + "epoch": 4.415900918897323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11053 + }, + { + "epoch": 4.416300439472633, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11054 + }, + { + "epoch": 4.416699960047943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11055 + }, + { + "epoch": 4.417099480623252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11056 + }, + { + "epoch": 4.4174990011985615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11057 + }, + { + "epoch": 4.417898521773871, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11058 + }, + { + "epoch": 4.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11059 + }, + { + "epoch": 4.418697562924491, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11060 + }, + { + "epoch": 4.4190970834998, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11061 + }, + { + "epoch": 4.41949660407511, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11062 + }, + { + "epoch": 4.419896124650419, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11063 + }, + { + "epoch": 4.420295645225729, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11064 + }, + { + "epoch": 4.420695165801039, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11065 + }, + { + "epoch": 4.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11066 + }, + { + "epoch": 4.421494206951658, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11067 + }, + { + "epoch": 4.421893727526967, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11068 + }, + { + "epoch": 4.422293248102277, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11069 + }, + { + "epoch": 4.422692768677587, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11070 + }, + { + "epoch": 4.423092289252897, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11071 + }, + { + "epoch": 4.4234918098282066, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11072 + }, + { + "epoch": 4.4238913304035155, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11073 + }, + { + "epoch": 4.424290850978825, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11074 + }, + { + "epoch": 4.424690371554135, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11075 + }, + { + "epoch": 4.425089892129445, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11076 + }, + { + "epoch": 4.425489412704755, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11077 + }, + { + "epoch": 4.4258889332800635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11078 + }, + { + "epoch": 4.426288453855373, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11079 + }, + { + "epoch": 4.426687974430683, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11080 + }, + { + "epoch": 4.427087495005993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11081 + }, + { + "epoch": 4.427487015581303, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11082 + }, + { + "epoch": 4.427886536156612, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11083 + }, + { + "epoch": 4.428286056731921, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11084 + }, + { + "epoch": 4.428685577307231, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11085 + }, + { + "epoch": 4.429085097882541, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11086 + }, + { + "epoch": 4.429484618457851, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11087 + }, + { + "epoch": 4.4298841390331605, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 11088 + }, + { + "epoch": 4.43028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11089 + }, + { + "epoch": 4.430683180183779, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11090 + }, + { + "epoch": 4.431082700759089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11091 + }, + { + "epoch": 4.431482221334399, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 11092 + }, + { + "epoch": 4.4318817419097085, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11093 + }, + { + "epoch": 4.432281262485018, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11094 + }, + { + "epoch": 4.432680783060327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11095 + }, + { + "epoch": 4.433080303635637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11096 + }, + { + "epoch": 4.433479824210947, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11097 + }, + { + "epoch": 4.433879344786257, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 11098 + }, + { + "epoch": 4.434278865361566, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11099 + }, + { + "epoch": 4.434678385936876, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11100 + }, + { + "epoch": 4.435077906512185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11101 + }, + { + "epoch": 4.435477427087495, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11102 + }, + { + "epoch": 4.435876947662805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11103 + }, + { + "epoch": 4.436276468238114, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11104 + }, + { + "epoch": 4.436675988813424, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11105 + }, + { + "epoch": 4.437075509388734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11106 + }, + { + "epoch": 4.437475029964043, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 11107 + }, + { + "epoch": 4.437874550539353, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 11108 + }, + { + "epoch": 4.438274071114662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11109 + }, + { + "epoch": 4.438673591689972, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11110 + }, + { + "epoch": 4.439073112265282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11111 + }, + { + "epoch": 4.439472632840591, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11112 + }, + { + "epoch": 4.439872153415901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11113 + }, + { + "epoch": 4.4402716739912105, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11114 + }, + { + "epoch": 4.44067119456652, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11115 + }, + { + "epoch": 4.44107071514183, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11116 + }, + { + "epoch": 4.44147023571714, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11117 + }, + { + "epoch": 4.441869756292449, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11118 + }, + { + "epoch": 4.4422692768677585, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 11119 + }, + { + "epoch": 4.442668797443068, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11120 + }, + { + "epoch": 4.443068318018378, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11121 + }, + { + "epoch": 4.443467838593688, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11122 + }, + { + "epoch": 4.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11123 + }, + { + "epoch": 4.444266879744307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11124 + }, + { + "epoch": 4.444666400319616, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11125 + }, + { + "epoch": 4.445065920894926, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11126 + }, + { + "epoch": 4.445465441470236, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11127 + }, + { + "epoch": 4.445864962045546, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11128 + }, + { + "epoch": 4.446264482620855, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11129 + }, + { + "epoch": 4.446664003196164, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11130 + }, + { + "epoch": 4.447063523771474, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11131 + }, + { + "epoch": 4.447463044346784, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11132 + }, + { + "epoch": 4.447862564922094, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11133 + }, + { + "epoch": 4.448262085497403, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11134 + }, + { + "epoch": 4.448661606072712, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11135 + }, + { + "epoch": 4.449061126648022, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11136 + }, + { + "epoch": 4.449460647223332, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 11137 + }, + { + "epoch": 4.449860167798642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11138 + }, + { + "epoch": 4.450259688373952, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11139 + }, + { + "epoch": 4.4506592089492605, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 11140 + }, + { + "epoch": 4.45105872952457, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11141 + }, + { + "epoch": 4.45145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11142 + }, + { + "epoch": 4.45185777067519, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11143 + }, + { + "epoch": 4.4522572912505, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11144 + }, + { + "epoch": 4.452656811825809, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11145 + }, + { + "epoch": 4.453056332401118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11146 + }, + { + "epoch": 4.453455852976428, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11147 + }, + { + "epoch": 4.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11148 + }, + { + "epoch": 4.454254894127048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11149 + }, + { + "epoch": 4.4546544147023575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11150 + }, + { + "epoch": 4.455053935277666, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11151 + }, + { + "epoch": 4.455453455852976, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11152 + }, + { + "epoch": 4.455852976428286, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11153 + }, + { + "epoch": 4.456252497003596, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11154 + }, + { + "epoch": 4.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11155 + }, + { + "epoch": 4.457051538154215, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11156 + }, + { + "epoch": 4.457451058729524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11157 + }, + { + "epoch": 4.457850579304834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11158 + }, + { + "epoch": 4.458250099880144, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11159 + }, + { + "epoch": 4.4586496204554535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11160 + }, + { + "epoch": 4.459049141030763, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11161 + }, + { + "epoch": 4.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11162 + }, + { + "epoch": 4.459848182181382, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11163 + }, + { + "epoch": 4.460247702756692, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11164 + }, + { + "epoch": 4.460647223332002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11165 + }, + { + "epoch": 4.461046743907311, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11166 + }, + { + "epoch": 4.461446264482621, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11167 + }, + { + "epoch": 4.46184578505793, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11168 + }, + { + "epoch": 4.46224530563324, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11169 + }, + { + "epoch": 4.46264482620855, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11170 + }, + { + "epoch": 4.463044346783859, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11171 + }, + { + "epoch": 4.463443867359169, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 11172 + }, + { + "epoch": 4.463843387934479, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11173 + }, + { + "epoch": 4.464242908509788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11174 + }, + { + "epoch": 4.464642429085098, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11175 + }, + { + "epoch": 4.4650419496604075, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11176 + }, + { + "epoch": 4.465441470235717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11177 + }, + { + "epoch": 4.465840990811027, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11178 + }, + { + "epoch": 4.466240511386337, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11179 + }, + { + "epoch": 4.466640031961646, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11180 + }, + { + "epoch": 4.4670395525369555, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11181 + }, + { + "epoch": 4.467439073112265, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11182 + }, + { + "epoch": 4.467838593687575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11183 + }, + { + "epoch": 4.468238114262885, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11184 + }, + { + "epoch": 4.468637634838194, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 11185 + }, + { + "epoch": 4.4690371554135035, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11186 + }, + { + "epoch": 4.469436675988813, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11187 + }, + { + "epoch": 4.469836196564123, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11188 + }, + { + "epoch": 4.470235717139433, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11189 + }, + { + "epoch": 4.470635237714743, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 11190 + }, + { + "epoch": 4.471034758290052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11191 + }, + { + "epoch": 4.471434278865361, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11192 + }, + { + "epoch": 4.471833799440671, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11193 + }, + { + "epoch": 4.472233320015981, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11194 + }, + { + "epoch": 4.472632840591291, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11195 + }, + { + "epoch": 4.4730323611666005, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11196 + }, + { + "epoch": 4.473431881741909, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11197 + }, + { + "epoch": 4.473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11198 + }, + { + "epoch": 4.474230922892529, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11199 + }, + { + "epoch": 4.474630443467839, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11200 + }, + { + "epoch": 4.475029964043149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11201 + }, + { + "epoch": 4.4754294846184575, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 11202 + }, + { + "epoch": 4.475829005193767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 11203 + }, + { + "epoch": 4.476228525769077, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11204 + }, + { + "epoch": 4.476628046344387, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 11205 + }, + { + "epoch": 4.477027566919697, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11206 + }, + { + "epoch": 4.477427087495006, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11207 + }, + { + "epoch": 4.477826608070315, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11208 + }, + { + "epoch": 4.478226128645625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11209 + }, + { + "epoch": 4.478625649220935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11210 + }, + { + "epoch": 4.479025169796245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11211 + }, + { + "epoch": 4.479424690371554, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11212 + }, + { + "epoch": 4.479824210946864, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11213 + }, + { + "epoch": 4.480223731522173, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11214 + }, + { + "epoch": 4.480623252097483, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11215 + }, + { + "epoch": 4.481022772672793, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11216 + }, + { + "epoch": 4.4814222932481025, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11217 + }, + { + "epoch": 4.481821813823412, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11218 + }, + { + "epoch": 4.482221334398721, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11219 + }, + { + "epoch": 4.482620854974031, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11220 + }, + { + "epoch": 4.483020375549341, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11221 + }, + { + "epoch": 4.4834198961246505, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11222 + }, + { + "epoch": 4.48381941669996, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11223 + }, + { + "epoch": 4.484218937275269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11224 + }, + { + "epoch": 4.484618457850579, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11225 + }, + { + "epoch": 4.485017978425889, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11226 + }, + { + "epoch": 4.485417499001199, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11227 + }, + { + "epoch": 4.485817019576508, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11228 + }, + { + "epoch": 4.486216540151818, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11229 + }, + { + "epoch": 4.486616060727127, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11230 + }, + { + "epoch": 4.487015581302437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11231 + }, + { + "epoch": 4.487415101877747, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11232 + }, + { + "epoch": 4.487814622453056, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11233 + }, + { + "epoch": 4.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11234 + }, + { + "epoch": 4.488613663603676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11235 + }, + { + "epoch": 4.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11236 + }, + { + "epoch": 4.489412704754295, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11237 + }, + { + "epoch": 4.4898122253296044, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11238 + }, + { + "epoch": 4.490211745904914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11239 + }, + { + "epoch": 4.490611266480224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11240 + }, + { + "epoch": 4.491010787055533, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11241 + }, + { + "epoch": 4.491410307630843, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 11242 + }, + { + "epoch": 4.4918098282061525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11243 + }, + { + "epoch": 4.492209348781462, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11244 + }, + { + "epoch": 4.492608869356772, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11245 + }, + { + "epoch": 4.493008389932082, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11246 + }, + { + "epoch": 4.493407910507391, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11247 + }, + { + "epoch": 4.4938074310827005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11248 + }, + { + "epoch": 4.49420695165801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11249 + }, + { + "epoch": 4.49460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11250 + }, + { + "epoch": 4.49500599280863, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11251 + }, + { + "epoch": 4.49540551338394, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11252 + }, + { + "epoch": 4.495805033959249, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11253 + }, + { + "epoch": 4.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11254 + }, + { + "epoch": 4.496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11255 + }, + { + "epoch": 4.497003595685178, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11256 + }, + { + "epoch": 4.497403116260488, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11257 + }, + { + "epoch": 4.497802636835797, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 11258 + }, + { + "epoch": 4.498202157411106, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11259 + }, + { + "epoch": 4.498601677986416, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11260 + }, + { + "epoch": 4.499001198561726, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11261 + }, + { + "epoch": 4.499400719137036, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11262 + }, + { + "epoch": 4.4998002397123456, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11263 + }, + { + "epoch": 4.5001997602876544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11264 + }, + { + "epoch": 4.500599280862964, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11265 + }, + { + "epoch": 4.500998801438274, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11266 + }, + { + "epoch": 4.501398322013584, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11267 + }, + { + "epoch": 4.501797842588894, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11268 + }, + { + "epoch": 4.502197363164203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11269 + }, + { + "epoch": 4.502596883739512, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11270 + }, + { + "epoch": 4.502996404314822, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11271 + }, + { + "epoch": 4.503395924890132, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11272 + }, + { + "epoch": 4.503795445465442, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 11273 + }, + { + "epoch": 4.504194966040751, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11274 + }, + { + "epoch": 4.50459448661606, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11275 + }, + { + "epoch": 4.50499400719137, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11276 + }, + { + "epoch": 4.50539352776668, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11277 + }, + { + "epoch": 4.50579304834199, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11278 + }, + { + "epoch": 4.5061925689172995, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11279 + }, + { + "epoch": 4.506592089492609, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11280 + }, + { + "epoch": 4.506991610067918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11281 + }, + { + "epoch": 4.507391130643228, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11282 + }, + { + "epoch": 4.507790651218538, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11283 + }, + { + "epoch": 4.5081901717938475, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11284 + }, + { + "epoch": 4.508589692369157, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11285 + }, + { + "epoch": 4.508989212944467, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11286 + }, + { + "epoch": 4.509388733519776, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11287 + }, + { + "epoch": 4.509788254095086, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11288 + }, + { + "epoch": 4.5101877746703956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11289 + }, + { + "epoch": 4.510587295245705, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11290 + }, + { + "epoch": 4.510986815821015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11291 + }, + { + "epoch": 4.511386336396324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 11292 + }, + { + "epoch": 4.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11293 + }, + { + "epoch": 4.512185377546944, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11294 + }, + { + "epoch": 4.512584898122253, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11295 + }, + { + "epoch": 4.512984418697563, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11296 + }, + { + "epoch": 4.513383939272872, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11297 + }, + { + "epoch": 4.513783459848182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11298 + }, + { + "epoch": 4.514182980423492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11299 + }, + { + "epoch": 4.514582500998801, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11300 + }, + { + "epoch": 4.514982021574111, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11301 + }, + { + "epoch": 4.515381542149421, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.2975, + "step": 11302 + }, + { + "epoch": 4.515781062724731, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11303 + }, + { + "epoch": 4.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11304 + }, + { + "epoch": 4.5165801038753495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 11305 + }, + { + "epoch": 4.516979624450659, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 11306 + }, + { + "epoch": 4.517379145025969, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11307 + }, + { + "epoch": 4.517778665601279, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 11308 + }, + { + "epoch": 4.518178186176588, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11309 + }, + { + "epoch": 4.5185777067518975, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11310 + }, + { + "epoch": 4.518977227327207, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11311 + }, + { + "epoch": 4.519376747902517, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11312 + }, + { + "epoch": 4.519776268477827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11313 + }, + { + "epoch": 4.520175789053136, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11314 + }, + { + "epoch": 4.520575309628446, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11315 + }, + { + "epoch": 4.520974830203755, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 11316 + }, + { + "epoch": 4.521374350779065, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11317 + }, + { + "epoch": 4.521773871354375, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11318 + }, + { + "epoch": 4.522173391929685, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11319 + }, + { + "epoch": 4.5225729125049945, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 11320 + }, + { + "epoch": 4.522972433080303, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11321 + }, + { + "epoch": 4.523371953655613, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11322 + }, + { + "epoch": 4.523771474230923, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11323 + }, + { + "epoch": 4.524170994806233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11324 + }, + { + "epoch": 4.5245705153815425, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11325 + }, + { + "epoch": 4.524970035956851, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11326 + }, + { + "epoch": 4.525369556532161, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11327 + }, + { + "epoch": 4.525769077107471, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11328 + }, + { + "epoch": 4.526168597682781, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11329 + }, + { + "epoch": 4.526568118258091, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11330 + }, + { + "epoch": 4.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11331 + }, + { + "epoch": 4.527367159408709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11332 + }, + { + "epoch": 4.527766679984019, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11333 + }, + { + "epoch": 4.528166200559329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11334 + }, + { + "epoch": 4.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 11335 + }, + { + "epoch": 4.528965241709948, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11336 + }, + { + "epoch": 4.529364762285257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11337 + }, + { + "epoch": 4.529764282860567, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11338 + }, + { + "epoch": 4.530163803435877, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11339 + }, + { + "epoch": 4.530563324011187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11340 + }, + { + "epoch": 4.5309628445864965, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11341 + }, + { + "epoch": 4.531362365161806, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11342 + }, + { + "epoch": 4.531761885737115, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 11343 + }, + { + "epoch": 4.532161406312425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11344 + }, + { + "epoch": 4.532560926887735, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11345 + }, + { + "epoch": 4.5329604474630445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 11346 + }, + { + "epoch": 4.533359968038354, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11347 + }, + { + "epoch": 4.533759488613663, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11348 + }, + { + "epoch": 4.534159009188973, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11349 + }, + { + "epoch": 4.534558529764283, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 11350 + }, + { + "epoch": 4.5349580503395925, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11351 + }, + { + "epoch": 4.535357570914902, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11352 + }, + { + "epoch": 4.535757091490212, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 11353 + }, + { + "epoch": 4.536156612065521, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11354 + }, + { + "epoch": 4.536556132640831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11355 + }, + { + "epoch": 4.536955653216141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11356 + }, + { + "epoch": 4.53735517379145, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11357 + }, + { + "epoch": 4.53775469436676, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11358 + }, + { + "epoch": 4.53815421494207, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11359 + }, + { + "epoch": 4.538553735517379, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11360 + }, + { + "epoch": 4.538953256092689, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11361 + }, + { + "epoch": 4.539352776667998, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11362 + }, + { + "epoch": 4.539752297243308, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11363 + }, + { + "epoch": 4.540151817818618, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11364 + }, + { + "epoch": 4.540551338393927, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11365 + }, + { + "epoch": 4.540950858969237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11366 + }, + { + "epoch": 4.5413503795445465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11367 + }, + { + "epoch": 4.541749900119856, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11368 + }, + { + "epoch": 4.542149420695166, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11369 + }, + { + "epoch": 4.542548941270476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11370 + }, + { + "epoch": 4.542948461845785, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11371 + }, + { + "epoch": 4.5433479824210945, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11372 + }, + { + "epoch": 4.543747502996404, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11373 + }, + { + "epoch": 4.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11374 + }, + { + "epoch": 4.544546544147024, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11375 + }, + { + "epoch": 4.544946064722334, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11376 + }, + { + "epoch": 4.5453455852976425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11377 + }, + { + "epoch": 4.545745105872952, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11378 + }, + { + "epoch": 4.546144626448262, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11379 + }, + { + "epoch": 4.546544147023572, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11380 + }, + { + "epoch": 4.546943667598882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11381 + }, + { + "epoch": 4.547343188174191, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11382 + }, + { + "epoch": 4.5477427087495, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11383 + }, + { + "epoch": 4.54814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11384 + }, + { + "epoch": 4.54854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11385 + }, + { + "epoch": 4.54894127047543, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11386 + }, + { + "epoch": 4.549340791050739, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11387 + }, + { + "epoch": 4.549740311626048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11388 + }, + { + "epoch": 4.550139832201358, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11389 + }, + { + "epoch": 4.550539352776668, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11390 + }, + { + "epoch": 4.550938873351978, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11391 + }, + { + "epoch": 4.551338393927288, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11392 + }, + { + "epoch": 4.551737914502597, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11393 + }, + { + "epoch": 4.552137435077906, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11394 + }, + { + "epoch": 4.552536955653216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11395 + }, + { + "epoch": 4.552936476228526, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11396 + }, + { + "epoch": 4.553335996803836, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11397 + }, + { + "epoch": 4.553735517379145, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11398 + }, + { + "epoch": 4.554135037954454, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11399 + }, + { + "epoch": 4.554534558529764, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11400 + }, + { + "epoch": 4.554934079105074, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 11401 + }, + { + "epoch": 4.555333599680384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11402 + }, + { + "epoch": 4.555733120255693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11403 + }, + { + "epoch": 4.556132640831002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11404 + }, + { + "epoch": 4.556532161406312, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11405 + }, + { + "epoch": 4.556931681981622, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11406 + }, + { + "epoch": 4.557331202556932, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11407 + }, + { + "epoch": 4.5577307231322415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11408 + }, + { + "epoch": 4.558130243707551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 11409 + }, + { + "epoch": 4.558529764282861, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11410 + }, + { + "epoch": 4.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11411 + }, + { + "epoch": 4.55932880543348, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11412 + }, + { + "epoch": 4.5597283260087895, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11413 + }, + { + "epoch": 4.560127846584099, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11414 + }, + { + "epoch": 4.560527367159409, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 11415 + }, + { + "epoch": 4.560926887734718, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11416 + }, + { + "epoch": 4.561326408310028, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11417 + }, + { + "epoch": 4.561725928885338, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11418 + }, + { + "epoch": 4.562125449460647, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11419 + }, + { + "epoch": 4.562524970035957, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11420 + }, + { + "epoch": 4.562924490611266, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11421 + }, + { + "epoch": 4.563324011186576, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11422 + }, + { + "epoch": 4.563723531761886, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11423 + }, + { + "epoch": 4.564123052337195, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11424 + }, + { + "epoch": 4.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11425 + }, + { + "epoch": 4.564922093487815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11426 + }, + { + "epoch": 4.565321614063124, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11427 + }, + { + "epoch": 4.565721134638434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11428 + }, + { + "epoch": 4.566120655213743, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11429 + }, + { + "epoch": 4.566520175789053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11430 + }, + { + "epoch": 4.566919696364363, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11431 + }, + { + "epoch": 4.567319216939673, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11432 + }, + { + "epoch": 4.567718737514982, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11433 + }, + { + "epoch": 4.5681182580902915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11434 + }, + { + "epoch": 4.568517778665601, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11435 + }, + { + "epoch": 4.568917299240911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11436 + }, + { + "epoch": 4.569316819816221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11437 + }, + { + "epoch": 4.56971634039153, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11438 + }, + { + "epoch": 4.5701158609668395, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11439 + }, + { + "epoch": 4.570515381542149, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11440 + }, + { + "epoch": 4.570914902117459, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11441 + }, + { + "epoch": 4.571314422692769, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11442 + }, + { + "epoch": 4.571713943268079, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11443 + }, + { + "epoch": 4.572113463843388, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11444 + }, + { + "epoch": 4.572512984418697, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11445 + }, + { + "epoch": 4.572912504994007, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11446 + }, + { + "epoch": 4.573312025569317, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11447 + }, + { + "epoch": 4.573711546144627, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11448 + }, + { + "epoch": 4.5741110667199365, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11449 + }, + { + "epoch": 4.574510587295245, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11450 + }, + { + "epoch": 4.574910107870555, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11451 + }, + { + "epoch": 4.575309628445865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11452 + }, + { + "epoch": 4.575709149021175, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11453 + }, + { + "epoch": 4.5761086695964845, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11454 + }, + { + "epoch": 4.5765081901717934, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11455 + }, + { + "epoch": 4.576907710747103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11456 + }, + { + "epoch": 4.577307231322413, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11457 + }, + { + "epoch": 4.577706751897723, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11458 + }, + { + "epoch": 4.578106272473033, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11459 + }, + { + "epoch": 4.578505793048342, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11460 + }, + { + "epoch": 4.578905313623651, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11461 + }, + { + "epoch": 4.579304834198961, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11462 + }, + { + "epoch": 4.579704354774271, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11463 + }, + { + "epoch": 4.580103875349581, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11464 + }, + { + "epoch": 4.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11465 + }, + { + "epoch": 4.5809029165002, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11466 + }, + { + "epoch": 4.581302437075509, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11467 + }, + { + "epoch": 4.581701957650819, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11468 + }, + { + "epoch": 4.582101478226129, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11469 + }, + { + "epoch": 4.5825009988014385, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11470 + }, + { + "epoch": 4.582900519376748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11471 + }, + { + "epoch": 4.583300039952057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 11472 + }, + { + "epoch": 4.583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11473 + }, + { + "epoch": 4.584099081102677, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11474 + }, + { + "epoch": 4.5844986016779865, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11475 + }, + { + "epoch": 4.584898122253296, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11476 + }, + { + "epoch": 4.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11477 + }, + { + "epoch": 4.585697163403915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11478 + }, + { + "epoch": 4.586096683979225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11479 + }, + { + "epoch": 4.5864962045545346, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11480 + }, + { + "epoch": 4.586895725129844, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11481 + }, + { + "epoch": 4.587295245705154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11482 + }, + { + "epoch": 4.587694766280464, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11483 + }, + { + "epoch": 4.588094286855773, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11484 + }, + { + "epoch": 4.588493807431083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 11485 + }, + { + "epoch": 4.588893328006392, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11486 + }, + { + "epoch": 4.589292848581702, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 11487 + }, + { + "epoch": 4.589692369157012, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11488 + }, + { + "epoch": 4.590091889732321, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11489 + }, + { + "epoch": 4.590491410307631, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11490 + }, + { + "epoch": 4.59089093088294, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 11491 + }, + { + "epoch": 4.59129045145825, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11492 + }, + { + "epoch": 4.59168997203356, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11493 + }, + { + "epoch": 4.592089492608869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11494 + }, + { + "epoch": 4.592489013184179, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11495 + }, + { + "epoch": 4.5928885337594885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 11496 + }, + { + "epoch": 4.593288054334798, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11497 + }, + { + "epoch": 4.593687574910108, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11498 + }, + { + "epoch": 4.594087095485418, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11499 + }, + { + "epoch": 4.594486616060728, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11500 + }, + { + "epoch": 4.5948861366360365, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11501 + }, + { + "epoch": 4.595285657211346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11502 + }, + { + "epoch": 4.595685177786656, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11503 + }, + { + "epoch": 4.596084698361966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11504 + }, + { + "epoch": 4.596484218937276, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11505 + }, + { + "epoch": 4.5968837395125846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11506 + }, + { + "epoch": 4.597283260087894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11507 + }, + { + "epoch": 4.597682780663204, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11508 + }, + { + "epoch": 4.598082301238514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11509 + }, + { + "epoch": 4.598481821813824, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11510 + }, + { + "epoch": 4.598881342389133, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11511 + }, + { + "epoch": 4.599280862964442, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11512 + }, + { + "epoch": 4.599680383539752, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11513 + }, + { + "epoch": 4.600079904115062, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11514 + }, + { + "epoch": 4.600479424690372, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11515 + }, + { + "epoch": 4.6008789452656815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11516 + }, + { + "epoch": 4.60127846584099, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11517 + }, + { + "epoch": 4.6016779864163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11518 + }, + { + "epoch": 4.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11519 + }, + { + "epoch": 4.60247702756692, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11520 + }, + { + "epoch": 4.60287654814223, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11521 + }, + { + "epoch": 4.603276068717539, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11522 + }, + { + "epoch": 4.603675589292848, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11523 + }, + { + "epoch": 4.604075109868158, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11524 + }, + { + "epoch": 4.604474630443468, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11525 + }, + { + "epoch": 4.604874151018778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11526 + }, + { + "epoch": 4.605273671594087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11527 + }, + { + "epoch": 4.605673192169396, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11528 + }, + { + "epoch": 4.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11529 + }, + { + "epoch": 4.606472233320016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11530 + }, + { + "epoch": 4.606871753895326, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11531 + }, + { + "epoch": 4.6072712744706354, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 11532 + }, + { + "epoch": 4.607670795045945, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11533 + }, + { + "epoch": 4.608070315621254, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11534 + }, + { + "epoch": 4.608469836196564, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11535 + }, + { + "epoch": 4.608869356771874, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11536 + }, + { + "epoch": 4.6092688773471835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11537 + }, + { + "epoch": 4.609668397922493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11538 + }, + { + "epoch": 4.610067918497803, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11539 + }, + { + "epoch": 4.610467439073112, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11540 + }, + { + "epoch": 4.610866959648422, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11541 + }, + { + "epoch": 4.6112664802237315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11542 + }, + { + "epoch": 4.611666000799041, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11543 + }, + { + "epoch": 4.612065521374351, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11544 + }, + { + "epoch": 4.61246504194966, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11545 + }, + { + "epoch": 4.61286456252497, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11546 + }, + { + "epoch": 4.61326408310028, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11547 + }, + { + "epoch": 4.613663603675589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11548 + }, + { + "epoch": 4.614063124250899, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11549 + }, + { + "epoch": 4.614462644826209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11550 + }, + { + "epoch": 4.614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11551 + }, + { + "epoch": 4.615261685976828, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 11552 + }, + { + "epoch": 4.615661206552137, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11553 + }, + { + "epoch": 4.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11554 + }, + { + "epoch": 4.616460247702757, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11555 + }, + { + "epoch": 4.616859768278067, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11556 + }, + { + "epoch": 4.617259288853376, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11557 + }, + { + "epoch": 4.6176588094286855, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11558 + }, + { + "epoch": 4.618058330003995, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11559 + }, + { + "epoch": 4.618457850579305, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11560 + }, + { + "epoch": 4.618857371154615, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11561 + }, + { + "epoch": 4.619256891729924, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11562 + }, + { + "epoch": 4.6196564123052335, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11563 + }, + { + "epoch": 4.620055932880543, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11564 + }, + { + "epoch": 4.620455453455853, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 11565 + }, + { + "epoch": 4.620854974031163, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11566 + }, + { + "epoch": 4.621254494606472, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11567 + }, + { + "epoch": 4.6216540151817815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11568 + }, + { + "epoch": 4.622053535757091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11569 + }, + { + "epoch": 4.622453056332401, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11570 + }, + { + "epoch": 4.622852576907711, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11571 + }, + { + "epoch": 4.623252097483021, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 11572 + }, + { + "epoch": 4.6236516180583305, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11573 + }, + { + "epoch": 4.624051138633639, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11574 + }, + { + "epoch": 4.624450659208949, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11575 + }, + { + "epoch": 4.624850179784259, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11576 + }, + { + "epoch": 4.625249700359569, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11577 + }, + { + "epoch": 4.6256492209348785, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11578 + }, + { + "epoch": 4.626048741510187, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11579 + }, + { + "epoch": 4.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11580 + }, + { + "epoch": 4.626847782660807, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11581 + }, + { + "epoch": 4.627247303236117, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11582 + }, + { + "epoch": 4.627646823811427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 11583 + }, + { + "epoch": 4.6280463443867355, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11584 + }, + { + "epoch": 4.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11585 + }, + { + "epoch": 4.628845385537355, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11586 + }, + { + "epoch": 4.629244906112665, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11587 + }, + { + "epoch": 4.629644426687975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11588 + }, + { + "epoch": 4.630043947263284, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11589 + }, + { + "epoch": 4.630443467838594, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11590 + }, + { + "epoch": 4.630842988413903, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 11591 + }, + { + "epoch": 4.631242508989213, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11592 + }, + { + "epoch": 4.631642029564523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11593 + }, + { + "epoch": 4.632041550139832, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11594 + }, + { + "epoch": 4.632441070715142, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 11595 + }, + { + "epoch": 4.632840591290451, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11596 + }, + { + "epoch": 4.633240111865761, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11597 + }, + { + "epoch": 4.633639632441071, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11598 + }, + { + "epoch": 4.6340391530163805, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11599 + }, + { + "epoch": 4.63443867359169, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11600 + }, + { + "epoch": 4.634838194166999, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11601 + }, + { + "epoch": 4.635237714742309, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11602 + }, + { + "epoch": 4.635637235317619, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11603 + }, + { + "epoch": 4.6360367558929285, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11604 + }, + { + "epoch": 4.636436276468238, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11605 + }, + { + "epoch": 4.636835797043548, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11606 + }, + { + "epoch": 4.637235317618857, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11607 + }, + { + "epoch": 4.637634838194167, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11608 + }, + { + "epoch": 4.638034358769477, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11609 + }, + { + "epoch": 4.638433879344786, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 11610 + }, + { + "epoch": 4.638833399920096, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11611 + }, + { + "epoch": 4.639232920495406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11612 + }, + { + "epoch": 4.639632441070715, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11613 + }, + { + "epoch": 4.640031961646025, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11614 + }, + { + "epoch": 4.640431482221334, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11615 + }, + { + "epoch": 4.640831002796644, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11616 + }, + { + "epoch": 4.641230523371954, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11617 + }, + { + "epoch": 4.641630043947263, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11618 + }, + { + "epoch": 4.642029564522573, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11619 + }, + { + "epoch": 4.642429085097882, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11620 + }, + { + "epoch": 4.642828605673192, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11621 + }, + { + "epoch": 4.643228126248502, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11622 + }, + { + "epoch": 4.643627646823812, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11623 + }, + { + "epoch": 4.644027167399121, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11624 + }, + { + "epoch": 4.6444266879744305, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11625 + }, + { + "epoch": 4.64482620854974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11626 + }, + { + "epoch": 4.64522572912505, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11627 + }, + { + "epoch": 4.64562524970036, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11628 + }, + { + "epoch": 4.64602477027567, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11629 + }, + { + "epoch": 4.6464242908509785, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11630 + }, + { + "epoch": 4.646823811426288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11631 + }, + { + "epoch": 4.647223332001598, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11632 + }, + { + "epoch": 4.647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11633 + }, + { + "epoch": 4.648022373152218, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11634 + }, + { + "epoch": 4.648421893727527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11635 + }, + { + "epoch": 4.648821414302836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11636 + }, + { + "epoch": 4.649220934878146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11637 + }, + { + "epoch": 4.649620455453456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11638 + }, + { + "epoch": 4.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11639 + }, + { + "epoch": 4.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11640 + }, + { + "epoch": 4.650819017179384, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11641 + }, + { + "epoch": 4.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 11642 + }, + { + "epoch": 4.651618058330004, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11643 + }, + { + "epoch": 4.652017578905314, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11644 + }, + { + "epoch": 4.6524170994806235, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11645 + }, + { + "epoch": 4.652816620055933, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11646 + }, + { + "epoch": 4.653216140631242, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11647 + }, + { + "epoch": 4.653615661206552, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11648 + }, + { + "epoch": 4.654015181781862, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11649 + }, + { + "epoch": 4.654414702357172, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11650 + }, + { + "epoch": 4.654814222932481, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11651 + }, + { + "epoch": 4.65521374350779, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11652 + }, + { + "epoch": 4.6556132640831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11653 + }, + { + "epoch": 4.65601278465841, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11654 + }, + { + "epoch": 4.65641230523372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11655 + }, + { + "epoch": 4.656811825809029, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11656 + }, + { + "epoch": 4.657211346384338, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11657 + }, + { + "epoch": 4.657610866959648, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11658 + }, + { + "epoch": 4.658010387534958, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11659 + }, + { + "epoch": 4.658409908110268, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11660 + }, + { + "epoch": 4.6588094286855775, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11661 + }, + { + "epoch": 4.659208949260887, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11662 + }, + { + "epoch": 4.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11663 + }, + { + "epoch": 4.660007990411506, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11664 + }, + { + "epoch": 4.660407510986816, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11665 + }, + { + "epoch": 4.6608070315621255, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11666 + }, + { + "epoch": 4.661206552137435, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11667 + }, + { + "epoch": 4.661606072712745, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11668 + }, + { + "epoch": 4.662005593288054, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11669 + }, + { + "epoch": 4.662405113863364, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11670 + }, + { + "epoch": 4.6628046344386735, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11671 + }, + { + "epoch": 4.663204155013983, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11672 + }, + { + "epoch": 4.663603675589293, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11673 + }, + { + "epoch": 4.664003196164602, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11674 + }, + { + "epoch": 4.664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11675 + }, + { + "epoch": 4.664802237315222, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11676 + }, + { + "epoch": 4.665201757890531, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11677 + }, + { + "epoch": 4.665601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11678 + }, + { + "epoch": 4.666000799041151, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11679 + }, + { + "epoch": 4.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11680 + }, + { + "epoch": 4.66679984019177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11681 + }, + { + "epoch": 4.667199360767079, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11682 + }, + { + "epoch": 4.667598881342389, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11683 + }, + { + "epoch": 4.667998401917699, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11684 + }, + { + "epoch": 4.668397922493009, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11685 + }, + { + "epoch": 4.668797443068318, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11686 + }, + { + "epoch": 4.6691969636436275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11687 + }, + { + "epoch": 4.669596484218937, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11688 + }, + { + "epoch": 4.669996004794247, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11689 + }, + { + "epoch": 4.670395525369557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11690 + }, + { + "epoch": 4.670795045944866, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11691 + }, + { + "epoch": 4.6711945665201755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11692 + }, + { + "epoch": 4.671594087095485, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11693 + }, + { + "epoch": 4.671993607670795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11694 + }, + { + "epoch": 4.672393128246105, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11695 + }, + { + "epoch": 4.672792648821415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11696 + }, + { + "epoch": 4.6731921693967235, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11697 + }, + { + "epoch": 4.673591689972033, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11698 + }, + { + "epoch": 4.673991210547343, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11699 + }, + { + "epoch": 4.674390731122653, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 11700 + }, + { + "epoch": 4.674790251697963, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11701 + }, + { + "epoch": 4.6751897722732725, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11702 + }, + { + "epoch": 4.675589292848581, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11703 + }, + { + "epoch": 4.675988813423891, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11704 + }, + { + "epoch": 4.676388333999201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11705 + }, + { + "epoch": 4.676787854574511, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 11706 + }, + { + "epoch": 4.6771873751498205, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11707 + }, + { + "epoch": 4.677586895725129, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11708 + }, + { + "epoch": 4.677986416300439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11709 + }, + { + "epoch": 4.678385936875749, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11710 + }, + { + "epoch": 4.678785457451059, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11711 + }, + { + "epoch": 4.679184978026369, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11712 + }, + { + "epoch": 4.679584498601678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11713 + }, + { + "epoch": 4.679984019176987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11714 + }, + { + "epoch": 4.680383539752297, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11715 + }, + { + "epoch": 4.680783060327607, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11716 + }, + { + "epoch": 4.681182580902917, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11717 + }, + { + "epoch": 4.681582101478226, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11718 + }, + { + "epoch": 4.681981622053536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11719 + }, + { + "epoch": 4.682381142628845, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11720 + }, + { + "epoch": 4.682780663204155, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11721 + }, + { + "epoch": 4.683180183779465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11722 + }, + { + "epoch": 4.683579704354774, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 11723 + }, + { + "epoch": 4.683979224930084, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11724 + }, + { + "epoch": 4.684378745505393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11725 + }, + { + "epoch": 4.684778266080703, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11726 + }, + { + "epoch": 4.685177786656013, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11727 + }, + { + "epoch": 4.6855773072313225, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11728 + }, + { + "epoch": 4.685976827806632, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11729 + }, + { + "epoch": 4.686376348381942, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11730 + }, + { + "epoch": 4.686775868957251, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11731 + }, + { + "epoch": 4.687175389532561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11732 + }, + { + "epoch": 4.6875749101078705, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11733 + }, + { + "epoch": 4.68797443068318, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11734 + }, + { + "epoch": 4.68837395125849, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11735 + }, + { + "epoch": 4.6887734718338, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11736 + }, + { + "epoch": 4.689172992409109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11737 + }, + { + "epoch": 4.689572512984419, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11738 + }, + { + "epoch": 4.689972033559728, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11739 + }, + { + "epoch": 4.690371554135038, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 11740 + }, + { + "epoch": 4.690771074710348, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11741 + }, + { + "epoch": 4.691170595285657, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 11742 + }, + { + "epoch": 4.691570115860967, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11743 + }, + { + "epoch": 4.691969636436276, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11744 + }, + { + "epoch": 4.692369157011586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11745 + }, + { + "epoch": 4.692768677586896, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11746 + }, + { + "epoch": 4.693168198162205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11747 + }, + { + "epoch": 4.693567718737515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11748 + }, + { + "epoch": 4.6939672393128244, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 11749 + }, + { + "epoch": 4.694366759888134, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11750 + }, + { + "epoch": 4.694766280463444, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11751 + }, + { + "epoch": 4.695165801038754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11752 + }, + { + "epoch": 4.695565321614064, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11753 + }, + { + "epoch": 4.6959648421893725, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11754 + }, + { + "epoch": 4.696364362764682, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11755 + }, + { + "epoch": 4.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11756 + }, + { + "epoch": 4.697163403915302, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11757 + }, + { + "epoch": 4.697562924490612, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11758 + }, + { + "epoch": 4.6979624450659205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11759 + }, + { + "epoch": 4.69836196564123, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11760 + }, + { + "epoch": 4.69876148621654, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11761 + }, + { + "epoch": 4.69916100679185, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 11762 + }, + { + "epoch": 4.69956052736716, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11763 + }, + { + "epoch": 4.699960047942469, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11764 + }, + { + "epoch": 4.700359568517778, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11765 + }, + { + "epoch": 4.700759089093088, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11766 + }, + { + "epoch": 4.701158609668398, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11767 + }, + { + "epoch": 4.701558130243708, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11768 + }, + { + "epoch": 4.7019576508190175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11769 + }, + { + "epoch": 4.702357171394327, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11770 + }, + { + "epoch": 4.702756691969636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11771 + }, + { + "epoch": 4.703156212544946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11772 + }, + { + "epoch": 4.703555733120256, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 11773 + }, + { + "epoch": 4.7039552536955656, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11774 + }, + { + "epoch": 4.704354774270875, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11775 + }, + { + "epoch": 4.704754294846184, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11776 + }, + { + "epoch": 4.705153815421494, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11777 + }, + { + "epoch": 4.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11778 + }, + { + "epoch": 4.705952856572114, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11779 + }, + { + "epoch": 4.706352377147423, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11780 + }, + { + "epoch": 4.706751897722732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11781 + }, + { + "epoch": 4.707151418298042, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 11782 + }, + { + "epoch": 4.707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11783 + }, + { + "epoch": 4.707950459448662, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11784 + }, + { + "epoch": 4.708349980023971, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11785 + }, + { + "epoch": 4.708749500599281, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11786 + }, + { + "epoch": 4.70914902117459, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11787 + }, + { + "epoch": 4.7095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11788 + }, + { + "epoch": 4.70994806232521, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11789 + }, + { + "epoch": 4.7103475829005195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11790 + }, + { + "epoch": 4.710747103475829, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11791 + }, + { + "epoch": 4.711146624051139, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11792 + }, + { + "epoch": 4.711546144626448, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11793 + }, + { + "epoch": 4.711945665201758, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11794 + }, + { + "epoch": 4.7123451857770675, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 11795 + }, + { + "epoch": 4.712744706352377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11796 + }, + { + "epoch": 4.713144226927687, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11797 + }, + { + "epoch": 4.713543747502996, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11798 + }, + { + "epoch": 4.713943268078306, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11799 + }, + { + "epoch": 4.714342788653616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11800 + }, + { + "epoch": 4.714742309228925, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 11801 + }, + { + "epoch": 4.715141829804235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11802 + }, + { + "epoch": 4.715541350379545, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11803 + }, + { + "epoch": 4.715940870954854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11804 + }, + { + "epoch": 4.716340391530164, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11805 + }, + { + "epoch": 4.716739912105473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11806 + }, + { + "epoch": 4.717139432680783, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11807 + }, + { + "epoch": 4.717538953256093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11808 + }, + { + "epoch": 4.717938473831403, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11809 + }, + { + "epoch": 4.718337994406712, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11810 + }, + { + "epoch": 4.718737514982021, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11811 + }, + { + "epoch": 4.719137035557331, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11812 + }, + { + "epoch": 4.719536556132641, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11813 + }, + { + "epoch": 4.719936076707951, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11814 + }, + { + "epoch": 4.72033559728326, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11815 + }, + { + "epoch": 4.7207351178585695, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11816 + }, + { + "epoch": 4.721134638433879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11817 + }, + { + "epoch": 4.721534159009189, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11818 + }, + { + "epoch": 4.721933679584499, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11819 + }, + { + "epoch": 4.722333200159808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11820 + }, + { + "epoch": 4.7227327207351175, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11821 + }, + { + "epoch": 4.723132241310427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 11822 + }, + { + "epoch": 4.723531761885737, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11823 + }, + { + "epoch": 4.723931282461047, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11824 + }, + { + "epoch": 4.724330803036357, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11825 + }, + { + "epoch": 4.7247303236116664, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11826 + }, + { + "epoch": 4.725129844186975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11827 + }, + { + "epoch": 4.725529364762285, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11828 + }, + { + "epoch": 4.725928885337595, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11829 + }, + { + "epoch": 4.726328405912905, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11830 + }, + { + "epoch": 4.7267279264882145, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11831 + }, + { + "epoch": 4.727127447063523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11832 + }, + { + "epoch": 4.727526967638833, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11833 + }, + { + "epoch": 4.727926488214143, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11834 + }, + { + "epoch": 4.728326008789453, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11835 + }, + { + "epoch": 4.7287255293647625, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 11836 + }, + { + "epoch": 4.729125049940071, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11837 + }, + { + "epoch": 4.729524570515381, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11838 + }, + { + "epoch": 4.729924091090691, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11839 + }, + { + "epoch": 4.730323611666001, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11840 + }, + { + "epoch": 4.730723132241311, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11841 + }, + { + "epoch": 4.73112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11842 + }, + { + "epoch": 4.73152217339193, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11843 + }, + { + "epoch": 4.731921693967239, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11844 + }, + { + "epoch": 4.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11845 + }, + { + "epoch": 4.732720735117859, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11846 + }, + { + "epoch": 4.733120255693168, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11847 + }, + { + "epoch": 4.733519776268478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11848 + }, + { + "epoch": 4.733919296843787, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11849 + }, + { + "epoch": 4.734318817419097, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11850 + }, + { + "epoch": 4.734718337994407, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 11851 + }, + { + "epoch": 4.7351178585697165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11852 + }, + { + "epoch": 4.735517379145026, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11853 + }, + { + "epoch": 4.735916899720335, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11854 + }, + { + "epoch": 4.736316420295645, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11855 + }, + { + "epoch": 4.736715940870955, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11856 + }, + { + "epoch": 4.7371154614462645, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11857 + }, + { + "epoch": 4.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11858 + }, + { + "epoch": 4.737914502596884, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 11859 + }, + { + "epoch": 4.738314023172194, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11860 + }, + { + "epoch": 4.738713543747503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11861 + }, + { + "epoch": 4.7391130643228125, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11862 + }, + { + "epoch": 4.739512584898122, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11863 + }, + { + "epoch": 4.739912105473432, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11864 + }, + { + "epoch": 4.740311626048742, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11865 + }, + { + "epoch": 4.740711146624051, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11866 + }, + { + "epoch": 4.741110667199361, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11867 + }, + { + "epoch": 4.74151018777467, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11868 + }, + { + "epoch": 4.74190970834998, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11869 + }, + { + "epoch": 4.74230922892529, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11870 + }, + { + "epoch": 4.742708749500599, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11871 + }, + { + "epoch": 4.743108270075909, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11872 + }, + { + "epoch": 4.743507790651218, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11873 + }, + { + "epoch": 4.743907311226528, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11874 + }, + { + "epoch": 4.744306831801838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11875 + }, + { + "epoch": 4.744706352377148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11876 + }, + { + "epoch": 4.745105872952457, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11877 + }, + { + "epoch": 4.7455053935277665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11878 + }, + { + "epoch": 4.745904914103076, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11879 + }, + { + "epoch": 4.746304434678386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11880 + }, + { + "epoch": 4.746703955253696, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11881 + }, + { + "epoch": 4.747103475829006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11882 + }, + { + "epoch": 4.7475029964043145, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11883 + }, + { + "epoch": 4.747902516979624, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11884 + }, + { + "epoch": 4.748302037554934, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11885 + }, + { + "epoch": 4.748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11886 + }, + { + "epoch": 4.749101078705554, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11887 + }, + { + "epoch": 4.7495005992808625, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11888 + }, + { + "epoch": 4.749900119856172, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11889 + }, + { + "epoch": 4.750299640431482, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11890 + }, + { + "epoch": 4.750699161006792, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11891 + }, + { + "epoch": 4.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11892 + }, + { + "epoch": 4.7514982021574115, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11893 + }, + { + "epoch": 4.75189772273272, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11894 + }, + { + "epoch": 4.75229724330803, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11895 + }, + { + "epoch": 4.75269676388334, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11896 + }, + { + "epoch": 4.75309628445865, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11897 + }, + { + "epoch": 4.7534958050339595, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11898 + }, + { + "epoch": 4.753895325609269, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11899 + }, + { + "epoch": 4.754294846184578, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11900 + }, + { + "epoch": 4.754694366759888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11901 + }, + { + "epoch": 4.755093887335198, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11902 + }, + { + "epoch": 4.755493407910508, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11903 + }, + { + "epoch": 4.755892928485817, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11904 + }, + { + "epoch": 4.756292449061126, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11905 + }, + { + "epoch": 4.756691969636436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11906 + }, + { + "epoch": 4.757091490211746, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11907 + }, + { + "epoch": 4.757491010787056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11908 + }, + { + "epoch": 4.757890531362365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11909 + }, + { + "epoch": 4.758290051937674, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11910 + }, + { + "epoch": 4.758689572512984, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11911 + }, + { + "epoch": 4.759089093088294, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11912 + }, + { + "epoch": 4.759488613663604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11913 + }, + { + "epoch": 4.759888134238913, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11914 + }, + { + "epoch": 4.760287654814223, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11915 + }, + { + "epoch": 4.760687175389533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11916 + }, + { + "epoch": 4.761086695964842, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11917 + }, + { + "epoch": 4.761486216540152, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11918 + }, + { + "epoch": 4.7618857371154615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11919 + }, + { + "epoch": 4.762285257690771, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 11920 + }, + { + "epoch": 4.762684778266081, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11921 + }, + { + "epoch": 4.76308429884139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11922 + }, + { + "epoch": 4.7634838194167, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 11923 + }, + { + "epoch": 4.7638833399920095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11924 + }, + { + "epoch": 4.764282860567319, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11925 + }, + { + "epoch": 4.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11926 + }, + { + "epoch": 4.765081901717938, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11927 + }, + { + "epoch": 4.765481422293248, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11928 + }, + { + "epoch": 4.765880942868558, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11929 + }, + { + "epoch": 4.766280463443867, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11930 + }, + { + "epoch": 4.766679984019177, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11931 + }, + { + "epoch": 4.767079504594487, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11932 + }, + { + "epoch": 4.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11933 + }, + { + "epoch": 4.767878545745106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11934 + }, + { + "epoch": 4.768278066320415, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11935 + }, + { + "epoch": 4.768677586895725, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11936 + }, + { + "epoch": 4.769077107471035, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11937 + }, + { + "epoch": 4.769476628046345, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11938 + }, + { + "epoch": 4.769876148621654, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11939 + }, + { + "epoch": 4.770275669196963, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11940 + }, + { + "epoch": 4.770675189772273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11941 + }, + { + "epoch": 4.771074710347583, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11942 + }, + { + "epoch": 4.771474230922893, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11943 + }, + { + "epoch": 4.771873751498202, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11944 + }, + { + "epoch": 4.7722732720735115, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11945 + }, + { + "epoch": 4.772672792648821, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11946 + }, + { + "epoch": 4.773072313224131, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11947 + }, + { + "epoch": 4.773471833799441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11948 + }, + { + "epoch": 4.773871354374751, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11949 + }, + { + "epoch": 4.77427087495006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11950 + }, + { + "epoch": 4.774670395525369, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11951 + }, + { + "epoch": 4.775069916100679, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 11952 + }, + { + "epoch": 4.775469436675989, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11953 + }, + { + "epoch": 4.775868957251299, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11954 + }, + { + "epoch": 4.7762684778266085, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11955 + }, + { + "epoch": 4.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11956 + }, + { + "epoch": 4.777067518977227, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11957 + }, + { + "epoch": 4.777467039552537, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11958 + }, + { + "epoch": 4.777866560127847, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11959 + }, + { + "epoch": 4.7782660807031565, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11960 + }, + { + "epoch": 4.778665601278465, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11961 + }, + { + "epoch": 4.779065121853775, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11962 + }, + { + "epoch": 4.779464642429085, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11963 + }, + { + "epoch": 4.779864163004395, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11964 + }, + { + "epoch": 4.7802636835797045, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11965 + }, + { + "epoch": 4.780663204155014, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11966 + }, + { + "epoch": 4.781062724730323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 11967 + }, + { + "epoch": 4.781462245305633, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11968 + }, + { + "epoch": 4.781861765880943, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11969 + }, + { + "epoch": 4.782261286456253, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11970 + }, + { + "epoch": 4.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11971 + }, + { + "epoch": 4.783060327606872, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11972 + }, + { + "epoch": 4.783459848182181, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11973 + }, + { + "epoch": 4.783859368757491, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11974 + }, + { + "epoch": 4.784258889332801, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 11975 + }, + { + "epoch": 4.78465840990811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11976 + }, + { + "epoch": 4.78505793048342, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11977 + }, + { + "epoch": 4.785457451058729, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11978 + }, + { + "epoch": 4.785856971634039, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11979 + }, + { + "epoch": 4.786256492209349, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11980 + }, + { + "epoch": 4.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11981 + }, + { + "epoch": 4.787055533359968, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11982 + }, + { + "epoch": 4.787455053935278, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11983 + }, + { + "epoch": 4.787854574510587, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11984 + }, + { + "epoch": 4.788254095085897, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11985 + }, + { + "epoch": 4.7886536156612065, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11986 + }, + { + "epoch": 4.789053136236516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11987 + }, + { + "epoch": 4.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11988 + }, + { + "epoch": 4.789852177387136, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11989 + }, + { + "epoch": 4.790251697962445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11990 + }, + { + "epoch": 4.7906512185377546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11991 + }, + { + "epoch": 4.791050739113064, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11992 + }, + { + "epoch": 4.791450259688374, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11993 + }, + { + "epoch": 4.791849780263684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11994 + }, + { + "epoch": 4.792249300838993, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11995 + }, + { + "epoch": 4.792648821414303, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11996 + }, + { + "epoch": 4.793048341989612, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11997 + }, + { + "epoch": 4.793447862564922, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11998 + }, + { + "epoch": 4.793847383140232, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 11999 + }, + { + "epoch": 4.794246903715541, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12000 + }, + { + "epoch": 4.794646424290851, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12001 + }, + { + "epoch": 4.79504594486616, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12002 + }, + { + "epoch": 4.79544546544147, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12003 + }, + { + "epoch": 4.79584498601678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12004 + }, + { + "epoch": 4.79624450659209, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12005 + }, + { + "epoch": 4.7966440271674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12006 + }, + { + "epoch": 4.7970435477427085, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12007 + }, + { + "epoch": 4.797443068318018, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12008 + }, + { + "epoch": 4.797842588893328, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12009 + }, + { + "epoch": 4.798242109468638, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12010 + }, + { + "epoch": 4.798641630043948, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12011 + }, + { + "epoch": 4.7990411506192565, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12012 + }, + { + "epoch": 4.799440671194566, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 12013 + }, + { + "epoch": 4.799840191769876, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12014 + }, + { + "epoch": 4.800239712345186, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12015 + }, + { + "epoch": 4.800639232920496, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12016 + }, + { + "epoch": 4.801038753495805, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12017 + }, + { + "epoch": 4.801438274071114, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12018 + }, + { + "epoch": 4.801837794646424, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12019 + }, + { + "epoch": 4.802237315221734, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12020 + }, + { + "epoch": 4.802636835797044, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12021 + }, + { + "epoch": 4.8030363563723535, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12022 + }, + { + "epoch": 4.803435876947663, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12023 + }, + { + "epoch": 4.803835397522972, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12024 + }, + { + "epoch": 4.804234918098282, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12025 + }, + { + "epoch": 4.804634438673592, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12026 + }, + { + "epoch": 4.8050339592489015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12027 + }, + { + "epoch": 4.805433479824211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12028 + }, + { + "epoch": 4.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12029 + }, + { + "epoch": 4.80623252097483, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12030 + }, + { + "epoch": 4.80663204155014, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12031 + }, + { + "epoch": 4.80703156212545, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12032 + }, + { + "epoch": 4.807431082700759, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12033 + }, + { + "epoch": 4.807830603276068, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12034 + }, + { + "epoch": 4.808230123851378, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12035 + }, + { + "epoch": 4.808629644426688, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12036 + }, + { + "epoch": 4.809029165001998, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12037 + }, + { + "epoch": 4.809428685577307, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12038 + }, + { + "epoch": 4.809828206152617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12039 + }, + { + "epoch": 4.810227726727927, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12040 + }, + { + "epoch": 4.810627247303236, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12041 + }, + { + "epoch": 4.811026767878546, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12042 + }, + { + "epoch": 4.8114262884538554, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12043 + }, + { + "epoch": 4.811825809029165, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12044 + }, + { + "epoch": 4.812225329604475, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 12045 + }, + { + "epoch": 4.812624850179784, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12046 + }, + { + "epoch": 4.813024370755094, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12047 + }, + { + "epoch": 4.8134238913304035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12048 + }, + { + "epoch": 4.813823411905713, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12049 + }, + { + "epoch": 4.814222932481023, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12050 + }, + { + "epoch": 4.814622453056332, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12051 + }, + { + "epoch": 4.815021973631642, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12052 + }, + { + "epoch": 4.8154214942069515, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12053 + }, + { + "epoch": 4.815821014782261, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12054 + }, + { + "epoch": 4.816220535357571, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12055 + }, + { + "epoch": 4.816620055932881, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12056 + }, + { + "epoch": 4.81701957650819, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12057 + }, + { + "epoch": 4.8174190970835, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12058 + }, + { + "epoch": 4.817818617658809, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12059 + }, + { + "epoch": 4.818218138234119, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12060 + }, + { + "epoch": 4.818617658809429, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12061 + }, + { + "epoch": 4.819017179384739, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12062 + }, + { + "epoch": 4.819416699960048, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12063 + }, + { + "epoch": 4.819816220535357, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12064 + }, + { + "epoch": 4.820215741110667, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12065 + }, + { + "epoch": 4.820615261685977, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12066 + }, + { + "epoch": 4.821014782261287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12067 + }, + { + "epoch": 4.821414302836596, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12068 + }, + { + "epoch": 4.8218138234119055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12069 + }, + { + "epoch": 4.822213343987215, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12070 + }, + { + "epoch": 4.822612864562525, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12071 + }, + { + "epoch": 4.823012385137835, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12072 + }, + { + "epoch": 4.823411905713145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12073 + }, + { + "epoch": 4.8238114262884535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12074 + }, + { + "epoch": 4.824210946863763, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12075 + }, + { + "epoch": 4.824610467439073, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12076 + }, + { + "epoch": 4.825009988014383, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12077 + }, + { + "epoch": 4.825409508589693, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 12078 + }, + { + "epoch": 4.825809029165002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12079 + }, + { + "epoch": 4.826208549740311, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12080 + }, + { + "epoch": 4.826608070315621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12081 + }, + { + "epoch": 4.827007590890931, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12082 + }, + { + "epoch": 4.827407111466241, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12083 + }, + { + "epoch": 4.8278066320415505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12084 + }, + { + "epoch": 4.828206152616859, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12085 + }, + { + "epoch": 4.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12086 + }, + { + "epoch": 4.829005193767479, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12087 + }, + { + "epoch": 4.829404714342789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12088 + }, + { + "epoch": 4.8298042349180985, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12089 + }, + { + "epoch": 4.830203755493407, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12090 + }, + { + "epoch": 4.830603276068717, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12091 + }, + { + "epoch": 4.831002796644027, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12092 + }, + { + "epoch": 4.831402317219337, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12093 + }, + { + "epoch": 4.831801837794647, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 12094 + }, + { + "epoch": 4.832201358369956, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12095 + }, + { + "epoch": 4.832600878945266, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12096 + }, + { + "epoch": 4.833000399520575, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12097 + }, + { + "epoch": 4.833399920095885, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12098 + }, + { + "epoch": 4.833799440671195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12099 + }, + { + "epoch": 4.834198961246504, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12100 + }, + { + "epoch": 4.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12101 + }, + { + "epoch": 4.834998002397123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12102 + }, + { + "epoch": 4.835397522972433, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12103 + }, + { + "epoch": 4.835797043547743, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12104 + }, + { + "epoch": 4.836196564123052, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12105 + }, + { + "epoch": 4.836596084698362, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12106 + }, + { + "epoch": 4.836995605273671, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12107 + }, + { + "epoch": 4.837395125848981, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12108 + }, + { + "epoch": 4.837794646424291, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12109 + }, + { + "epoch": 4.8381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12110 + }, + { + "epoch": 4.83859368757491, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12111 + }, + { + "epoch": 4.83899320815022, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12112 + }, + { + "epoch": 4.83939272872553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12113 + }, + { + "epoch": 4.839792249300839, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12114 + }, + { + "epoch": 4.8401917698761485, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12115 + }, + { + "epoch": 4.840591290451458, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12116 + }, + { + "epoch": 4.840990811026768, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12117 + }, + { + "epoch": 4.841390331602078, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12118 + }, + { + "epoch": 4.841789852177387, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12119 + }, + { + "epoch": 4.842189372752697, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12120 + }, + { + "epoch": 4.842588893328006, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12121 + }, + { + "epoch": 4.842988413903316, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12122 + }, + { + "epoch": 4.843387934478626, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 12123 + }, + { + "epoch": 4.843787455053935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12124 + }, + { + "epoch": 4.844186975629245, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12125 + }, + { + "epoch": 4.844586496204554, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12126 + }, + { + "epoch": 4.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12127 + }, + { + "epoch": 4.845385537355174, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12128 + }, + { + "epoch": 4.845785057930484, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12129 + }, + { + "epoch": 4.8461845785057935, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 12130 + }, + { + "epoch": 4.846584099081102, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12131 + }, + { + "epoch": 4.846983619656412, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12132 + }, + { + "epoch": 4.847383140231722, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12133 + }, + { + "epoch": 4.847782660807032, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12134 + }, + { + "epoch": 4.848182181382342, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12135 + }, + { + "epoch": 4.8485817019576505, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12136 + }, + { + "epoch": 4.84898122253296, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 12137 + }, + { + "epoch": 4.84938074310827, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12138 + }, + { + "epoch": 4.84978026368358, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12139 + }, + { + "epoch": 4.85017978425889, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12140 + }, + { + "epoch": 4.8505793048341985, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12141 + }, + { + "epoch": 4.850978825409508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12142 + }, + { + "epoch": 4.851378345984818, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12143 + }, + { + "epoch": 4.851777866560128, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12144 + }, + { + "epoch": 4.852177387135438, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12145 + }, + { + "epoch": 4.8525769077107475, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12146 + }, + { + "epoch": 4.852976428286056, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12147 + }, + { + "epoch": 4.853375948861366, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12148 + }, + { + "epoch": 4.853775469436676, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12149 + }, + { + "epoch": 4.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12150 + }, + { + "epoch": 4.8545745105872955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12151 + }, + { + "epoch": 4.854974031162605, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 12152 + }, + { + "epoch": 4.855373551737914, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12153 + }, + { + "epoch": 4.855773072313224, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12154 + }, + { + "epoch": 4.856172592888534, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12155 + }, + { + "epoch": 4.8565721134638435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12156 + }, + { + "epoch": 4.856971634039153, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12157 + }, + { + "epoch": 4.857371154614462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12158 + }, + { + "epoch": 4.857770675189772, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12159 + }, + { + "epoch": 4.858170195765082, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12160 + }, + { + "epoch": 4.858569716340392, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12161 + }, + { + "epoch": 4.858969236915701, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12162 + }, + { + "epoch": 4.859368757491011, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12163 + }, + { + "epoch": 4.85976827806632, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12164 + }, + { + "epoch": 4.86016779864163, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12165 + }, + { + "epoch": 4.86056731921694, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12166 + }, + { + "epoch": 4.860966839792249, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 12167 + }, + { + "epoch": 4.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12168 + }, + { + "epoch": 4.861765880942869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12169 + }, + { + "epoch": 4.862165401518178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12170 + }, + { + "epoch": 4.862564922093488, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12171 + }, + { + "epoch": 4.8629644426687975, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12172 + }, + { + "epoch": 4.863363963244107, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12173 + }, + { + "epoch": 4.863763483819417, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12174 + }, + { + "epoch": 4.864163004394726, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12175 + }, + { + "epoch": 4.864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12176 + }, + { + "epoch": 4.8649620455453455, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 12177 + }, + { + "epoch": 4.865361566120655, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12178 + }, + { + "epoch": 4.865761086695965, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12179 + }, + { + "epoch": 4.866160607271274, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12180 + }, + { + "epoch": 4.866560127846584, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12181 + }, + { + "epoch": 4.8669596484218935, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12182 + }, + { + "epoch": 4.867359168997203, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12183 + }, + { + "epoch": 4.867758689572513, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12184 + }, + { + "epoch": 4.868158210147823, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12185 + }, + { + "epoch": 4.868557730723133, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12186 + }, + { + "epoch": 4.868957251298442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12187 + }, + { + "epoch": 4.869356771873751, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12188 + }, + { + "epoch": 4.869756292449061, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12189 + }, + { + "epoch": 4.870155813024371, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12190 + }, + { + "epoch": 4.870555333599681, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12191 + }, + { + "epoch": 4.87095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12192 + }, + { + "epoch": 4.871354374750299, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12193 + }, + { + "epoch": 4.871753895325609, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12194 + }, + { + "epoch": 4.872153415900919, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12195 + }, + { + "epoch": 4.872552936476229, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 12196 + }, + { + "epoch": 4.872952457051538, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12197 + }, + { + "epoch": 4.8733519776268475, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12198 + }, + { + "epoch": 4.873751498202157, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 12199 + }, + { + "epoch": 4.874151018777467, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12200 + }, + { + "epoch": 4.874550539352777, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 12201 + }, + { + "epoch": 4.874950059928087, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12202 + }, + { + "epoch": 4.875349580503396, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12203 + }, + { + "epoch": 4.875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12204 + }, + { + "epoch": 4.876148621654015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12205 + }, + { + "epoch": 4.876548142229325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12206 + }, + { + "epoch": 4.876947662804635, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 12207 + }, + { + "epoch": 4.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12208 + }, + { + "epoch": 4.877746703955253, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12209 + }, + { + "epoch": 4.878146224530563, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12210 + }, + { + "epoch": 4.878545745105873, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12211 + }, + { + "epoch": 4.878945265681183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12212 + }, + { + "epoch": 4.8793447862564925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12213 + }, + { + "epoch": 4.879744306831801, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12214 + }, + { + "epoch": 4.880143827407111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12215 + }, + { + "epoch": 4.880543347982421, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12216 + }, + { + "epoch": 4.880942868557731, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12217 + }, + { + "epoch": 4.8813423891330405, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 12218 + }, + { + "epoch": 4.88174190970835, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12219 + }, + { + "epoch": 4.88214143028366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12220 + }, + { + "epoch": 4.882540950858969, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12221 + }, + { + "epoch": 4.882940471434279, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12222 + }, + { + "epoch": 4.883339992009589, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12223 + }, + { + "epoch": 4.883739512584898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12224 + }, + { + "epoch": 4.884139033160208, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12225 + }, + { + "epoch": 4.884538553735517, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12226 + }, + { + "epoch": 4.884938074310827, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12227 + }, + { + "epoch": 4.885337594886137, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12228 + }, + { + "epoch": 4.885737115461446, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12229 + }, + { + "epoch": 4.886136636036756, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12230 + }, + { + "epoch": 4.886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12231 + }, + { + "epoch": 4.886935677187375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12232 + }, + { + "epoch": 4.887335197762685, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12233 + }, + { + "epoch": 4.887734718337994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12234 + }, + { + "epoch": 4.888134238913304, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12235 + }, + { + "epoch": 4.888533759488614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12236 + }, + { + "epoch": 4.888933280063923, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12237 + }, + { + "epoch": 4.889332800639233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12238 + }, + { + "epoch": 4.8897323212145425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12239 + }, + { + "epoch": 4.890131841789852, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12240 + }, + { + "epoch": 4.890531362365162, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12241 + }, + { + "epoch": 4.890930882940472, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12242 + }, + { + "epoch": 4.891330403515781, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12243 + }, + { + "epoch": 4.8917299240910905, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12244 + }, + { + "epoch": 4.8921294446664, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 12245 + }, + { + "epoch": 4.89252896524171, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12246 + }, + { + "epoch": 4.89292848581702, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12247 + }, + { + "epoch": 4.893328006392329, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12248 + }, + { + "epoch": 4.893727526967639, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12249 + }, + { + "epoch": 4.894127047542948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12250 + }, + { + "epoch": 4.894526568118258, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12251 + }, + { + "epoch": 4.894926088693568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12252 + }, + { + "epoch": 4.895325609268878, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12253 + }, + { + "epoch": 4.895725129844187, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12254 + }, + { + "epoch": 4.896124650419496, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12255 + }, + { + "epoch": 4.896524170994806, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12256 + }, + { + "epoch": 4.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12257 + }, + { + "epoch": 4.897323212145426, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12258 + }, + { + "epoch": 4.8977227327207356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12259 + }, + { + "epoch": 4.8981222532960444, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12260 + }, + { + "epoch": 4.898521773871354, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12261 + }, + { + "epoch": 4.898921294446664, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12262 + }, + { + "epoch": 4.899320815021974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12263 + }, + { + "epoch": 4.899720335597284, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12264 + }, + { + "epoch": 4.9001198561725925, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12265 + }, + { + "epoch": 4.900519376747902, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12266 + }, + { + "epoch": 4.900918897323212, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12267 + }, + { + "epoch": 4.901318417898522, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12268 + }, + { + "epoch": 4.901717938473832, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12269 + }, + { + "epoch": 4.9021174590491405, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12270 + }, + { + "epoch": 4.90251697962445, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 12271 + }, + { + "epoch": 4.90291650019976, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12272 + }, + { + "epoch": 4.90331602077507, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12273 + }, + { + "epoch": 4.90371554135038, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12274 + }, + { + "epoch": 4.9041150619256895, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 12275 + }, + { + "epoch": 4.904514582500999, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12276 + }, + { + "epoch": 4.904914103076308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 12277 + }, + { + "epoch": 4.905313623651618, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12278 + }, + { + "epoch": 4.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12279 + }, + { + "epoch": 4.9061126648022375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12280 + }, + { + "epoch": 4.906512185377547, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12281 + }, + { + "epoch": 4.906911705952856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12282 + }, + { + "epoch": 4.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 12283 + }, + { + "epoch": 4.907710747103476, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12284 + }, + { + "epoch": 4.9081102676787856, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12285 + }, + { + "epoch": 4.908509788254095, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12286 + }, + { + "epoch": 4.908909308829404, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12287 + }, + { + "epoch": 4.909308829404714, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12288 + }, + { + "epoch": 4.909708349980024, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12289 + }, + { + "epoch": 4.910107870555334, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12290 + }, + { + "epoch": 4.910507391130643, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12291 + }, + { + "epoch": 4.910906911705953, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12292 + }, + { + "epoch": 4.911306432281263, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12293 + }, + { + "epoch": 4.911705952856572, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 12294 + }, + { + "epoch": 4.912105473431882, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12295 + }, + { + "epoch": 4.912504994007191, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12296 + }, + { + "epoch": 4.912904514582501, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12297 + }, + { + "epoch": 4.913304035157811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12298 + }, + { + "epoch": 4.91370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12299 + }, + { + "epoch": 4.91410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12300 + }, + { + "epoch": 4.9145025968837395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12301 + }, + { + "epoch": 4.914902117459049, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12302 + }, + { + "epoch": 4.915301638034359, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12303 + }, + { + "epoch": 4.915701158609668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12304 + }, + { + "epoch": 4.916100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12305 + }, + { + "epoch": 4.9165001997602875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12306 + }, + { + "epoch": 4.916899720335597, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12307 + }, + { + "epoch": 4.917299240910907, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12308 + }, + { + "epoch": 4.917698761486217, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12309 + }, + { + "epoch": 4.918098282061527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12310 + }, + { + "epoch": 4.918497802636836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12311 + }, + { + "epoch": 4.918897323212145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12312 + }, + { + "epoch": 4.919296843787455, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12313 + }, + { + "epoch": 4.919696364362765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12314 + }, + { + "epoch": 4.920095884938075, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12315 + }, + { + "epoch": 4.920495405513384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12316 + }, + { + "epoch": 4.920894926088693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12317 + }, + { + "epoch": 4.921294446664003, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12318 + }, + { + "epoch": 4.921693967239313, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12319 + }, + { + "epoch": 4.922093487814623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12320 + }, + { + "epoch": 4.922493008389932, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12321 + }, + { + "epoch": 4.922892528965241, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12322 + }, + { + "epoch": 4.923292049540551, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12323 + }, + { + "epoch": 4.923691570115861, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12324 + }, + { + "epoch": 4.924091090691171, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12325 + }, + { + "epoch": 4.924490611266481, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12326 + }, + { + "epoch": 4.9248901318417895, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12327 + }, + { + "epoch": 4.925289652417099, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12328 + }, + { + "epoch": 4.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12329 + }, + { + "epoch": 4.926088693567719, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12330 + }, + { + "epoch": 4.926488214143029, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2973, + "step": 12331 + }, + { + "epoch": 4.926887734718338, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12332 + }, + { + "epoch": 4.927287255293647, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 12333 + }, + { + "epoch": 4.927686775868957, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12334 + }, + { + "epoch": 4.928086296444267, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12335 + }, + { + "epoch": 4.928485817019577, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12336 + }, + { + "epoch": 4.9288853375948865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12337 + }, + { + "epoch": 4.929284858170195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 12338 + }, + { + "epoch": 4.929684378745505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12339 + }, + { + "epoch": 4.930083899320815, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12340 + }, + { + "epoch": 4.930483419896125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12341 + }, + { + "epoch": 4.9308829404714345, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12342 + }, + { + "epoch": 4.931282461046744, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12343 + }, + { + "epoch": 4.931681981622053, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12344 + }, + { + "epoch": 4.932081502197363, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12345 + }, + { + "epoch": 4.932481022772673, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12346 + }, + { + "epoch": 4.9328805433479825, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12347 + }, + { + "epoch": 4.933280063923292, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12348 + }, + { + "epoch": 4.933679584498602, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12349 + }, + { + "epoch": 4.934079105073911, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12350 + }, + { + "epoch": 4.934478625649221, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12351 + }, + { + "epoch": 4.934878146224531, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12352 + }, + { + "epoch": 4.93527766679984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12353 + }, + { + "epoch": 4.93567718737515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12354 + }, + { + "epoch": 4.936076707950459, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12355 + }, + { + "epoch": 4.936476228525769, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12356 + }, + { + "epoch": 4.936875749101079, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12357 + }, + { + "epoch": 4.937275269676388, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12358 + }, + { + "epoch": 4.937674790251698, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12359 + }, + { + "epoch": 4.938074310827007, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12360 + }, + { + "epoch": 4.938473831402317, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12361 + }, + { + "epoch": 4.938873351977627, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12362 + }, + { + "epoch": 4.9392728725529365, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12363 + }, + { + "epoch": 4.939672393128246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12364 + }, + { + "epoch": 4.940071913703556, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12365 + }, + { + "epoch": 4.940471434278866, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12366 + }, + { + "epoch": 4.940870954854175, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12367 + }, + { + "epoch": 4.9412704754294845, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12368 + }, + { + "epoch": 4.941669996004794, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12369 + }, + { + "epoch": 4.942069516580104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12370 + }, + { + "epoch": 4.942469037155414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12371 + }, + { + "epoch": 4.942868557730723, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12372 + }, + { + "epoch": 4.9432680783060325, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12373 + }, + { + "epoch": 4.943667598881342, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12374 + }, + { + "epoch": 4.944067119456652, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 12375 + }, + { + "epoch": 4.944466640031962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12376 + }, + { + "epoch": 4.944866160607271, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12377 + }, + { + "epoch": 4.945265681182581, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12378 + }, + { + "epoch": 4.94566520175789, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12379 + }, + { + "epoch": 4.9460647223332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12380 + }, + { + "epoch": 4.94646424290851, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12381 + }, + { + "epoch": 4.94686376348382, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12382 + }, + { + "epoch": 4.9472632840591295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12383 + }, + { + "epoch": 4.947662804634438, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12384 + }, + { + "epoch": 4.948062325209748, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12385 + }, + { + "epoch": 4.948461845785058, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12386 + }, + { + "epoch": 4.948861366360368, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12387 + }, + { + "epoch": 4.949260886935678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12388 + }, + { + "epoch": 4.9496604075109865, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12389 + }, + { + "epoch": 4.950059928086296, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12390 + }, + { + "epoch": 4.950459448661606, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12391 + }, + { + "epoch": 4.950858969236916, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12392 + }, + { + "epoch": 4.951258489812226, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 12393 + }, + { + "epoch": 4.9516580103875345, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12394 + }, + { + "epoch": 4.952057530962844, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 12395 + }, + { + "epoch": 4.952457051538154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12396 + }, + { + "epoch": 4.952856572113464, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12397 + }, + { + "epoch": 4.953256092688774, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12398 + }, + { + "epoch": 4.953655613264083, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12399 + }, + { + "epoch": 4.954055133839393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12400 + }, + { + "epoch": 4.954454654414702, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12401 + }, + { + "epoch": 4.954854174990012, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12402 + }, + { + "epoch": 4.955253695565322, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12403 + }, + { + "epoch": 4.9556532161406315, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12404 + }, + { + "epoch": 4.956052736715941, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12405 + }, + { + "epoch": 4.95645225729125, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12406 + }, + { + "epoch": 4.95685177786656, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12407 + }, + { + "epoch": 4.95725129844187, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12408 + }, + { + "epoch": 4.9576508190171795, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 12409 + }, + { + "epoch": 4.958050339592489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12410 + }, + { + "epoch": 4.958449860167798, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12411 + }, + { + "epoch": 4.958849380743108, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 12412 + }, + { + "epoch": 4.959248901318418, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12413 + }, + { + "epoch": 4.959648421893728, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12414 + }, + { + "epoch": 4.960047942469037, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12415 + }, + { + "epoch": 4.960447463044347, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12416 + }, + { + "epoch": 4.960846983619656, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12417 + }, + { + "epoch": 4.961246504194966, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12418 + }, + { + "epoch": 4.961646024770276, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12419 + }, + { + "epoch": 4.962045545345585, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12420 + }, + { + "epoch": 4.962445065920895, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12421 + }, + { + "epoch": 4.962844586496205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12422 + }, + { + "epoch": 4.963244107071514, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12423 + }, + { + "epoch": 4.963643627646824, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12424 + }, + { + "epoch": 4.964043148222133, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12425 + }, + { + "epoch": 4.964442668797443, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12426 + }, + { + "epoch": 4.964842189372753, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12427 + }, + { + "epoch": 4.965241709948062, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12428 + }, + { + "epoch": 4.965641230523372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12429 + }, + { + "epoch": 4.9660407510986815, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 12430 + }, + { + "epoch": 4.966440271673991, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12431 + }, + { + "epoch": 4.966839792249301, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12432 + }, + { + "epoch": 4.967239312824611, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12433 + }, + { + "epoch": 4.96763883339992, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12434 + }, + { + "epoch": 4.9680383539752295, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12435 + }, + { + "epoch": 4.968437874550539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12436 + }, + { + "epoch": 4.968837395125849, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12437 + }, + { + "epoch": 4.969236915701159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12438 + }, + { + "epoch": 4.969636436276469, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12439 + }, + { + "epoch": 4.970035956851778, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12440 + }, + { + "epoch": 4.970435477427087, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12441 + }, + { + "epoch": 4.970834998002397, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12442 + }, + { + "epoch": 4.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12443 + }, + { + "epoch": 4.971634039153017, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12444 + }, + { + "epoch": 4.972033559728326, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12445 + }, + { + "epoch": 4.972433080303635, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12446 + }, + { + "epoch": 4.972832600878945, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12447 + }, + { + "epoch": 4.973232121454255, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12448 + }, + { + "epoch": 4.973631642029565, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12449 + }, + { + "epoch": 4.974031162604874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12450 + }, + { + "epoch": 4.974430683180183, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12451 + }, + { + "epoch": 4.974830203755493, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12452 + }, + { + "epoch": 4.975229724330803, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12453 + }, + { + "epoch": 4.975629244906113, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12454 + }, + { + "epoch": 4.976028765481423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12455 + }, + { + "epoch": 4.976428286056732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12456 + }, + { + "epoch": 4.976827806632041, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12457 + }, + { + "epoch": 4.977227327207351, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12458 + }, + { + "epoch": 4.977626847782661, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12459 + }, + { + "epoch": 4.978026368357971, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 12460 + }, + { + "epoch": 4.97842588893328, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12461 + }, + { + "epoch": 4.978825409508589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 12462 + }, + { + "epoch": 4.979224930083899, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12463 + }, + { + "epoch": 4.979624450659209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 12464 + }, + { + "epoch": 4.980023971234519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 12465 + }, + { + "epoch": 4.9804234918098285, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12466 + }, + { + "epoch": 4.980823012385137, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12467 + }, + { + "epoch": 4.981222532960447, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12468 + }, + { + "epoch": 4.981622053535757, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12469 + }, + { + "epoch": 4.982021574111067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12470 + }, + { + "epoch": 4.9824210946863765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12471 + }, + { + "epoch": 4.982820615261686, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12472 + }, + { + "epoch": 4.983220135836996, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12473 + }, + { + "epoch": 4.983619656412305, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12474 + }, + { + "epoch": 4.984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12475 + }, + { + "epoch": 4.9844186975629245, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12476 + }, + { + "epoch": 4.984818218138234, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12477 + }, + { + "epoch": 4.985217738713544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12478 + }, + { + "epoch": 4.985617259288853, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12479 + }, + { + "epoch": 4.986016779864163, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12480 + }, + { + "epoch": 4.986416300439473, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12481 + }, + { + "epoch": 4.986815821014782, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12482 + }, + { + "epoch": 4.987215341590092, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12483 + }, + { + "epoch": 4.987614862165401, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12484 + }, + { + "epoch": 4.988014382740711, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12485 + }, + { + "epoch": 4.988413903316021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12486 + }, + { + "epoch": 4.98881342389133, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12487 + }, + { + "epoch": 4.98921294446664, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12488 + }, + { + "epoch": 4.98961246504195, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12489 + }, + { + "epoch": 4.99001198561726, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12490 + }, + { + "epoch": 4.990411506192569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12491 + }, + { + "epoch": 4.9908110267678785, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12492 + }, + { + "epoch": 4.991210547343188, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12493 + }, + { + "epoch": 4.991610067918498, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12494 + }, + { + "epoch": 4.992009588493808, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12495 + }, + { + "epoch": 4.992409109069117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12496 + }, + { + "epoch": 4.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12497 + }, + { + "epoch": 4.993208150219736, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12498 + }, + { + "epoch": 4.993607670795046, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12499 + }, + { + "epoch": 4.994007191370356, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 12500 + }, + { + "epoch": 4.994406711945665, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12501 + }, + { + "epoch": 4.9948062325209746, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12502 + }, + { + "epoch": 4.995205753096284, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12503 + }, + { + "epoch": 4.995605273671594, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12504 + }, + { + "epoch": 4.996004794246904, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12505 + }, + { + "epoch": 4.996404314822214, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12506 + }, + { + "epoch": 4.996803835397523, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12507 + }, + { + "epoch": 4.997203355972832, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12508 + }, + { + "epoch": 4.997602876548142, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12509 + }, + { + "epoch": 4.998002397123452, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12510 + }, + { + "epoch": 4.998401917698762, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 12511 + }, + { + "epoch": 4.9988014382740715, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12512 + }, + { + "epoch": 4.99920095884938, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12513 + }, + { + "epoch": 4.99960047942469, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12514 + }, + { + "epoch": 5.0, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12515 + }, + { + "epoch": 5.00039952057531, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12516 + }, + { + "epoch": 5.00079904115062, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12517 + }, + { + "epoch": 5.0011985617259285, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12518 + }, + { + "epoch": 5.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12519 + }, + { + "epoch": 5.001997602876548, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12520 + }, + { + "epoch": 5.002397123451858, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12521 + }, + { + "epoch": 5.002796644027168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12522 + }, + { + "epoch": 5.003196164602477, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12523 + }, + { + "epoch": 5.003595685177786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12524 + }, + { + "epoch": 5.003995205753096, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12525 + }, + { + "epoch": 5.004394726328406, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12526 + }, + { + "epoch": 5.004794246903716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12527 + }, + { + "epoch": 5.0051937674790254, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12528 + }, + { + "epoch": 5.005593288054335, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12529 + }, + { + "epoch": 5.005992808629644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12530 + }, + { + "epoch": 5.006392329204954, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12531 + }, + { + "epoch": 5.006791849780264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12532 + }, + { + "epoch": 5.0071913703555735, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12533 + }, + { + "epoch": 5.007590890930883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12534 + }, + { + "epoch": 5.007990411506192, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12535 + }, + { + "epoch": 5.008389932081502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12536 + }, + { + "epoch": 5.008789452656812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12537 + }, + { + "epoch": 5.0091889732321215, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12538 + }, + { + "epoch": 5.009588493807431, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12539 + }, + { + "epoch": 5.009988014382741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12540 + }, + { + "epoch": 5.01038753495805, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12541 + }, + { + "epoch": 5.01078705553336, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12542 + }, + { + "epoch": 5.01118657610867, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12543 + }, + { + "epoch": 5.011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12544 + }, + { + "epoch": 5.011985617259289, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12545 + }, + { + "epoch": 5.012385137834598, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12546 + }, + { + "epoch": 5.012784658409908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12547 + }, + { + "epoch": 5.013184178985218, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12548 + }, + { + "epoch": 5.013583699560527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12549 + }, + { + "epoch": 5.013983220135837, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 12550 + }, + { + "epoch": 5.014382740711147, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12551 + }, + { + "epoch": 5.014782261286456, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12552 + }, + { + "epoch": 5.015181781861766, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 12553 + }, + { + "epoch": 5.0155813024370755, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12554 + }, + { + "epoch": 5.015980823012385, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12555 + }, + { + "epoch": 5.016380343587695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12556 + }, + { + "epoch": 5.016779864163005, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12557 + }, + { + "epoch": 5.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12558 + }, + { + "epoch": 5.0175789053136235, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12559 + }, + { + "epoch": 5.017978425888933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12560 + }, + { + "epoch": 5.018377946464243, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12561 + }, + { + "epoch": 5.018777467039553, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12562 + }, + { + "epoch": 5.019176987614862, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12563 + }, + { + "epoch": 5.0195765081901715, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12564 + }, + { + "epoch": 5.019976028765481, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12565 + }, + { + "epoch": 5.020375549340791, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12566 + }, + { + "epoch": 5.020775069916101, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12567 + }, + { + "epoch": 5.021174590491411, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12568 + }, + { + "epoch": 5.02157411106672, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12569 + }, + { + "epoch": 5.021973631642029, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 12570 + }, + { + "epoch": 5.022373152217339, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12571 + }, + { + "epoch": 5.022772672792649, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12572 + }, + { + "epoch": 5.023172193367959, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12573 + }, + { + "epoch": 5.0235717139432685, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12574 + }, + { + "epoch": 5.023971234518577, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12575 + }, + { + "epoch": 5.024370755093887, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12576 + }, + { + "epoch": 5.024770275669197, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12577 + }, + { + "epoch": 5.025169796244507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12578 + }, + { + "epoch": 5.025569316819817, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12579 + }, + { + "epoch": 5.0259688373951255, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12580 + }, + { + "epoch": 5.026368357970435, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12581 + }, + { + "epoch": 5.026767878545745, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12582 + }, + { + "epoch": 5.027167399121055, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12583 + }, + { + "epoch": 5.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12584 + }, + { + "epoch": 5.027966440271674, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12585 + }, + { + "epoch": 5.028365960846983, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12586 + }, + { + "epoch": 5.028765481422293, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12587 + }, + { + "epoch": 5.029165001997603, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12588 + }, + { + "epoch": 5.029564522572913, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12589 + }, + { + "epoch": 5.029964043148222, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12590 + }, + { + "epoch": 5.030363563723531, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12591 + }, + { + "epoch": 5.030763084298841, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12592 + }, + { + "epoch": 5.031162604874151, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12593 + }, + { + "epoch": 5.031562125449461, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12594 + }, + { + "epoch": 5.0319616460247705, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12595 + }, + { + "epoch": 5.03236116660008, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12596 + }, + { + "epoch": 5.032760687175389, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12597 + }, + { + "epoch": 5.033160207750699, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 12598 + }, + { + "epoch": 5.033559728326009, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12599 + }, + { + "epoch": 5.0339592489013185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12600 + }, + { + "epoch": 5.034358769476628, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12601 + }, + { + "epoch": 5.034758290051938, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12602 + }, + { + "epoch": 5.035157810627247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12603 + }, + { + "epoch": 5.035557331202557, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 12604 + }, + { + "epoch": 5.035956851777867, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12605 + }, + { + "epoch": 5.036356372353176, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12606 + }, + { + "epoch": 5.036755892928486, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12607 + }, + { + "epoch": 5.037155413503795, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12608 + }, + { + "epoch": 5.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12609 + }, + { + "epoch": 5.037954454654415, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12610 + }, + { + "epoch": 5.038353975229724, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12611 + }, + { + "epoch": 5.038753495805034, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12612 + }, + { + "epoch": 5.039153016380344, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12613 + }, + { + "epoch": 5.039552536955653, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12614 + }, + { + "epoch": 5.039952057530963, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12615 + }, + { + "epoch": 5.040351578106272, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 12616 + }, + { + "epoch": 5.040751098681582, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12617 + }, + { + "epoch": 5.041150619256892, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 12618 + }, + { + "epoch": 5.041550139832202, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12619 + }, + { + "epoch": 5.041949660407511, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12620 + }, + { + "epoch": 5.0423491809828205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12621 + }, + { + "epoch": 5.04274870155813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12622 + }, + { + "epoch": 5.04314822213344, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 12623 + }, + { + "epoch": 5.04354774270875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12624 + }, + { + "epoch": 5.043947263284059, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12625 + }, + { + "epoch": 5.0443467838593685, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12626 + }, + { + "epoch": 5.044746304434678, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12627 + }, + { + "epoch": 5.045145825009988, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12628 + }, + { + "epoch": 5.045545345585298, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12629 + }, + { + "epoch": 5.045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12630 + }, + { + "epoch": 5.046344386735917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12631 + }, + { + "epoch": 5.046743907311226, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12632 + }, + { + "epoch": 5.047143427886536, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12633 + }, + { + "epoch": 5.047542948461846, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12634 + }, + { + "epoch": 5.047942469037156, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12635 + }, + { + "epoch": 5.048341989612465, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12636 + }, + { + "epoch": 5.048741510187774, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12637 + }, + { + "epoch": 5.049141030763084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12638 + }, + { + "epoch": 5.049540551338394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12639 + }, + { + "epoch": 5.049940071913704, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12640 + }, + { + "epoch": 5.0503395924890135, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12641 + }, + { + "epoch": 5.050739113064322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12642 + }, + { + "epoch": 5.051138633639632, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12643 + }, + { + "epoch": 5.051538154214942, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12644 + }, + { + "epoch": 5.051937674790252, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 12645 + }, + { + "epoch": 5.052337195365562, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12646 + }, + { + "epoch": 5.052736715940871, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12647 + }, + { + "epoch": 5.05313623651618, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12648 + }, + { + "epoch": 5.05353575709149, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12649 + }, + { + "epoch": 5.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12650 + }, + { + "epoch": 5.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12651 + }, + { + "epoch": 5.054734318817419, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12652 + }, + { + "epoch": 5.055133839392728, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12653 + }, + { + "epoch": 5.055533359968038, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12654 + }, + { + "epoch": 5.055932880543348, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12655 + }, + { + "epoch": 5.056332401118658, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12656 + }, + { + "epoch": 5.0567319216939675, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12657 + }, + { + "epoch": 5.057131442269277, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12658 + }, + { + "epoch": 5.057530962844586, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12659 + }, + { + "epoch": 5.057930483419896, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12660 + }, + { + "epoch": 5.058330003995206, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12661 + }, + { + "epoch": 5.0587295245705155, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12662 + }, + { + "epoch": 5.059129045145825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12663 + }, + { + "epoch": 5.059528565721135, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12664 + }, + { + "epoch": 5.059928086296444, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12665 + }, + { + "epoch": 5.060327606871754, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12666 + }, + { + "epoch": 5.0607271274470635, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12667 + }, + { + "epoch": 5.061126648022373, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12668 + }, + { + "epoch": 5.061526168597683, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12669 + }, + { + "epoch": 5.061925689172992, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 12670 + }, + { + "epoch": 5.062325209748302, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12671 + }, + { + "epoch": 5.062724730323612, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12672 + }, + { + "epoch": 5.063124250898921, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12673 + }, + { + "epoch": 5.063523771474231, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12674 + }, + { + "epoch": 5.063923292049541, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12675 + }, + { + "epoch": 5.06432281262485, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12676 + }, + { + "epoch": 5.06472233320016, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12677 + }, + { + "epoch": 5.065121853775469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12678 + }, + { + "epoch": 5.065521374350779, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12679 + }, + { + "epoch": 5.065920894926089, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12680 + }, + { + "epoch": 5.066320415501398, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12681 + }, + { + "epoch": 5.066719936076708, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12682 + }, + { + "epoch": 5.0671194566520175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12683 + }, + { + "epoch": 5.067518977227327, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12684 + }, + { + "epoch": 5.067918497802637, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12685 + }, + { + "epoch": 5.068318018377947, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12686 + }, + { + "epoch": 5.068717538953256, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12687 + }, + { + "epoch": 5.0691170595285655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12688 + }, + { + "epoch": 5.069516580103875, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12689 + }, + { + "epoch": 5.069916100679185, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12690 + }, + { + "epoch": 5.070315621254495, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12691 + }, + { + "epoch": 5.070715141829805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12692 + }, + { + "epoch": 5.0711146624051135, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12693 + }, + { + "epoch": 5.071514182980423, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12694 + }, + { + "epoch": 5.071913703555733, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12695 + }, + { + "epoch": 5.072313224131043, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 12696 + }, + { + "epoch": 5.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12697 + }, + { + "epoch": 5.073112265281662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12698 + }, + { + "epoch": 5.073511785856971, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12699 + }, + { + "epoch": 5.073911306432281, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12700 + }, + { + "epoch": 5.074310827007591, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12701 + }, + { + "epoch": 5.074710347582901, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 12702 + }, + { + "epoch": 5.0751098681582105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12703 + }, + { + "epoch": 5.075509388733519, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12704 + }, + { + "epoch": 5.075908909308829, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12705 + }, + { + "epoch": 5.076308429884139, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12706 + }, + { + "epoch": 5.076707950459449, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12707 + }, + { + "epoch": 5.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12708 + }, + { + "epoch": 5.077506991610068, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12709 + }, + { + "epoch": 5.077906512185377, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 12710 + }, + { + "epoch": 5.078306032760687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12711 + }, + { + "epoch": 5.078705553335997, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12712 + }, + { + "epoch": 5.079105073911307, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12713 + }, + { + "epoch": 5.079504594486616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12714 + }, + { + "epoch": 5.079904115061925, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12715 + }, + { + "epoch": 5.080303635637235, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12716 + }, + { + "epoch": 5.080703156212545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12717 + }, + { + "epoch": 5.081102676787855, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12718 + }, + { + "epoch": 5.081502197363164, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12719 + }, + { + "epoch": 5.081901717938474, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12720 + }, + { + "epoch": 5.082301238513783, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12721 + }, + { + "epoch": 5.082700759089093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12722 + }, + { + "epoch": 5.083100279664403, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12723 + }, + { + "epoch": 5.0834998002397125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12724 + }, + { + "epoch": 5.083899320815022, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12725 + }, + { + "epoch": 5.084298841390331, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 12726 + }, + { + "epoch": 5.084698361965641, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12727 + }, + { + "epoch": 5.085097882540951, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12728 + }, + { + "epoch": 5.0854974031162605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12729 + }, + { + "epoch": 5.08589692369157, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12730 + }, + { + "epoch": 5.08629644426688, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12731 + }, + { + "epoch": 5.086695964842189, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12732 + }, + { + "epoch": 5.087095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12733 + }, + { + "epoch": 5.087495005992809, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 12734 + }, + { + "epoch": 5.087894526568118, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12735 + }, + { + "epoch": 5.088294047143428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12736 + }, + { + "epoch": 5.088693567718738, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12737 + }, + { + "epoch": 5.089093088294047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 12738 + }, + { + "epoch": 5.089492608869357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12739 + }, + { + "epoch": 5.089892129444666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12740 + }, + { + "epoch": 5.090291650019976, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12741 + }, + { + "epoch": 5.090691170595286, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12742 + }, + { + "epoch": 5.091090691170595, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12743 + }, + { + "epoch": 5.091490211745905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12744 + }, + { + "epoch": 5.0918897323212144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12745 + }, + { + "epoch": 5.092289252896524, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12746 + }, + { + "epoch": 5.092688773471834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12747 + }, + { + "epoch": 5.093088294047144, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 12748 + }, + { + "epoch": 5.093487814622453, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 12749 + }, + { + "epoch": 5.0938873351977625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12750 + }, + { + "epoch": 5.094286855773072, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12751 + }, + { + "epoch": 5.094686376348382, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12752 + }, + { + "epoch": 5.095085896923692, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12753 + }, + { + "epoch": 5.095485417499002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12754 + }, + { + "epoch": 5.0958849380743105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12755 + }, + { + "epoch": 5.09628445864962, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12756 + }, + { + "epoch": 5.09668397922493, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12757 + }, + { + "epoch": 5.09708349980024, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12758 + }, + { + "epoch": 5.09748302037555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12759 + }, + { + "epoch": 5.097882540950859, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12760 + }, + { + "epoch": 5.098282061526168, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 12761 + }, + { + "epoch": 5.098681582101478, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12762 + }, + { + "epoch": 5.099081102676788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12763 + }, + { + "epoch": 5.099480623252098, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12764 + }, + { + "epoch": 5.0998801438274075, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12765 + }, + { + "epoch": 5.100279664402716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 12766 + }, + { + "epoch": 5.100679184978026, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12767 + }, + { + "epoch": 5.101078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 12768 + }, + { + "epoch": 5.101478226128646, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12769 + }, + { + "epoch": 5.1018777467039556, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12770 + }, + { + "epoch": 5.1022772672792644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12771 + }, + { + "epoch": 5.102676787854574, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12772 + }, + { + "epoch": 5.103076308429884, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12773 + }, + { + "epoch": 5.103475829005194, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 12774 + }, + { + "epoch": 5.103875349580504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12775 + }, + { + "epoch": 5.104274870155813, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12776 + }, + { + "epoch": 5.104674390731122, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12777 + }, + { + "epoch": 5.105073911306432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12778 + }, + { + "epoch": 5.105473431881742, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12779 + }, + { + "epoch": 5.105872952457052, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12780 + }, + { + "epoch": 5.106272473032361, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12781 + }, + { + "epoch": 5.106671993607671, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12782 + }, + { + "epoch": 5.10707151418298, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12783 + }, + { + "epoch": 5.10747103475829, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12784 + }, + { + "epoch": 5.1078705553336, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12785 + }, + { + "epoch": 5.1082700759089095, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 12786 + }, + { + "epoch": 5.108669596484219, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12787 + }, + { + "epoch": 5.109069117059528, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12788 + }, + { + "epoch": 5.109468637634838, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12789 + }, + { + "epoch": 5.109868158210148, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12790 + }, + { + "epoch": 5.1102676787854575, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12791 + }, + { + "epoch": 5.110667199360767, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12792 + }, + { + "epoch": 5.111066719936077, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12793 + }, + { + "epoch": 5.111466240511386, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12794 + }, + { + "epoch": 5.111865761086696, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12795 + }, + { + "epoch": 5.112265281662006, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12796 + }, + { + "epoch": 5.112664802237315, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12797 + }, + { + "epoch": 5.113064322812625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12798 + }, + { + "epoch": 5.113463843387935, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12799 + }, + { + "epoch": 5.113863363963244, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12800 + }, + { + "epoch": 5.114262884538554, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12801 + }, + { + "epoch": 5.114662405113863, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12802 + }, + { + "epoch": 5.115061925689173, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12803 + }, + { + "epoch": 5.115461446264483, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12804 + }, + { + "epoch": 5.115860966839792, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12805 + }, + { + "epoch": 5.116260487415102, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12806 + }, + { + "epoch": 5.116660007990411, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12807 + }, + { + "epoch": 5.117059528565721, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12808 + }, + { + "epoch": 5.117459049141031, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12809 + }, + { + "epoch": 5.117858569716341, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12810 + }, + { + "epoch": 5.11825809029165, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12811 + }, + { + "epoch": 5.1186576108669595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12812 + }, + { + "epoch": 5.119057131442269, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12813 + }, + { + "epoch": 5.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12814 + }, + { + "epoch": 5.119856172592889, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12815 + }, + { + "epoch": 5.120255693168198, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12816 + }, + { + "epoch": 5.1206552137435075, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12817 + }, + { + "epoch": 5.121054734318817, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12818 + }, + { + "epoch": 5.121454254894127, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12819 + }, + { + "epoch": 5.121853775469437, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12820 + }, + { + "epoch": 5.122253296044747, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3011, + "step": 12821 + }, + { + "epoch": 5.122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12822 + }, + { + "epoch": 5.123052337195365, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12823 + }, + { + "epoch": 5.123451857770675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12824 + }, + { + "epoch": 5.123851378345985, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12825 + }, + { + "epoch": 5.124250898921295, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 12826 + }, + { + "epoch": 5.1246504194966045, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12827 + }, + { + "epoch": 5.125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12828 + }, + { + "epoch": 5.125449460647223, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 12829 + }, + { + "epoch": 5.125848981222533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12830 + }, + { + "epoch": 5.126248501797843, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12831 + }, + { + "epoch": 5.1266480223731525, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12832 + }, + { + "epoch": 5.127047542948461, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12833 + }, + { + "epoch": 5.127447063523771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12834 + }, + { + "epoch": 5.127846584099081, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12835 + }, + { + "epoch": 5.128246104674391, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12836 + }, + { + "epoch": 5.128645625249701, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12837 + }, + { + "epoch": 5.12904514582501, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12838 + }, + { + "epoch": 5.129444666400319, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12839 + }, + { + "epoch": 5.129844186975629, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12840 + }, + { + "epoch": 5.130243707550939, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12841 + }, + { + "epoch": 5.130643228126249, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12842 + }, + { + "epoch": 5.131042748701558, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 12843 + }, + { + "epoch": 5.131442269276867, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12844 + }, + { + "epoch": 5.131841789852177, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12845 + }, + { + "epoch": 5.132241310427487, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12846 + }, + { + "epoch": 5.132640831002797, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12847 + }, + { + "epoch": 5.1330403515781065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12848 + }, + { + "epoch": 5.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12849 + }, + { + "epoch": 5.133839392728725, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12850 + }, + { + "epoch": 5.134238913304035, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 12851 + }, + { + "epoch": 5.134638433879345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12852 + }, + { + "epoch": 5.1350379544546545, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12853 + }, + { + "epoch": 5.135437475029964, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12854 + }, + { + "epoch": 5.135836995605274, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12855 + }, + { + "epoch": 5.136236516180583, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12856 + }, + { + "epoch": 5.136636036755893, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12857 + }, + { + "epoch": 5.1370355573312025, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12858 + }, + { + "epoch": 5.137435077906512, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12859 + }, + { + "epoch": 5.137834598481822, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12860 + }, + { + "epoch": 5.138234119057131, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12861 + }, + { + "epoch": 5.138633639632441, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12862 + }, + { + "epoch": 5.139033160207751, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12863 + }, + { + "epoch": 5.13943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12864 + }, + { + "epoch": 5.13983220135837, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12865 + }, + { + "epoch": 5.14023172193368, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12866 + }, + { + "epoch": 5.140631242508989, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12867 + }, + { + "epoch": 5.141030763084299, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12868 + }, + { + "epoch": 5.141430283659608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12869 + }, + { + "epoch": 5.141829804234918, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12870 + }, + { + "epoch": 5.142229324810228, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12871 + }, + { + "epoch": 5.142628845385538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12872 + }, + { + "epoch": 5.143028365960847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12873 + }, + { + "epoch": 5.1434278865361565, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12874 + }, + { + "epoch": 5.143827407111466, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12875 + }, + { + "epoch": 5.144226927686776, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12876 + }, + { + "epoch": 5.144626448262086, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12877 + }, + { + "epoch": 5.145025968837395, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12878 + }, + { + "epoch": 5.1454254894127045, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12879 + }, + { + "epoch": 5.145825009988014, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12880 + }, + { + "epoch": 5.146224530563324, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12881 + }, + { + "epoch": 5.146624051138634, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12882 + }, + { + "epoch": 5.147023571713944, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12883 + }, + { + "epoch": 5.1474230922892525, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12884 + }, + { + "epoch": 5.147822612864562, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12885 + }, + { + "epoch": 5.148222133439872, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12886 + }, + { + "epoch": 5.148621654015182, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12887 + }, + { + "epoch": 5.149021174590492, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12888 + }, + { + "epoch": 5.1494206951658015, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12889 + }, + { + "epoch": 5.14982021574111, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 12890 + }, + { + "epoch": 5.15021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12891 + }, + { + "epoch": 5.15061925689173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12892 + }, + { + "epoch": 5.15101877746704, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12893 + }, + { + "epoch": 5.1514182980423495, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12894 + }, + { + "epoch": 5.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12895 + }, + { + "epoch": 5.152217339192968, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12896 + }, + { + "epoch": 5.152616859768278, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12897 + }, + { + "epoch": 5.153016380343588, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12898 + }, + { + "epoch": 5.153415900918898, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12899 + }, + { + "epoch": 5.153815421494207, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12900 + }, + { + "epoch": 5.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12901 + }, + { + "epoch": 5.154614462644826, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12902 + }, + { + "epoch": 5.155013983220136, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12903 + }, + { + "epoch": 5.155413503795446, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12904 + }, + { + "epoch": 5.155813024370755, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12905 + }, + { + "epoch": 5.156212544946065, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12906 + }, + { + "epoch": 5.156612065521374, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12907 + }, + { + "epoch": 5.157011586096684, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 12908 + }, + { + "epoch": 5.157411106671994, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12909 + }, + { + "epoch": 5.157810627247303, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12910 + }, + { + "epoch": 5.158210147822613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12911 + }, + { + "epoch": 5.158609668397922, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12912 + }, + { + "epoch": 5.159009188973232, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12913 + }, + { + "epoch": 5.159408709548542, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12914 + }, + { + "epoch": 5.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12915 + }, + { + "epoch": 5.160207750699161, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12916 + }, + { + "epoch": 5.160607271274471, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12917 + }, + { + "epoch": 5.16100679184978, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12918 + }, + { + "epoch": 5.16140631242509, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12919 + }, + { + "epoch": 5.1618058330003995, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12920 + }, + { + "epoch": 5.162205353575709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12921 + }, + { + "epoch": 5.162604874151019, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12922 + }, + { + "epoch": 5.163004394726328, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12923 + }, + { + "epoch": 5.163403915301638, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12924 + }, + { + "epoch": 5.163803435876948, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12925 + }, + { + "epoch": 5.164202956452257, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12926 + }, + { + "epoch": 5.164602477027567, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12927 + }, + { + "epoch": 5.165001997602877, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12928 + }, + { + "epoch": 5.165401518178186, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12929 + }, + { + "epoch": 5.165801038753496, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12930 + }, + { + "epoch": 5.166200559328805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12931 + }, + { + "epoch": 5.166600079904115, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12932 + }, + { + "epoch": 5.166999600479425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12933 + }, + { + "epoch": 5.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12934 + }, + { + "epoch": 5.167798641630044, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12935 + }, + { + "epoch": 5.168198162205353, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12936 + }, + { + "epoch": 5.168597682780663, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12937 + }, + { + "epoch": 5.168997203355973, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12938 + }, + { + "epoch": 5.169396723931283, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12939 + }, + { + "epoch": 5.169796244506592, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12940 + }, + { + "epoch": 5.1701957650819015, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12941 + }, + { + "epoch": 5.170595285657211, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12942 + }, + { + "epoch": 5.170994806232521, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12943 + }, + { + "epoch": 5.171394326807831, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 12944 + }, + { + "epoch": 5.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12945 + }, + { + "epoch": 5.1721933679584495, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12946 + }, + { + "epoch": 5.172592888533759, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12947 + }, + { + "epoch": 5.172992409109069, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12948 + }, + { + "epoch": 5.173391929684379, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 12949 + }, + { + "epoch": 5.173791450259689, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12950 + }, + { + "epoch": 5.174190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12951 + }, + { + "epoch": 5.174590491410307, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12952 + }, + { + "epoch": 5.174990011985617, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12953 + }, + { + "epoch": 5.175389532560927, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12954 + }, + { + "epoch": 5.175789053136237, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12955 + }, + { + "epoch": 5.1761885737115465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12956 + }, + { + "epoch": 5.176588094286855, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12957 + }, + { + "epoch": 5.176987614862165, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12958 + }, + { + "epoch": 5.177387135437475, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12959 + }, + { + "epoch": 5.177786656012785, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12960 + }, + { + "epoch": 5.1781861765880945, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12961 + }, + { + "epoch": 5.178585697163404, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12962 + }, + { + "epoch": 5.178985217738713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12963 + }, + { + "epoch": 5.179384738314023, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12964 + }, + { + "epoch": 5.179784258889333, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12965 + }, + { + "epoch": 5.180183779464643, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12966 + }, + { + "epoch": 5.180583300039952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12967 + }, + { + "epoch": 5.180982820615261, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12968 + }, + { + "epoch": 5.181382341190571, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12969 + }, + { + "epoch": 5.181781861765881, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12970 + }, + { + "epoch": 5.182181382341191, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12971 + }, + { + "epoch": 5.1825809029165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12972 + }, + { + "epoch": 5.18298042349181, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12973 + }, + { + "epoch": 5.183379944067119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12974 + }, + { + "epoch": 5.183779464642429, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12975 + }, + { + "epoch": 5.184178985217739, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12976 + }, + { + "epoch": 5.1845785057930485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12977 + }, + { + "epoch": 5.184978026368358, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12978 + }, + { + "epoch": 5.185377546943668, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 12979 + }, + { + "epoch": 5.185777067518977, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12980 + }, + { + "epoch": 5.186176588094287, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12981 + }, + { + "epoch": 5.1865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12982 + }, + { + "epoch": 5.186975629244906, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12983 + }, + { + "epoch": 5.187375149820216, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12984 + }, + { + "epoch": 5.187774670395525, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12985 + }, + { + "epoch": 5.188174190970835, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12986 + }, + { + "epoch": 5.1885737115461446, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12987 + }, + { + "epoch": 5.188973232121454, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12988 + }, + { + "epoch": 5.189372752696764, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12989 + }, + { + "epoch": 5.189772273272074, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12990 + }, + { + "epoch": 5.190171793847383, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12991 + }, + { + "epoch": 5.190571314422693, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12992 + }, + { + "epoch": 5.190970834998002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12993 + }, + { + "epoch": 5.191370355573312, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 12994 + }, + { + "epoch": 5.191769876148622, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12995 + }, + { + "epoch": 5.192169396723932, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12996 + }, + { + "epoch": 5.192568917299241, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12997 + }, + { + "epoch": 5.19296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12998 + }, + { + "epoch": 5.19336795844986, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12999 + }, + { + "epoch": 5.19376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13000 + }, + { + "epoch": 5.19416699960048, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13001 + }, + { + "epoch": 5.194566520175789, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13002 + }, + { + "epoch": 5.1949660407510985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13003 + }, + { + "epoch": 5.195365561326408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13004 + }, + { + "epoch": 5.195765081901718, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13005 + }, + { + "epoch": 5.196164602477028, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13006 + }, + { + "epoch": 5.196564123052338, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13007 + }, + { + "epoch": 5.1969636436276465, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13008 + }, + { + "epoch": 5.197363164202956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13009 + }, + { + "epoch": 5.197762684778266, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13010 + }, + { + "epoch": 5.198162205353576, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13011 + }, + { + "epoch": 5.198561725928886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13012 + }, + { + "epoch": 5.1989612465041946, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13013 + }, + { + "epoch": 5.199360767079504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13014 + }, + { + "epoch": 5.199760287654814, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13015 + }, + { + "epoch": 5.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13016 + }, + { + "epoch": 5.200559328805434, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13017 + }, + { + "epoch": 5.2009588493807435, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13018 + }, + { + "epoch": 5.201358369956052, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13019 + }, + { + "epoch": 5.201757890531362, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13020 + }, + { + "epoch": 5.202157411106672, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13021 + }, + { + "epoch": 5.202556931681982, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13022 + }, + { + "epoch": 5.2029564522572915, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 13023 + }, + { + "epoch": 5.2033559728326, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13024 + }, + { + "epoch": 5.20375549340791, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13025 + }, + { + "epoch": 5.20415501398322, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13026 + }, + { + "epoch": 5.20455453455853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 13027 + }, + { + "epoch": 5.20495405513384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13028 + }, + { + "epoch": 5.205353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13029 + }, + { + "epoch": 5.205753096284458, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13030 + }, + { + "epoch": 5.206152616859768, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13031 + }, + { + "epoch": 5.206552137435078, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13032 + }, + { + "epoch": 5.206951658010388, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13033 + }, + { + "epoch": 5.207351178585697, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13034 + }, + { + "epoch": 5.207750699161007, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13035 + }, + { + "epoch": 5.208150219736316, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13036 + }, + { + "epoch": 5.208549740311626, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13037 + }, + { + "epoch": 5.208949260886936, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 13038 + }, + { + "epoch": 5.2093487814622454, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13039 + }, + { + "epoch": 5.209748302037555, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13040 + }, + { + "epoch": 5.210147822612864, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13041 + }, + { + "epoch": 5.210547343188174, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13042 + }, + { + "epoch": 5.210946863763484, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13043 + }, + { + "epoch": 5.2113463843387935, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13044 + }, + { + "epoch": 5.211745904914103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13045 + }, + { + "epoch": 5.212145425489413, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13046 + }, + { + "epoch": 5.212544946064722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13047 + }, + { + "epoch": 5.212944466640032, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13048 + }, + { + "epoch": 5.2133439872153415, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 13049 + }, + { + "epoch": 5.213743507790651, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13050 + }, + { + "epoch": 5.214143028365961, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13051 + }, + { + "epoch": 5.214542548941271, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13052 + }, + { + "epoch": 5.21494206951658, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13053 + }, + { + "epoch": 5.21534159009189, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13054 + }, + { + "epoch": 5.215741110667199, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13055 + }, + { + "epoch": 5.216140631242509, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 13056 + }, + { + "epoch": 5.216540151817819, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13057 + }, + { + "epoch": 5.216939672393128, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13058 + }, + { + "epoch": 5.217339192968438, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13059 + }, + { + "epoch": 5.217738713543747, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13060 + }, + { + "epoch": 5.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13061 + }, + { + "epoch": 5.218537754694367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 13062 + }, + { + "epoch": 5.218937275269677, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13063 + }, + { + "epoch": 5.219336795844986, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13064 + }, + { + "epoch": 5.2197363164202955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13065 + }, + { + "epoch": 5.220135836995605, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13066 + }, + { + "epoch": 5.220535357570915, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13067 + }, + { + "epoch": 5.220934878146225, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13068 + }, + { + "epoch": 5.221334398721535, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13069 + }, + { + "epoch": 5.2217339192968435, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13070 + }, + { + "epoch": 5.222133439872153, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13071 + }, + { + "epoch": 5.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13072 + }, + { + "epoch": 5.222932481022773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13073 + }, + { + "epoch": 5.223332001598083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13074 + }, + { + "epoch": 5.2237315221733915, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13075 + }, + { + "epoch": 5.224131042748701, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13076 + }, + { + "epoch": 5.224530563324011, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13077 + }, + { + "epoch": 5.224930083899321, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13078 + }, + { + "epoch": 5.225329604474631, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13079 + }, + { + "epoch": 5.2257291250499405, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13080 + }, + { + "epoch": 5.226128645625249, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13081 + }, + { + "epoch": 5.226528166200559, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13082 + }, + { + "epoch": 5.226927686775869, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13083 + }, + { + "epoch": 5.227327207351179, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13084 + }, + { + "epoch": 5.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13085 + }, + { + "epoch": 5.228126248501797, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13086 + }, + { + "epoch": 5.228525769077107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13087 + }, + { + "epoch": 5.228925289652417, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13088 + }, + { + "epoch": 5.229324810227727, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13089 + }, + { + "epoch": 5.229724330803037, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13090 + }, + { + "epoch": 5.230123851378346, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13091 + }, + { + "epoch": 5.230523371953655, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13092 + }, + { + "epoch": 5.230922892528965, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13093 + }, + { + "epoch": 5.231322413104275, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13094 + }, + { + "epoch": 5.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13095 + }, + { + "epoch": 5.232121454254894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13096 + }, + { + "epoch": 5.232520974830204, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13097 + }, + { + "epoch": 5.232920495405513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13098 + }, + { + "epoch": 5.233320015980823, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13099 + }, + { + "epoch": 5.233719536556133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13100 + }, + { + "epoch": 5.234119057131442, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13101 + }, + { + "epoch": 5.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13102 + }, + { + "epoch": 5.234918098282061, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13103 + }, + { + "epoch": 5.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13104 + }, + { + "epoch": 5.235717139432681, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13105 + }, + { + "epoch": 5.2361166600079905, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13106 + }, + { + "epoch": 5.2365161805833, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13107 + }, + { + "epoch": 5.23691570115861, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13108 + }, + { + "epoch": 5.237315221733919, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13109 + }, + { + "epoch": 5.237714742309229, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13110 + }, + { + "epoch": 5.2381142628845385, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13111 + }, + { + "epoch": 5.238513783459848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13112 + }, + { + "epoch": 5.238913304035158, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13113 + }, + { + "epoch": 5.239312824610467, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13114 + }, + { + "epoch": 5.239712345185777, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13115 + }, + { + "epoch": 5.240111865761087, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13116 + }, + { + "epoch": 5.240511386336396, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13117 + }, + { + "epoch": 5.240910906911706, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13118 + }, + { + "epoch": 5.241310427487016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13119 + }, + { + "epoch": 5.241709948062325, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13120 + }, + { + "epoch": 5.242109468637635, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13121 + }, + { + "epoch": 5.242508989212944, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13122 + }, + { + "epoch": 5.242908509788254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13123 + }, + { + "epoch": 5.243308030363564, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13124 + }, + { + "epoch": 5.243707550938874, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13125 + }, + { + "epoch": 5.244107071514183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13126 + }, + { + "epoch": 5.244506592089492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13127 + }, + { + "epoch": 5.244906112664802, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13128 + }, + { + "epoch": 5.245305633240112, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 13129 + }, + { + "epoch": 5.245705153815422, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13130 + }, + { + "epoch": 5.246104674390731, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13131 + }, + { + "epoch": 5.2465041949660405, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13132 + }, + { + "epoch": 5.24690371554135, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13133 + }, + { + "epoch": 5.24730323611666, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13134 + }, + { + "epoch": 5.24770275669197, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13135 + }, + { + "epoch": 5.24810227726728, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13136 + }, + { + "epoch": 5.2485017978425885, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13137 + }, + { + "epoch": 5.248901318417898, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13138 + }, + { + "epoch": 5.249300838993208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13139 + }, + { + "epoch": 5.249700359568518, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13140 + }, + { + "epoch": 5.250099880143828, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13141 + }, + { + "epoch": 5.2504994007191375, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13142 + }, + { + "epoch": 5.250898921294446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13143 + }, + { + "epoch": 5.251298441869756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13144 + }, + { + "epoch": 5.251697962445066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13145 + }, + { + "epoch": 5.252097483020376, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13146 + }, + { + "epoch": 5.2524970035956855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13147 + }, + { + "epoch": 5.252896524170994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13148 + }, + { + "epoch": 5.253296044746304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13149 + }, + { + "epoch": 5.253695565321614, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13150 + }, + { + "epoch": 5.254095085896924, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13151 + }, + { + "epoch": 5.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13152 + }, + { + "epoch": 5.254894127047543, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13153 + }, + { + "epoch": 5.255293647622852, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13154 + }, + { + "epoch": 5.255693168198162, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13155 + }, + { + "epoch": 5.256092688773472, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13156 + }, + { + "epoch": 5.256492209348782, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13157 + }, + { + "epoch": 5.256891729924091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13158 + }, + { + "epoch": 5.257291250499401, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13159 + }, + { + "epoch": 5.25769077107471, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13160 + }, + { + "epoch": 5.25809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13161 + }, + { + "epoch": 5.25848981222533, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13162 + }, + { + "epoch": 5.258889332800639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13163 + }, + { + "epoch": 5.259288853375949, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13164 + }, + { + "epoch": 5.259688373951258, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13165 + }, + { + "epoch": 5.260087894526568, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13166 + }, + { + "epoch": 5.260487415101878, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13167 + }, + { + "epoch": 5.2608869356771875, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13168 + }, + { + "epoch": 5.261286456252497, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13169 + }, + { + "epoch": 5.261685976827807, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13170 + }, + { + "epoch": 5.262085497403116, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13171 + }, + { + "epoch": 5.262485017978426, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13172 + }, + { + "epoch": 5.2628845385537355, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13173 + }, + { + "epoch": 5.263284059129045, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13174 + }, + { + "epoch": 5.263683579704355, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13175 + }, + { + "epoch": 5.264083100279665, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13176 + }, + { + "epoch": 5.264482620854974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13177 + }, + { + "epoch": 5.2648821414302835, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13178 + }, + { + "epoch": 5.265281662005593, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13179 + }, + { + "epoch": 5.265681182580903, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13180 + }, + { + "epoch": 5.266080703156213, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13181 + }, + { + "epoch": 5.266480223731522, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13182 + }, + { + "epoch": 5.266879744306832, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13183 + }, + { + "epoch": 5.267279264882141, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13184 + }, + { + "epoch": 5.267678785457451, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13185 + }, + { + "epoch": 5.268078306032761, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13186 + }, + { + "epoch": 5.26847782660807, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13187 + }, + { + "epoch": 5.26887734718338, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13188 + }, + { + "epoch": 5.269276867758689, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13189 + }, + { + "epoch": 5.269676388333999, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13190 + }, + { + "epoch": 5.270075908909309, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13191 + }, + { + "epoch": 5.270475429484619, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13192 + }, + { + "epoch": 5.270874950059928, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13193 + }, + { + "epoch": 5.2712744706352375, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13194 + }, + { + "epoch": 5.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13195 + }, + { + "epoch": 5.272073511785857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13196 + }, + { + "epoch": 5.272473032361167, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13197 + }, + { + "epoch": 5.272872552936477, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13198 + }, + { + "epoch": 5.2732720735117855, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13199 + }, + { + "epoch": 5.273671594087095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13200 + }, + { + "epoch": 5.274071114662405, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13201 + }, + { + "epoch": 5.274470635237715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13202 + }, + { + "epoch": 5.274870155813025, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13203 + }, + { + "epoch": 5.2752696763883336, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 13204 + }, + { + "epoch": 5.275669196963643, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13205 + }, + { + "epoch": 5.276068717538953, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13206 + }, + { + "epoch": 5.276468238114263, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13207 + }, + { + "epoch": 5.276867758689573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13208 + }, + { + "epoch": 5.2772672792648825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13209 + }, + { + "epoch": 5.277666799840191, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13210 + }, + { + "epoch": 5.278066320415501, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13211 + }, + { + "epoch": 5.278465840990811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 13212 + }, + { + "epoch": 5.278865361566121, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13213 + }, + { + "epoch": 5.2792648821414305, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13214 + }, + { + "epoch": 5.27966440271674, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13215 + }, + { + "epoch": 5.280063923292049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13216 + }, + { + "epoch": 5.280463443867359, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13217 + }, + { + "epoch": 5.280862964442669, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13218 + }, + { + "epoch": 5.281262485017979, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13219 + }, + { + "epoch": 5.281662005593288, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13220 + }, + { + "epoch": 5.282061526168597, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13221 + }, + { + "epoch": 5.282461046743907, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13222 + }, + { + "epoch": 5.282860567319217, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13223 + }, + { + "epoch": 5.283260087894527, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13224 + }, + { + "epoch": 5.283659608469836, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13225 + }, + { + "epoch": 5.284059129045146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13226 + }, + { + "epoch": 5.284458649620455, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13227 + }, + { + "epoch": 5.284858170195765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 13228 + }, + { + "epoch": 5.285257690771075, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 13229 + }, + { + "epoch": 5.285657211346384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13230 + }, + { + "epoch": 5.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13231 + }, + { + "epoch": 5.286456252497004, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13232 + }, + { + "epoch": 5.286855773072313, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 13233 + }, + { + "epoch": 5.287255293647623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13234 + }, + { + "epoch": 5.2876548142229325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13235 + }, + { + "epoch": 5.288054334798242, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13236 + }, + { + "epoch": 5.288453855373552, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13237 + }, + { + "epoch": 5.288853375948861, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13238 + }, + { + "epoch": 5.289252896524171, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13239 + }, + { + "epoch": 5.2896524170994805, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13240 + }, + { + "epoch": 5.29005193767479, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13241 + }, + { + "epoch": 5.2904514582501, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13242 + }, + { + "epoch": 5.29085097882541, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13243 + }, + { + "epoch": 5.291250499400719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 13244 + }, + { + "epoch": 5.291650019976029, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13245 + }, + { + "epoch": 5.292049540551338, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13246 + }, + { + "epoch": 5.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13247 + }, + { + "epoch": 5.292848581701958, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13248 + }, + { + "epoch": 5.293248102277268, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13249 + }, + { + "epoch": 5.293647622852577, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13250 + }, + { + "epoch": 5.294047143427886, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13251 + }, + { + "epoch": 5.294446664003196, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13252 + }, + { + "epoch": 5.294846184578506, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13253 + }, + { + "epoch": 5.295245705153816, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13254 + }, + { + "epoch": 5.295645225729125, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13255 + }, + { + "epoch": 5.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13256 + }, + { + "epoch": 5.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.2965, + "step": 13257 + }, + { + "epoch": 5.296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13258 + }, + { + "epoch": 5.297243308030364, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13259 + }, + { + "epoch": 5.297642828605674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13260 + }, + { + "epoch": 5.2980423491809825, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13261 + }, + { + "epoch": 5.298441869756292, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13262 + }, + { + "epoch": 5.298841390331602, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13263 + }, + { + "epoch": 5.299240910906912, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13264 + }, + { + "epoch": 5.299640431482222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13265 + }, + { + "epoch": 5.300039952057531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13266 + }, + { + "epoch": 5.30043947263284, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13267 + }, + { + "epoch": 5.30083899320815, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13268 + }, + { + "epoch": 5.30123851378346, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13269 + }, + { + "epoch": 5.30163803435877, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13270 + }, + { + "epoch": 5.3020375549340795, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13271 + }, + { + "epoch": 5.302437075509388, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13272 + }, + { + "epoch": 5.302836596084698, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13273 + }, + { + "epoch": 5.303236116660008, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13274 + }, + { + "epoch": 5.303635637235318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13275 + }, + { + "epoch": 5.3040351578106275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13276 + }, + { + "epoch": 5.304434678385936, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13277 + }, + { + "epoch": 5.304834198961246, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13278 + }, + { + "epoch": 5.305233719536556, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13279 + }, + { + "epoch": 5.305633240111866, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13280 + }, + { + "epoch": 5.3060327606871756, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13281 + }, + { + "epoch": 5.306432281262485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13282 + }, + { + "epoch": 5.306831801837794, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13283 + }, + { + "epoch": 5.307231322413104, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13284 + }, + { + "epoch": 5.307630842988414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13285 + }, + { + "epoch": 5.308030363563724, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13286 + }, + { + "epoch": 5.308429884139033, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13287 + }, + { + "epoch": 5.308829404714343, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13288 + }, + { + "epoch": 5.309228925289652, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13289 + }, + { + "epoch": 5.309628445864962, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13290 + }, + { + "epoch": 5.310027966440272, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13291 + }, + { + "epoch": 5.310427487015581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 13292 + }, + { + "epoch": 5.310827007590891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13293 + }, + { + "epoch": 5.3112265281662, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13294 + }, + { + "epoch": 5.31162604874151, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13295 + }, + { + "epoch": 5.31202556931682, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13296 + }, + { + "epoch": 5.3124250898921295, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13297 + }, + { + "epoch": 5.312824610467439, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13298 + }, + { + "epoch": 5.313224131042749, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13299 + }, + { + "epoch": 5.313623651618058, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13300 + }, + { + "epoch": 5.314023172193368, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13301 + }, + { + "epoch": 5.3144226927686775, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13302 + }, + { + "epoch": 5.314822213343987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13303 + }, + { + "epoch": 5.315221733919297, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13304 + }, + { + "epoch": 5.315621254494607, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13305 + }, + { + "epoch": 5.316020775069916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13306 + }, + { + "epoch": 5.316420295645226, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13307 + }, + { + "epoch": 5.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13308 + }, + { + "epoch": 5.317219336795845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13309 + }, + { + "epoch": 5.317618857371155, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13310 + }, + { + "epoch": 5.318018377946464, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13311 + }, + { + "epoch": 5.318417898521774, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13312 + }, + { + "epoch": 5.318817419097083, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13313 + }, + { + "epoch": 5.319216939672393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 13314 + }, + { + "epoch": 5.319616460247703, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13315 + }, + { + "epoch": 5.320015980823013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 13316 + }, + { + "epoch": 5.320415501398322, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13317 + }, + { + "epoch": 5.320815021973631, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13318 + }, + { + "epoch": 5.321214542548941, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13319 + }, + { + "epoch": 5.321614063124251, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13320 + }, + { + "epoch": 5.322013583699561, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13321 + }, + { + "epoch": 5.322413104274871, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13322 + }, + { + "epoch": 5.3228126248501795, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13323 + }, + { + "epoch": 5.323212145425489, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13324 + }, + { + "epoch": 5.323611666000799, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13325 + }, + { + "epoch": 5.324011186576109, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13326 + }, + { + "epoch": 5.324410707151419, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13327 + }, + { + "epoch": 5.3248102277267275, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13328 + }, + { + "epoch": 5.325209748302037, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13329 + }, + { + "epoch": 5.325609268877347, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13330 + }, + { + "epoch": 5.326008789452657, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13331 + }, + { + "epoch": 5.326408310027967, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13332 + }, + { + "epoch": 5.3268078306032765, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13333 + }, + { + "epoch": 5.327207351178585, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13334 + }, + { + "epoch": 5.327606871753895, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13335 + }, + { + "epoch": 5.328006392329205, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 13336 + }, + { + "epoch": 5.328405912904515, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13337 + }, + { + "epoch": 5.3288054334798245, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13338 + }, + { + "epoch": 5.329204954055134, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13339 + }, + { + "epoch": 5.329604474630443, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13340 + }, + { + "epoch": 5.330003995205753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13341 + }, + { + "epoch": 5.330403515781063, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13342 + }, + { + "epoch": 5.3308030363563725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13343 + }, + { + "epoch": 5.331202556931682, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13344 + }, + { + "epoch": 5.331602077506991, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13345 + }, + { + "epoch": 5.332001598082301, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13346 + }, + { + "epoch": 5.332401118657611, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13347 + }, + { + "epoch": 5.332800639232921, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13348 + }, + { + "epoch": 5.33320015980823, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13349 + }, + { + "epoch": 5.33359968038354, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13350 + }, + { + "epoch": 5.333999200958849, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13351 + }, + { + "epoch": 5.334398721534159, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13352 + }, + { + "epoch": 5.334798242109469, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13353 + }, + { + "epoch": 5.335197762684778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13354 + }, + { + "epoch": 5.335597283260088, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13355 + }, + { + "epoch": 5.335996803835398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13356 + }, + { + "epoch": 5.336396324410707, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13357 + }, + { + "epoch": 5.336795844986017, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13358 + }, + { + "epoch": 5.3371953655613265, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13359 + }, + { + "epoch": 5.337594886136636, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13360 + }, + { + "epoch": 5.337994406711946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13361 + }, + { + "epoch": 5.338393927287255, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13362 + }, + { + "epoch": 5.338793447862565, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13363 + }, + { + "epoch": 5.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13364 + }, + { + "epoch": 5.339592489013184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13365 + }, + { + "epoch": 5.339992009588494, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13366 + }, + { + "epoch": 5.340391530163803, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13367 + }, + { + "epoch": 5.340791050739113, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13368 + }, + { + "epoch": 5.3411905713144225, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13369 + }, + { + "epoch": 5.341590091889732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13370 + }, + { + "epoch": 5.341989612465042, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13371 + }, + { + "epoch": 5.342389133040352, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13372 + }, + { + "epoch": 5.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13373 + }, + { + "epoch": 5.343188174190971, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 13374 + }, + { + "epoch": 5.34358769476628, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13375 + }, + { + "epoch": 5.34398721534159, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13376 + }, + { + "epoch": 5.3443867359169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13377 + }, + { + "epoch": 5.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13378 + }, + { + "epoch": 5.345185777067519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13379 + }, + { + "epoch": 5.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13380 + }, + { + "epoch": 5.345984818218138, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13381 + }, + { + "epoch": 5.346384338793448, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13382 + }, + { + "epoch": 5.346783859368758, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13383 + }, + { + "epoch": 5.347183379944067, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13384 + }, + { + "epoch": 5.3475829005193765, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13385 + }, + { + "epoch": 5.347982421094686, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13386 + }, + { + "epoch": 5.348381941669996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13387 + }, + { + "epoch": 5.348781462245306, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13388 + }, + { + "epoch": 5.349180982820616, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13389 + }, + { + "epoch": 5.3495805033959245, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13390 + }, + { + "epoch": 5.349980023971234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13391 + }, + { + "epoch": 5.350379544546544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13392 + }, + { + "epoch": 5.350779065121854, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13393 + }, + { + "epoch": 5.351178585697164, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13394 + }, + { + "epoch": 5.351578106272473, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13395 + }, + { + "epoch": 5.351977626847782, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13396 + }, + { + "epoch": 5.352377147423092, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13397 + }, + { + "epoch": 5.352776667998402, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13398 + }, + { + "epoch": 5.353176188573712, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13399 + }, + { + "epoch": 5.3535757091490215, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13400 + }, + { + "epoch": 5.35397522972433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13401 + }, + { + "epoch": 5.35437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13402 + }, + { + "epoch": 5.35477427087495, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13403 + }, + { + "epoch": 5.35517379145026, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13404 + }, + { + "epoch": 5.3555733120255695, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13405 + }, + { + "epoch": 5.355972832600879, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13406 + }, + { + "epoch": 5.356372353176188, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13407 + }, + { + "epoch": 5.356771873751498, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 13408 + }, + { + "epoch": 5.357171394326808, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13409 + }, + { + "epoch": 5.357570914902118, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13410 + }, + { + "epoch": 5.357970435477427, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13411 + }, + { + "epoch": 5.358369956052737, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13412 + }, + { + "epoch": 5.358769476628046, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13413 + }, + { + "epoch": 5.359168997203356, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13414 + }, + { + "epoch": 5.359568517778666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13415 + }, + { + "epoch": 5.359968038353975, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 13416 + }, + { + "epoch": 5.360367558929285, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13417 + }, + { + "epoch": 5.360767079504594, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13418 + }, + { + "epoch": 5.361166600079904, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13419 + }, + { + "epoch": 5.361566120655214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13420 + }, + { + "epoch": 5.361965641230523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13421 + }, + { + "epoch": 5.362365161805833, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13422 + }, + { + "epoch": 5.362764682381143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13423 + }, + { + "epoch": 5.363164202956452, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13424 + }, + { + "epoch": 5.363563723531762, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 13425 + }, + { + "epoch": 5.3639632441070715, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13426 + }, + { + "epoch": 5.364362764682381, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13427 + }, + { + "epoch": 5.364762285257691, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13428 + }, + { + "epoch": 5.365161805833001, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13429 + }, + { + "epoch": 5.36556132640831, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 13430 + }, + { + "epoch": 5.3659608469836195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13431 + }, + { + "epoch": 5.366360367558929, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13432 + }, + { + "epoch": 5.366759888134239, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13433 + }, + { + "epoch": 5.367159408709549, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13434 + }, + { + "epoch": 5.367558929284858, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 13435 + }, + { + "epoch": 5.367958449860168, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13436 + }, + { + "epoch": 5.368357970435477, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 13437 + }, + { + "epoch": 5.368757491010787, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13438 + }, + { + "epoch": 5.369157011586097, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13439 + }, + { + "epoch": 5.369556532161407, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13440 + }, + { + "epoch": 5.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13441 + }, + { + "epoch": 5.370355573312025, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13442 + }, + { + "epoch": 5.370755093887335, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 13443 + }, + { + "epoch": 5.371154614462645, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13444 + }, + { + "epoch": 5.371554135037955, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13445 + }, + { + "epoch": 5.3719536556132645, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13446 + }, + { + "epoch": 5.372353176188573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13447 + }, + { + "epoch": 5.372752696763883, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13448 + }, + { + "epoch": 5.373152217339193, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13449 + }, + { + "epoch": 5.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13450 + }, + { + "epoch": 5.373951258489813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13451 + }, + { + "epoch": 5.3743507790651215, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13452 + }, + { + "epoch": 5.374750299640431, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13453 + }, + { + "epoch": 5.375149820215741, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13454 + }, + { + "epoch": 5.375549340791051, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13455 + }, + { + "epoch": 5.375948861366361, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13456 + }, + { + "epoch": 5.3763483819416695, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13457 + }, + { + "epoch": 5.376747902516979, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13458 + }, + { + "epoch": 5.377147423092289, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13459 + }, + { + "epoch": 5.377546943667599, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13460 + }, + { + "epoch": 5.377946464242909, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13461 + }, + { + "epoch": 5.3783459848182185, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13462 + }, + { + "epoch": 5.378745505393527, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13463 + }, + { + "epoch": 5.379145025968837, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13464 + }, + { + "epoch": 5.379544546544147, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13465 + }, + { + "epoch": 5.379944067119457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13466 + }, + { + "epoch": 5.3803435876947665, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13467 + }, + { + "epoch": 5.380743108270076, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13468 + }, + { + "epoch": 5.381142628845385, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13469 + }, + { + "epoch": 5.381542149420695, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13470 + }, + { + "epoch": 5.381941669996005, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13471 + }, + { + "epoch": 5.3823411905713145, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13472 + }, + { + "epoch": 5.382740711146624, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13473 + }, + { + "epoch": 5.383140231721933, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13474 + }, + { + "epoch": 5.383539752297243, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13475 + }, + { + "epoch": 5.383939272872553, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13476 + }, + { + "epoch": 5.384338793447863, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13477 + }, + { + "epoch": 5.384738314023172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13478 + }, + { + "epoch": 5.385137834598482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13479 + }, + { + "epoch": 5.385537355173791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13480 + }, + { + "epoch": 5.385936875749101, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13481 + }, + { + "epoch": 5.386336396324411, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13482 + }, + { + "epoch": 5.38673591689972, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13483 + }, + { + "epoch": 5.38713543747503, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13484 + }, + { + "epoch": 5.38753495805034, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13485 + }, + { + "epoch": 5.387934478625649, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13486 + }, + { + "epoch": 5.388333999200959, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 13487 + }, + { + "epoch": 5.3887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13488 + }, + { + "epoch": 5.389133040351578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13489 + }, + { + "epoch": 5.389532560926888, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13490 + }, + { + "epoch": 5.389932081502197, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13491 + }, + { + "epoch": 5.390331602077507, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13492 + }, + { + "epoch": 5.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13493 + }, + { + "epoch": 5.391130643228126, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13494 + }, + { + "epoch": 5.391530163803436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13495 + }, + { + "epoch": 5.391929684378746, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13496 + }, + { + "epoch": 5.392329204954055, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13497 + }, + { + "epoch": 5.3927287255293646, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 13498 + }, + { + "epoch": 5.393128246104674, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13499 + }, + { + "epoch": 5.393527766679984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13500 + }, + { + "epoch": 5.393927287255294, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13501 + }, + { + "epoch": 5.394326807830604, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13502 + }, + { + "epoch": 5.394726328405913, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13503 + }, + { + "epoch": 5.395125848981222, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13504 + }, + { + "epoch": 5.395525369556532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13505 + }, + { + "epoch": 5.395924890131842, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13506 + }, + { + "epoch": 5.396324410707152, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13507 + }, + { + "epoch": 5.396723931282461, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13508 + }, + { + "epoch": 5.39712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13509 + }, + { + "epoch": 5.39752297243308, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13510 + }, + { + "epoch": 5.39792249300839, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13511 + }, + { + "epoch": 5.3983220135837, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13512 + }, + { + "epoch": 5.39872153415901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13513 + }, + { + "epoch": 5.3991210547343185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13514 + }, + { + "epoch": 5.399520575309628, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13515 + }, + { + "epoch": 5.399920095884938, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13516 + }, + { + "epoch": 5.400319616460248, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13517 + }, + { + "epoch": 5.400719137035558, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13518 + }, + { + "epoch": 5.401118657610867, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13519 + }, + { + "epoch": 5.401518178186176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13520 + }, + { + "epoch": 5.401917698761486, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13521 + }, + { + "epoch": 5.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13522 + }, + { + "epoch": 5.402716739912106, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13523 + }, + { + "epoch": 5.4031162604874154, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13524 + }, + { + "epoch": 5.403515781062724, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13525 + }, + { + "epoch": 5.403915301638034, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13526 + }, + { + "epoch": 5.404314822213344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13527 + }, + { + "epoch": 5.404714342788654, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13528 + }, + { + "epoch": 5.4051138633639635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13529 + }, + { + "epoch": 5.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.2988, + "step": 13530 + }, + { + "epoch": 5.405912904514582, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13531 + }, + { + "epoch": 5.406312425089892, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13532 + }, + { + "epoch": 5.406711945665202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13533 + }, + { + "epoch": 5.4071114662405115, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13534 + }, + { + "epoch": 5.407510986815821, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13535 + }, + { + "epoch": 5.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13536 + }, + { + "epoch": 5.40831002796644, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 13537 + }, + { + "epoch": 5.40870954854175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13538 + }, + { + "epoch": 5.40910906911706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13539 + }, + { + "epoch": 5.409508589692369, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13540 + }, + { + "epoch": 5.409908110267679, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13541 + }, + { + "epoch": 5.410307630842988, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13542 + }, + { + "epoch": 5.410707151418298, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13543 + }, + { + "epoch": 5.411106671993608, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13544 + }, + { + "epoch": 5.411506192568917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13545 + }, + { + "epoch": 5.411905713144227, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13546 + }, + { + "epoch": 5.412305233719536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13547 + }, + { + "epoch": 5.412704754294846, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13548 + }, + { + "epoch": 5.413104274870156, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13549 + }, + { + "epoch": 5.4135037954454654, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13550 + }, + { + "epoch": 5.413903316020775, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13551 + }, + { + "epoch": 5.414302836596085, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13552 + }, + { + "epoch": 5.414702357171394, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 13553 + }, + { + "epoch": 5.415101877746704, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13554 + }, + { + "epoch": 5.4155013983220135, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13555 + }, + { + "epoch": 5.415900918897323, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 13556 + }, + { + "epoch": 5.416300439472633, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13557 + }, + { + "epoch": 5.416699960047943, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13558 + }, + { + "epoch": 5.417099480623252, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13559 + }, + { + "epoch": 5.4174990011985615, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 13560 + }, + { + "epoch": 5.417898521773871, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13561 + }, + { + "epoch": 5.418298042349181, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13562 + }, + { + "epoch": 5.418697562924491, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13563 + }, + { + "epoch": 5.4190970834998, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13564 + }, + { + "epoch": 5.41949660407511, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13565 + }, + { + "epoch": 5.419896124650419, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13566 + }, + { + "epoch": 5.420295645225729, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13567 + }, + { + "epoch": 5.420695165801039, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13568 + }, + { + "epoch": 5.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13569 + }, + { + "epoch": 5.421494206951658, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13570 + }, + { + "epoch": 5.421893727526967, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13571 + }, + { + "epoch": 5.422293248102277, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13572 + }, + { + "epoch": 5.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13573 + }, + { + "epoch": 5.423092289252897, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13574 + }, + { + "epoch": 5.4234918098282066, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13575 + }, + { + "epoch": 5.4238913304035155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13576 + }, + { + "epoch": 5.424290850978825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13577 + }, + { + "epoch": 5.424690371554135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13578 + }, + { + "epoch": 5.425089892129445, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13579 + }, + { + "epoch": 5.425489412704755, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13580 + }, + { + "epoch": 5.4258889332800635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13581 + }, + { + "epoch": 5.426288453855373, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13582 + }, + { + "epoch": 5.426687974430683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13583 + }, + { + "epoch": 5.427087495005993, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13584 + }, + { + "epoch": 5.427487015581303, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13585 + }, + { + "epoch": 5.427886536156612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13586 + }, + { + "epoch": 5.428286056731921, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13587 + }, + { + "epoch": 5.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 13588 + }, + { + "epoch": 5.429085097882541, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13589 + }, + { + "epoch": 5.429484618457851, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13590 + }, + { + "epoch": 5.4298841390331605, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13591 + }, + { + "epoch": 5.43028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13592 + }, + { + "epoch": 5.430683180183779, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13593 + }, + { + "epoch": 5.431082700759089, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13594 + }, + { + "epoch": 5.431482221334399, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13595 + }, + { + "epoch": 5.4318817419097085, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13596 + }, + { + "epoch": 5.432281262485018, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13597 + }, + { + "epoch": 5.432680783060327, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13598 + }, + { + "epoch": 5.433080303635637, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13599 + }, + { + "epoch": 5.433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13600 + }, + { + "epoch": 5.433879344786257, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13601 + }, + { + "epoch": 5.434278865361566, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13602 + }, + { + "epoch": 5.434678385936876, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 13603 + }, + { + "epoch": 5.435077906512185, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13604 + }, + { + "epoch": 5.435477427087495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13605 + }, + { + "epoch": 5.435876947662805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13606 + }, + { + "epoch": 5.436276468238114, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13607 + }, + { + "epoch": 5.436675988813424, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13608 + }, + { + "epoch": 5.437075509388734, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13609 + }, + { + "epoch": 5.437475029964043, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13610 + }, + { + "epoch": 5.437874550539353, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13611 + }, + { + "epoch": 5.438274071114662, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13612 + }, + { + "epoch": 5.438673591689972, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13613 + }, + { + "epoch": 5.439073112265282, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13614 + }, + { + "epoch": 5.439472632840591, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13615 + }, + { + "epoch": 5.439872153415901, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13616 + }, + { + "epoch": 5.4402716739912105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 13617 + }, + { + "epoch": 5.44067119456652, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13618 + }, + { + "epoch": 5.44107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 13619 + }, + { + "epoch": 5.44147023571714, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13620 + }, + { + "epoch": 5.441869756292449, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13621 + }, + { + "epoch": 5.4422692768677585, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13622 + }, + { + "epoch": 5.442668797443068, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13623 + }, + { + "epoch": 5.443068318018378, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13624 + }, + { + "epoch": 5.443467838593688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13625 + }, + { + "epoch": 5.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13626 + }, + { + "epoch": 5.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13627 + }, + { + "epoch": 5.444666400319616, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13628 + }, + { + "epoch": 5.445065920894926, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13629 + }, + { + "epoch": 5.445465441470236, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13630 + }, + { + "epoch": 5.445864962045546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13631 + }, + { + "epoch": 5.446264482620855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13632 + }, + { + "epoch": 5.446664003196164, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13633 + }, + { + "epoch": 5.447063523771474, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13634 + }, + { + "epoch": 5.447463044346784, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13635 + }, + { + "epoch": 5.447862564922094, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13636 + }, + { + "epoch": 5.448262085497403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13637 + }, + { + "epoch": 5.448661606072712, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13638 + }, + { + "epoch": 5.449061126648022, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13639 + }, + { + "epoch": 5.449460647223332, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13640 + }, + { + "epoch": 5.449860167798642, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13641 + }, + { + "epoch": 5.450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13642 + }, + { + "epoch": 5.4506592089492605, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13643 + }, + { + "epoch": 5.45105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13644 + }, + { + "epoch": 5.45145825009988, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 13645 + }, + { + "epoch": 5.45185777067519, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13646 + }, + { + "epoch": 5.4522572912505, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13647 + }, + { + "epoch": 5.452656811825809, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13648 + }, + { + "epoch": 5.453056332401118, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13649 + }, + { + "epoch": 5.453455852976428, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13650 + }, + { + "epoch": 5.453855373551738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13651 + }, + { + "epoch": 5.454254894127048, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 13652 + }, + { + "epoch": 5.4546544147023575, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13653 + }, + { + "epoch": 5.455053935277666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13654 + }, + { + "epoch": 5.455453455852976, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13655 + }, + { + "epoch": 5.455852976428286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13656 + }, + { + "epoch": 5.456252497003596, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13657 + }, + { + "epoch": 5.4566520175789055, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13658 + }, + { + "epoch": 5.457051538154215, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13659 + }, + { + "epoch": 5.457451058729524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13660 + }, + { + "epoch": 5.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13661 + }, + { + "epoch": 5.458250099880144, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13662 + }, + { + "epoch": 5.4586496204554535, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13663 + }, + { + "epoch": 5.459049141030763, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 13664 + }, + { + "epoch": 5.459448661606073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13665 + }, + { + "epoch": 5.459848182181382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13666 + }, + { + "epoch": 5.460247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13667 + }, + { + "epoch": 5.460647223332002, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 13668 + }, + { + "epoch": 5.461046743907311, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13669 + }, + { + "epoch": 5.461446264482621, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13670 + }, + { + "epoch": 5.46184578505793, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13671 + }, + { + "epoch": 5.46224530563324, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13672 + }, + { + "epoch": 5.46264482620855, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13673 + }, + { + "epoch": 5.463044346783859, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13674 + }, + { + "epoch": 5.463443867359169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13675 + }, + { + "epoch": 5.463843387934479, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13676 + }, + { + "epoch": 5.464242908509788, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13677 + }, + { + "epoch": 5.464642429085098, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13678 + }, + { + "epoch": 5.4650419496604075, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13679 + }, + { + "epoch": 5.465441470235717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13680 + }, + { + "epoch": 5.465840990811027, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13681 + }, + { + "epoch": 5.466240511386337, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13682 + }, + { + "epoch": 5.466640031961646, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13683 + }, + { + "epoch": 5.4670395525369555, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13684 + }, + { + "epoch": 5.467439073112265, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13685 + }, + { + "epoch": 5.467838593687575, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13686 + }, + { + "epoch": 5.468238114262885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13687 + }, + { + "epoch": 5.468637634838194, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13688 + }, + { + "epoch": 5.4690371554135035, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13689 + }, + { + "epoch": 5.469436675988813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13690 + }, + { + "epoch": 5.469836196564123, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13691 + }, + { + "epoch": 5.470235717139433, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13692 + }, + { + "epoch": 5.470635237714743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13693 + }, + { + "epoch": 5.471034758290052, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13694 + }, + { + "epoch": 5.471434278865361, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13695 + }, + { + "epoch": 5.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13696 + }, + { + "epoch": 5.472233320015981, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13697 + }, + { + "epoch": 5.472632840591291, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13698 + }, + { + "epoch": 5.4730323611666005, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13699 + }, + { + "epoch": 5.473431881741909, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13700 + }, + { + "epoch": 5.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13701 + }, + { + "epoch": 5.474230922892529, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13702 + }, + { + "epoch": 5.474630443467839, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13703 + }, + { + "epoch": 5.475029964043149, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 13704 + }, + { + "epoch": 5.4754294846184575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13705 + }, + { + "epoch": 5.475829005193767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13706 + }, + { + "epoch": 5.476228525769077, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13707 + }, + { + "epoch": 5.476628046344387, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13708 + }, + { + "epoch": 5.477027566919697, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13709 + }, + { + "epoch": 5.477427087495006, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13710 + }, + { + "epoch": 5.477826608070315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13711 + }, + { + "epoch": 5.478226128645625, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13712 + }, + { + "epoch": 5.478625649220935, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13713 + }, + { + "epoch": 5.479025169796245, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 13714 + }, + { + "epoch": 5.479424690371554, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13715 + }, + { + "epoch": 5.479824210946864, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13716 + }, + { + "epoch": 5.480223731522173, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 13717 + }, + { + "epoch": 5.480623252097483, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13718 + }, + { + "epoch": 5.481022772672793, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13719 + }, + { + "epoch": 5.4814222932481025, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13720 + }, + { + "epoch": 5.481821813823412, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13721 + }, + { + "epoch": 5.482221334398721, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13722 + }, + { + "epoch": 5.482620854974031, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13723 + }, + { + "epoch": 5.483020375549341, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13724 + }, + { + "epoch": 5.4834198961246505, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13725 + }, + { + "epoch": 5.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 13726 + }, + { + "epoch": 5.484218937275269, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13727 + }, + { + "epoch": 5.484618457850579, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13728 + }, + { + "epoch": 5.485017978425889, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13729 + }, + { + "epoch": 5.485417499001199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13730 + }, + { + "epoch": 5.485817019576508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13731 + }, + { + "epoch": 5.486216540151818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13732 + }, + { + "epoch": 5.486616060727127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13733 + }, + { + "epoch": 5.487015581302437, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13734 + }, + { + "epoch": 5.487415101877747, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13735 + }, + { + "epoch": 5.487814622453056, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13736 + }, + { + "epoch": 5.488214143028366, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13737 + }, + { + "epoch": 5.488613663603676, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13738 + }, + { + "epoch": 5.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13739 + }, + { + "epoch": 5.489412704754295, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13740 + }, + { + "epoch": 5.4898122253296044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13741 + }, + { + "epoch": 5.490211745904914, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13742 + }, + { + "epoch": 5.490611266480224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13743 + }, + { + "epoch": 5.491010787055533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 13744 + }, + { + "epoch": 5.491410307630843, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13745 + }, + { + "epoch": 5.4918098282061525, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13746 + }, + { + "epoch": 5.492209348781462, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13747 + }, + { + "epoch": 5.492608869356772, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13748 + }, + { + "epoch": 5.493008389932082, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13749 + }, + { + "epoch": 5.493407910507391, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13750 + }, + { + "epoch": 5.4938074310827005, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13751 + }, + { + "epoch": 5.49420695165801, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13752 + }, + { + "epoch": 5.49460647223332, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13753 + }, + { + "epoch": 5.49500599280863, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13754 + }, + { + "epoch": 5.49540551338394, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13755 + }, + { + "epoch": 5.495805033959249, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13756 + }, + { + "epoch": 5.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13757 + }, + { + "epoch": 5.496604075109868, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13758 + }, + { + "epoch": 5.497003595685178, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13759 + }, + { + "epoch": 5.497403116260488, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13760 + }, + { + "epoch": 5.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13761 + }, + { + "epoch": 5.498202157411106, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13762 + }, + { + "epoch": 5.498601677986416, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13763 + }, + { + "epoch": 5.499001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13764 + }, + { + "epoch": 5.499400719137036, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13765 + }, + { + "epoch": 5.4998002397123456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13766 + }, + { + "epoch": 5.5001997602876544, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13767 + }, + { + "epoch": 5.500599280862964, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13768 + }, + { + "epoch": 5.500998801438274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13769 + }, + { + "epoch": 5.501398322013584, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13770 + }, + { + "epoch": 5.501797842588894, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13771 + }, + { + "epoch": 5.502197363164203, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13772 + }, + { + "epoch": 5.502596883739512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13773 + }, + { + "epoch": 5.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13774 + }, + { + "epoch": 5.503395924890132, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13775 + }, + { + "epoch": 5.503795445465442, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13776 + }, + { + "epoch": 5.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13777 + }, + { + "epoch": 5.50459448661606, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13778 + }, + { + "epoch": 5.50499400719137, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13779 + }, + { + "epoch": 5.50539352776668, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13780 + }, + { + "epoch": 5.50579304834199, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13781 + }, + { + "epoch": 5.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13782 + }, + { + "epoch": 5.506592089492609, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 13783 + }, + { + "epoch": 5.506991610067918, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13784 + }, + { + "epoch": 5.507391130643228, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13785 + }, + { + "epoch": 5.507790651218538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13786 + }, + { + "epoch": 5.5081901717938475, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13787 + }, + { + "epoch": 5.508589692369157, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13788 + }, + { + "epoch": 5.508989212944467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13789 + }, + { + "epoch": 5.509388733519776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13790 + }, + { + "epoch": 5.509788254095086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13791 + }, + { + "epoch": 5.5101877746703956, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13792 + }, + { + "epoch": 5.510587295245705, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 13793 + }, + { + "epoch": 5.510986815821015, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13794 + }, + { + "epoch": 5.511386336396324, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13795 + }, + { + "epoch": 5.511785856971634, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13796 + }, + { + "epoch": 5.512185377546944, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13797 + }, + { + "epoch": 5.512584898122253, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13798 + }, + { + "epoch": 5.512984418697563, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13799 + }, + { + "epoch": 5.513383939272872, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13800 + }, + { + "epoch": 5.513783459848182, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13801 + }, + { + "epoch": 5.514182980423492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13802 + }, + { + "epoch": 5.514582500998801, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13803 + }, + { + "epoch": 5.514982021574111, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13804 + }, + { + "epoch": 5.515381542149421, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13805 + }, + { + "epoch": 5.515781062724731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13806 + }, + { + "epoch": 5.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13807 + }, + { + "epoch": 5.5165801038753495, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13808 + }, + { + "epoch": 5.516979624450659, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13809 + }, + { + "epoch": 5.517379145025969, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13810 + }, + { + "epoch": 5.517778665601279, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13811 + }, + { + "epoch": 5.518178186176588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13812 + }, + { + "epoch": 5.5185777067518975, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13813 + }, + { + "epoch": 5.518977227327207, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13814 + }, + { + "epoch": 5.519376747902517, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13815 + }, + { + "epoch": 5.519776268477827, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13816 + }, + { + "epoch": 5.520175789053136, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13817 + }, + { + "epoch": 5.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13818 + }, + { + "epoch": 5.520974830203755, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13819 + }, + { + "epoch": 5.521374350779065, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13820 + }, + { + "epoch": 5.521773871354375, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13821 + }, + { + "epoch": 5.522173391929685, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13822 + }, + { + "epoch": 5.5225729125049945, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13823 + }, + { + "epoch": 5.522972433080303, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13824 + }, + { + "epoch": 5.523371953655613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13825 + }, + { + "epoch": 5.523771474230923, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13826 + }, + { + "epoch": 5.524170994806233, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13827 + }, + { + "epoch": 5.5245705153815425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13828 + }, + { + "epoch": 5.524970035956851, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13829 + }, + { + "epoch": 5.525369556532161, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13830 + }, + { + "epoch": 5.525769077107471, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13831 + }, + { + "epoch": 5.526168597682781, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13832 + }, + { + "epoch": 5.526568118258091, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13833 + }, + { + "epoch": 5.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13834 + }, + { + "epoch": 5.527367159408709, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13835 + }, + { + "epoch": 5.527766679984019, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13836 + }, + { + "epoch": 5.528166200559329, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13837 + }, + { + "epoch": 5.528565721134639, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13838 + }, + { + "epoch": 5.528965241709948, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13839 + }, + { + "epoch": 5.529364762285257, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13840 + }, + { + "epoch": 5.529764282860567, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13841 + }, + { + "epoch": 5.530163803435877, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13842 + }, + { + "epoch": 5.530563324011187, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13843 + }, + { + "epoch": 5.5309628445864965, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13844 + }, + { + "epoch": 5.531362365161806, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13845 + }, + { + "epoch": 5.531761885737115, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13846 + }, + { + "epoch": 5.532161406312425, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13847 + }, + { + "epoch": 5.532560926887735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13848 + }, + { + "epoch": 5.5329604474630445, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13849 + }, + { + "epoch": 5.533359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13850 + }, + { + "epoch": 5.533759488613663, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13851 + }, + { + "epoch": 5.534159009188973, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 13852 + }, + { + "epoch": 5.534558529764283, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13853 + }, + { + "epoch": 5.5349580503395925, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13854 + }, + { + "epoch": 5.535357570914902, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13855 + }, + { + "epoch": 5.535757091490212, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13856 + }, + { + "epoch": 5.536156612065521, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13857 + }, + { + "epoch": 5.536556132640831, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13858 + }, + { + "epoch": 5.536955653216141, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13859 + }, + { + "epoch": 5.53735517379145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 13860 + }, + { + "epoch": 5.53775469436676, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 13861 + }, + { + "epoch": 5.53815421494207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13862 + }, + { + "epoch": 5.538553735517379, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 13863 + }, + { + "epoch": 5.538953256092689, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13864 + }, + { + "epoch": 5.539352776667998, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13865 + }, + { + "epoch": 5.539752297243308, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13866 + }, + { + "epoch": 5.540151817818618, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 13867 + }, + { + "epoch": 5.540551338393927, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13868 + }, + { + "epoch": 5.540950858969237, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13869 + }, + { + "epoch": 5.5413503795445465, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13870 + }, + { + "epoch": 5.541749900119856, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13871 + }, + { + "epoch": 5.542149420695166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13872 + }, + { + "epoch": 5.542548941270476, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13873 + }, + { + "epoch": 5.542948461845785, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13874 + }, + { + "epoch": 5.5433479824210945, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13875 + }, + { + "epoch": 5.543747502996404, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 13876 + }, + { + "epoch": 5.544147023571714, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13877 + }, + { + "epoch": 5.544546544147024, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13878 + }, + { + "epoch": 5.544946064722334, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13879 + }, + { + "epoch": 5.5453455852976425, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13880 + }, + { + "epoch": 5.545745105872952, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13881 + }, + { + "epoch": 5.546144626448262, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13882 + }, + { + "epoch": 5.546544147023572, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13883 + }, + { + "epoch": 5.546943667598882, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13884 + }, + { + "epoch": 5.547343188174191, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13885 + }, + { + "epoch": 5.5477427087495, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13886 + }, + { + "epoch": 5.54814222932481, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 13887 + }, + { + "epoch": 5.54854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13888 + }, + { + "epoch": 5.54894127047543, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13889 + }, + { + "epoch": 5.549340791050739, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13890 + }, + { + "epoch": 5.549740311626048, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13891 + }, + { + "epoch": 5.550139832201358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13892 + }, + { + "epoch": 5.550539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13893 + }, + { + "epoch": 5.550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13894 + }, + { + "epoch": 5.551338393927288, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13895 + }, + { + "epoch": 5.551737914502597, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13896 + }, + { + "epoch": 5.552137435077906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13897 + }, + { + "epoch": 5.552536955653216, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13898 + }, + { + "epoch": 5.552936476228526, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13899 + }, + { + "epoch": 5.553335996803836, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13900 + }, + { + "epoch": 5.553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13901 + }, + { + "epoch": 5.554135037954454, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13902 + }, + { + "epoch": 5.554534558529764, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13903 + }, + { + "epoch": 5.554934079105074, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13904 + }, + { + "epoch": 5.555333599680384, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13905 + }, + { + "epoch": 5.555733120255693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13906 + }, + { + "epoch": 5.556132640831002, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13907 + }, + { + "epoch": 5.556532161406312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13908 + }, + { + "epoch": 5.556931681981622, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13909 + }, + { + "epoch": 5.557331202556932, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13910 + }, + { + "epoch": 5.5577307231322415, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13911 + }, + { + "epoch": 5.558130243707551, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13912 + }, + { + "epoch": 5.558529764282861, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13913 + }, + { + "epoch": 5.55892928485817, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13914 + }, + { + "epoch": 5.55932880543348, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13915 + }, + { + "epoch": 5.5597283260087895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13916 + }, + { + "epoch": 5.560127846584099, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13917 + }, + { + "epoch": 5.560527367159409, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13918 + }, + { + "epoch": 5.560926887734718, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13919 + }, + { + "epoch": 5.561326408310028, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13920 + }, + { + "epoch": 5.561725928885338, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13921 + }, + { + "epoch": 5.562125449460647, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13922 + }, + { + "epoch": 5.562524970035957, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13923 + }, + { + "epoch": 5.562924490611266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13924 + }, + { + "epoch": 5.563324011186576, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13925 + }, + { + "epoch": 5.563723531761886, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13926 + }, + { + "epoch": 5.564123052337195, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 13927 + }, + { + "epoch": 5.564522572912505, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13928 + }, + { + "epoch": 5.564922093487815, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13929 + }, + { + "epoch": 5.565321614063124, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13930 + }, + { + "epoch": 5.565721134638434, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13931 + }, + { + "epoch": 5.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13932 + }, + { + "epoch": 5.566520175789053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13933 + }, + { + "epoch": 5.566919696364363, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13934 + }, + { + "epoch": 5.567319216939673, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13935 + }, + { + "epoch": 5.567718737514982, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13936 + }, + { + "epoch": 5.5681182580902915, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13937 + }, + { + "epoch": 5.568517778665601, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13938 + }, + { + "epoch": 5.568917299240911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13939 + }, + { + "epoch": 5.569316819816221, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13940 + }, + { + "epoch": 5.56971634039153, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13941 + }, + { + "epoch": 5.5701158609668395, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13942 + }, + { + "epoch": 5.570515381542149, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13943 + }, + { + "epoch": 5.570914902117459, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13944 + }, + { + "epoch": 5.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13945 + }, + { + "epoch": 5.571713943268079, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13946 + }, + { + "epoch": 5.572113463843388, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13947 + }, + { + "epoch": 5.572512984418697, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13948 + }, + { + "epoch": 5.572912504994007, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13949 + }, + { + "epoch": 5.573312025569317, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13950 + }, + { + "epoch": 5.573711546144627, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13951 + }, + { + "epoch": 5.5741110667199365, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13952 + }, + { + "epoch": 5.574510587295245, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13953 + }, + { + "epoch": 5.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13954 + }, + { + "epoch": 5.575309628445865, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13955 + }, + { + "epoch": 5.575709149021175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13956 + }, + { + "epoch": 5.5761086695964845, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13957 + }, + { + "epoch": 5.5765081901717934, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13958 + }, + { + "epoch": 5.576907710747103, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 13959 + }, + { + "epoch": 5.577307231322413, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13960 + }, + { + "epoch": 5.577706751897723, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13961 + }, + { + "epoch": 5.578106272473033, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13962 + }, + { + "epoch": 5.578505793048342, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13963 + }, + { + "epoch": 5.578905313623651, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13964 + }, + { + "epoch": 5.579304834198961, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13965 + }, + { + "epoch": 5.579704354774271, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13966 + }, + { + "epoch": 5.580103875349581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13967 + }, + { + "epoch": 5.58050339592489, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13968 + }, + { + "epoch": 5.5809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 13969 + }, + { + "epoch": 5.581302437075509, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13970 + }, + { + "epoch": 5.581701957650819, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13971 + }, + { + "epoch": 5.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13972 + }, + { + "epoch": 5.5825009988014385, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13973 + }, + { + "epoch": 5.582900519376748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13974 + }, + { + "epoch": 5.583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13975 + }, + { + "epoch": 5.583699560527367, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13976 + }, + { + "epoch": 5.584099081102677, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13977 + }, + { + "epoch": 5.5844986016779865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13978 + }, + { + "epoch": 5.584898122253296, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13979 + }, + { + "epoch": 5.585297642828605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13980 + }, + { + "epoch": 5.585697163403915, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13981 + }, + { + "epoch": 5.586096683979225, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13982 + }, + { + "epoch": 5.5864962045545346, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13983 + }, + { + "epoch": 5.586895725129844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13984 + }, + { + "epoch": 5.587295245705154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13985 + }, + { + "epoch": 5.587694766280464, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13986 + }, + { + "epoch": 5.588094286855773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13987 + }, + { + "epoch": 5.588493807431083, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 13988 + }, + { + "epoch": 5.588893328006392, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13989 + }, + { + "epoch": 5.589292848581702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13990 + }, + { + "epoch": 5.589692369157012, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13991 + }, + { + "epoch": 5.590091889732321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13992 + }, + { + "epoch": 5.590491410307631, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13993 + }, + { + "epoch": 5.59089093088294, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 13994 + }, + { + "epoch": 5.59129045145825, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13995 + }, + { + "epoch": 5.59168997203356, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13996 + }, + { + "epoch": 5.592089492608869, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13997 + }, + { + "epoch": 5.592489013184179, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13998 + }, + { + "epoch": 5.5928885337594885, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13999 + }, + { + "epoch": 5.593288054334798, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14000 + }, + { + "epoch": 5.593687574910108, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14001 + }, + { + "epoch": 5.594087095485418, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14002 + }, + { + "epoch": 5.594486616060728, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 14003 + }, + { + "epoch": 5.5948861366360365, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14004 + }, + { + "epoch": 5.595285657211346, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14005 + }, + { + "epoch": 5.595685177786656, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14006 + }, + { + "epoch": 5.596084698361966, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14007 + }, + { + "epoch": 5.596484218937276, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14008 + }, + { + "epoch": 5.5968837395125846, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14009 + }, + { + "epoch": 5.597283260087894, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.2923, + "step": 14010 + }, + { + "epoch": 5.597682780663204, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14011 + }, + { + "epoch": 5.598082301238514, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14012 + }, + { + "epoch": 5.598481821813824, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14013 + }, + { + "epoch": 5.598881342389133, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14014 + }, + { + "epoch": 5.599280862964442, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14015 + }, + { + "epoch": 5.599680383539752, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14016 + }, + { + "epoch": 5.600079904115062, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 14017 + }, + { + "epoch": 5.600479424690372, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14018 + }, + { + "epoch": 5.6008789452656815, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14019 + }, + { + "epoch": 5.60127846584099, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14020 + }, + { + "epoch": 5.6016779864163, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14021 + }, + { + "epoch": 5.60207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14022 + }, + { + "epoch": 5.60247702756692, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14023 + }, + { + "epoch": 5.60287654814223, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 14024 + }, + { + "epoch": 5.603276068717539, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14025 + }, + { + "epoch": 5.603675589292848, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14026 + }, + { + "epoch": 5.604075109868158, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14027 + }, + { + "epoch": 5.604474630443468, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14028 + }, + { + "epoch": 5.604874151018778, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14029 + }, + { + "epoch": 5.605273671594087, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 14030 + }, + { + "epoch": 5.605673192169396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14031 + }, + { + "epoch": 5.606072712744706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14032 + }, + { + "epoch": 5.606472233320016, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14033 + }, + { + "epoch": 5.606871753895326, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14034 + }, + { + "epoch": 5.6072712744706354, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14035 + }, + { + "epoch": 5.607670795045945, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14036 + }, + { + "epoch": 5.608070315621254, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14037 + }, + { + "epoch": 5.608469836196564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14038 + }, + { + "epoch": 5.608869356771874, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14039 + }, + { + "epoch": 5.6092688773471835, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14040 + }, + { + "epoch": 5.609668397922493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14041 + }, + { + "epoch": 5.610067918497803, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14042 + }, + { + "epoch": 5.610467439073112, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14043 + }, + { + "epoch": 5.610866959648422, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14044 + }, + { + "epoch": 5.6112664802237315, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14045 + }, + { + "epoch": 5.611666000799041, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14046 + }, + { + "epoch": 5.612065521374351, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14047 + }, + { + "epoch": 5.61246504194966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14048 + }, + { + "epoch": 5.61286456252497, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14049 + }, + { + "epoch": 5.61326408310028, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14050 + }, + { + "epoch": 5.613663603675589, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14051 + }, + { + "epoch": 5.614063124250899, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14052 + }, + { + "epoch": 5.614462644826209, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14053 + }, + { + "epoch": 5.614862165401518, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14054 + }, + { + "epoch": 5.615261685976828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14055 + }, + { + "epoch": 5.615661206552137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14056 + }, + { + "epoch": 5.616060727127447, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14057 + }, + { + "epoch": 5.616460247702757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14058 + }, + { + "epoch": 5.616859768278067, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14059 + }, + { + "epoch": 5.617259288853376, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14060 + }, + { + "epoch": 5.6176588094286855, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14061 + }, + { + "epoch": 5.618058330003995, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14062 + }, + { + "epoch": 5.618457850579305, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14063 + }, + { + "epoch": 5.618857371154615, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14064 + }, + { + "epoch": 5.619256891729924, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14065 + }, + { + "epoch": 5.6196564123052335, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14066 + }, + { + "epoch": 5.620055932880543, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14067 + }, + { + "epoch": 5.620455453455853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14068 + }, + { + "epoch": 5.620854974031163, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14069 + }, + { + "epoch": 5.621254494606472, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14070 + }, + { + "epoch": 5.6216540151817815, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14071 + }, + { + "epoch": 5.622053535757091, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14072 + }, + { + "epoch": 5.622453056332401, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14073 + }, + { + "epoch": 5.622852576907711, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14074 + }, + { + "epoch": 5.623252097483021, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14075 + }, + { + "epoch": 5.6236516180583305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14076 + }, + { + "epoch": 5.624051138633639, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14077 + }, + { + "epoch": 5.624450659208949, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14078 + }, + { + "epoch": 5.624850179784259, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14079 + }, + { + "epoch": 5.625249700359569, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.2954, + "step": 14080 + }, + { + "epoch": 5.6256492209348785, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14081 + }, + { + "epoch": 5.626048741510187, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14082 + }, + { + "epoch": 5.626448262085497, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14083 + }, + { + "epoch": 5.626847782660807, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14084 + }, + { + "epoch": 5.627247303236117, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 14085 + }, + { + "epoch": 5.627646823811427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14086 + }, + { + "epoch": 5.6280463443867355, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14087 + }, + { + "epoch": 5.628445864962045, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14088 + }, + { + "epoch": 5.628845385537355, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14089 + }, + { + "epoch": 5.629244906112665, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14090 + }, + { + "epoch": 5.629644426687975, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14091 + }, + { + "epoch": 5.630043947263284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14092 + }, + { + "epoch": 5.630443467838594, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14093 + }, + { + "epoch": 5.630842988413903, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14094 + }, + { + "epoch": 5.631242508989213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14095 + }, + { + "epoch": 5.631642029564523, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14096 + }, + { + "epoch": 5.632041550139832, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14097 + }, + { + "epoch": 5.632441070715142, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14098 + }, + { + "epoch": 5.632840591290451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14099 + }, + { + "epoch": 5.633240111865761, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14100 + }, + { + "epoch": 5.633639632441071, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14101 + }, + { + "epoch": 5.6340391530163805, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14102 + }, + { + "epoch": 5.63443867359169, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14103 + }, + { + "epoch": 5.634838194166999, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14104 + }, + { + "epoch": 5.635237714742309, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14105 + }, + { + "epoch": 5.635637235317619, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14106 + }, + { + "epoch": 5.6360367558929285, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14107 + }, + { + "epoch": 5.636436276468238, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14108 + }, + { + "epoch": 5.636835797043548, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14109 + }, + { + "epoch": 5.637235317618857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 14110 + }, + { + "epoch": 5.637634838194167, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14111 + }, + { + "epoch": 5.638034358769477, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14112 + }, + { + "epoch": 5.638433879344786, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14113 + }, + { + "epoch": 5.638833399920096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14114 + }, + { + "epoch": 5.639232920495406, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14115 + }, + { + "epoch": 5.639632441070715, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14116 + }, + { + "epoch": 5.640031961646025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14117 + }, + { + "epoch": 5.640431482221334, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14118 + }, + { + "epoch": 5.640831002796644, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14119 + }, + { + "epoch": 5.641230523371954, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14120 + }, + { + "epoch": 5.641630043947263, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14121 + }, + { + "epoch": 5.642029564522573, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14122 + }, + { + "epoch": 5.642429085097882, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 14123 + }, + { + "epoch": 5.642828605673192, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14124 + }, + { + "epoch": 5.643228126248502, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14125 + }, + { + "epoch": 5.643627646823812, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14126 + }, + { + "epoch": 5.644027167399121, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14127 + }, + { + "epoch": 5.6444266879744305, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14128 + }, + { + "epoch": 5.64482620854974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14129 + }, + { + "epoch": 5.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14130 + }, + { + "epoch": 5.64562524970036, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14131 + }, + { + "epoch": 5.64602477027567, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14132 + }, + { + "epoch": 5.6464242908509785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14133 + }, + { + "epoch": 5.646823811426288, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14134 + }, + { + "epoch": 5.647223332001598, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14135 + }, + { + "epoch": 5.647622852576908, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14136 + }, + { + "epoch": 5.648022373152218, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14137 + }, + { + "epoch": 5.648421893727527, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14138 + }, + { + "epoch": 5.648821414302836, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14139 + }, + { + "epoch": 5.649220934878146, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14140 + }, + { + "epoch": 5.649620455453456, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14141 + }, + { + "epoch": 5.650019976028766, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14142 + }, + { + "epoch": 5.6504194966040755, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14143 + }, + { + "epoch": 5.650819017179384, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14144 + }, + { + "epoch": 5.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14145 + }, + { + "epoch": 5.651618058330004, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14146 + }, + { + "epoch": 5.652017578905314, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14147 + }, + { + "epoch": 5.6524170994806235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14148 + }, + { + "epoch": 5.652816620055933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14149 + }, + { + "epoch": 5.653216140631242, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 14150 + }, + { + "epoch": 5.653615661206552, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14151 + }, + { + "epoch": 5.654015181781862, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14152 + }, + { + "epoch": 5.654414702357172, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14153 + }, + { + "epoch": 5.654814222932481, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14154 + }, + { + "epoch": 5.65521374350779, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14155 + }, + { + "epoch": 5.6556132640831, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14156 + }, + { + "epoch": 5.65601278465841, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14157 + }, + { + "epoch": 5.65641230523372, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14158 + }, + { + "epoch": 5.656811825809029, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14159 + }, + { + "epoch": 5.657211346384338, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14160 + }, + { + "epoch": 5.657610866959648, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 14161 + }, + { + "epoch": 5.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14162 + }, + { + "epoch": 5.658409908110268, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14163 + }, + { + "epoch": 5.6588094286855775, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14164 + }, + { + "epoch": 5.659208949260887, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14165 + }, + { + "epoch": 5.659608469836197, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14166 + }, + { + "epoch": 5.660007990411506, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14167 + }, + { + "epoch": 5.660407510986816, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14168 + }, + { + "epoch": 5.6608070315621255, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14169 + }, + { + "epoch": 5.661206552137435, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14170 + }, + { + "epoch": 5.661606072712745, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14171 + }, + { + "epoch": 5.662005593288054, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14172 + }, + { + "epoch": 5.662405113863364, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14173 + }, + { + "epoch": 5.6628046344386735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14174 + }, + { + "epoch": 5.663204155013983, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14175 + }, + { + "epoch": 5.663603675589293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 14176 + }, + { + "epoch": 5.664003196164602, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 14177 + }, + { + "epoch": 5.664402716739912, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14178 + }, + { + "epoch": 5.664802237315222, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14179 + }, + { + "epoch": 5.665201757890531, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14180 + }, + { + "epoch": 5.665601278465841, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14181 + }, + { + "epoch": 5.666000799041151, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14182 + }, + { + "epoch": 5.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14183 + }, + { + "epoch": 5.66679984019177, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14184 + }, + { + "epoch": 5.667199360767079, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14185 + }, + { + "epoch": 5.667598881342389, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14186 + }, + { + "epoch": 5.667998401917699, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14187 + }, + { + "epoch": 5.668397922493009, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14188 + }, + { + "epoch": 5.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14189 + }, + { + "epoch": 5.6691969636436275, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14190 + }, + { + "epoch": 5.669596484218937, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14191 + }, + { + "epoch": 5.669996004794247, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14192 + }, + { + "epoch": 5.670395525369557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14193 + }, + { + "epoch": 5.670795045944866, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14194 + }, + { + "epoch": 5.6711945665201755, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14195 + }, + { + "epoch": 5.671594087095485, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14196 + }, + { + "epoch": 5.671993607670795, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14197 + }, + { + "epoch": 5.672393128246105, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14198 + }, + { + "epoch": 5.672792648821415, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14199 + }, + { + "epoch": 5.6731921693967235, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14200 + }, + { + "epoch": 5.673591689972033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14201 + }, + { + "epoch": 5.673991210547343, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14202 + }, + { + "epoch": 5.674390731122653, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14203 + }, + { + "epoch": 5.674790251697963, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14204 + }, + { + "epoch": 5.6751897722732725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 14205 + }, + { + "epoch": 5.675589292848581, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14206 + }, + { + "epoch": 5.675988813423891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14207 + }, + { + "epoch": 5.676388333999201, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14208 + }, + { + "epoch": 5.676787854574511, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14209 + }, + { + "epoch": 5.6771873751498205, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14210 + }, + { + "epoch": 5.677586895725129, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14211 + }, + { + "epoch": 5.677986416300439, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14212 + }, + { + "epoch": 5.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14213 + }, + { + "epoch": 5.678785457451059, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14214 + }, + { + "epoch": 5.679184978026369, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14215 + }, + { + "epoch": 5.679584498601678, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14216 + }, + { + "epoch": 5.679984019176987, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14217 + }, + { + "epoch": 5.680383539752297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14218 + }, + { + "epoch": 5.680783060327607, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14219 + }, + { + "epoch": 5.681182580902917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14220 + }, + { + "epoch": 5.681582101478226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14221 + }, + { + "epoch": 5.681981622053536, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14222 + }, + { + "epoch": 5.682381142628845, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14223 + }, + { + "epoch": 5.682780663204155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14224 + }, + { + "epoch": 5.683180183779465, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14225 + }, + { + "epoch": 5.683579704354774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 14226 + }, + { + "epoch": 5.683979224930084, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14227 + }, + { + "epoch": 5.684378745505393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14228 + }, + { + "epoch": 5.684778266080703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14229 + }, + { + "epoch": 5.685177786656013, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14230 + }, + { + "epoch": 5.6855773072313225, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14231 + }, + { + "epoch": 5.685976827806632, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14232 + }, + { + "epoch": 5.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14233 + }, + { + "epoch": 5.686775868957251, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14234 + }, + { + "epoch": 5.687175389532561, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14235 + }, + { + "epoch": 5.6875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14236 + }, + { + "epoch": 5.68797443068318, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14237 + }, + { + "epoch": 5.68837395125849, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14238 + }, + { + "epoch": 5.6887734718338, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14239 + }, + { + "epoch": 5.689172992409109, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14240 + }, + { + "epoch": 5.689572512984419, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14241 + }, + { + "epoch": 5.689972033559728, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 14242 + }, + { + "epoch": 5.690371554135038, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14243 + }, + { + "epoch": 5.690771074710348, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14244 + }, + { + "epoch": 5.691170595285657, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14245 + }, + { + "epoch": 5.691570115860967, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14246 + }, + { + "epoch": 5.691969636436276, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14247 + }, + { + "epoch": 5.692369157011586, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14248 + }, + { + "epoch": 5.692768677586896, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14249 + }, + { + "epoch": 5.693168198162205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14250 + }, + { + "epoch": 5.693567718737515, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 14251 + }, + { + "epoch": 5.6939672393128244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14252 + }, + { + "epoch": 5.694366759888134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14253 + }, + { + "epoch": 5.694766280463444, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14254 + }, + { + "epoch": 5.695165801038754, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14255 + }, + { + "epoch": 5.695565321614064, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14256 + }, + { + "epoch": 5.6959648421893725, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14257 + }, + { + "epoch": 5.696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14258 + }, + { + "epoch": 5.696763883339992, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14259 + }, + { + "epoch": 5.697163403915302, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14260 + }, + { + "epoch": 5.697562924490612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14261 + }, + { + "epoch": 5.6979624450659205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14262 + }, + { + "epoch": 5.69836196564123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14263 + }, + { + "epoch": 5.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14264 + }, + { + "epoch": 5.69916100679185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14265 + }, + { + "epoch": 5.69956052736716, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14266 + }, + { + "epoch": 5.699960047942469, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14267 + }, + { + "epoch": 5.700359568517778, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 14268 + }, + { + "epoch": 5.700759089093088, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14269 + }, + { + "epoch": 5.701158609668398, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14270 + }, + { + "epoch": 5.701558130243708, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14271 + }, + { + "epoch": 5.7019576508190175, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 14272 + }, + { + "epoch": 5.702357171394327, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14273 + }, + { + "epoch": 5.702756691969636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14274 + }, + { + "epoch": 5.703156212544946, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14275 + }, + { + "epoch": 5.703555733120256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14276 + }, + { + "epoch": 5.7039552536955656, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14277 + }, + { + "epoch": 5.704354774270875, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14278 + }, + { + "epoch": 5.704754294846184, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14279 + }, + { + "epoch": 5.705153815421494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14280 + }, + { + "epoch": 5.705553335996804, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14281 + }, + { + "epoch": 5.705952856572114, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14282 + }, + { + "epoch": 5.706352377147423, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14283 + }, + { + "epoch": 5.706751897722732, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14284 + }, + { + "epoch": 5.707151418298042, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14285 + }, + { + "epoch": 5.707550938873352, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14286 + }, + { + "epoch": 5.707950459448662, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14287 + }, + { + "epoch": 5.708349980023971, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14288 + }, + { + "epoch": 5.708749500599281, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14289 + }, + { + "epoch": 5.70914902117459, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 14290 + }, + { + "epoch": 5.7095485417499, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14291 + }, + { + "epoch": 5.70994806232521, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14292 + }, + { + "epoch": 5.7103475829005195, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14293 + }, + { + "epoch": 5.710747103475829, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14294 + }, + { + "epoch": 5.711146624051139, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14295 + }, + { + "epoch": 5.711546144626448, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14296 + }, + { + "epoch": 5.711945665201758, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 14297 + }, + { + "epoch": 5.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14298 + }, + { + "epoch": 5.712744706352377, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 14299 + }, + { + "epoch": 5.713144226927687, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14300 + }, + { + "epoch": 5.713543747502996, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14301 + }, + { + "epoch": 5.713943268078306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14302 + }, + { + "epoch": 5.714342788653616, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14303 + }, + { + "epoch": 5.714742309228925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14304 + }, + { + "epoch": 5.715141829804235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14305 + }, + { + "epoch": 5.715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14306 + }, + { + "epoch": 5.715940870954854, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14307 + }, + { + "epoch": 5.716340391530164, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14308 + }, + { + "epoch": 5.716739912105473, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14309 + }, + { + "epoch": 5.717139432680783, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14310 + }, + { + "epoch": 5.717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14311 + }, + { + "epoch": 5.717938473831403, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14312 + }, + { + "epoch": 5.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14313 + }, + { + "epoch": 5.718737514982021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14314 + }, + { + "epoch": 5.719137035557331, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14315 + }, + { + "epoch": 5.719536556132641, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14316 + }, + { + "epoch": 5.719936076707951, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 14317 + }, + { + "epoch": 5.72033559728326, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14318 + }, + { + "epoch": 5.7207351178585695, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14319 + }, + { + "epoch": 5.721134638433879, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14320 + }, + { + "epoch": 5.721534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14321 + }, + { + "epoch": 5.721933679584499, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14322 + }, + { + "epoch": 5.722333200159808, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14323 + }, + { + "epoch": 5.7227327207351175, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 14324 + }, + { + "epoch": 5.723132241310427, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14325 + }, + { + "epoch": 5.723531761885737, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14326 + }, + { + "epoch": 5.723931282461047, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14327 + }, + { + "epoch": 5.724330803036357, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14328 + }, + { + "epoch": 5.7247303236116664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14329 + }, + { + "epoch": 5.725129844186975, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14330 + }, + { + "epoch": 5.725529364762285, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14331 + }, + { + "epoch": 5.725928885337595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.301, + "step": 14332 + }, + { + "epoch": 5.726328405912905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14333 + }, + { + "epoch": 5.7267279264882145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14334 + }, + { + "epoch": 5.727127447063523, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14335 + }, + { + "epoch": 5.727526967638833, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14336 + }, + { + "epoch": 5.727926488214143, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14337 + }, + { + "epoch": 5.728326008789453, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14338 + }, + { + "epoch": 5.7287255293647625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14339 + }, + { + "epoch": 5.729125049940071, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14340 + }, + { + "epoch": 5.729524570515381, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14341 + }, + { + "epoch": 5.729924091090691, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14342 + }, + { + "epoch": 5.730323611666001, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14343 + }, + { + "epoch": 5.730723132241311, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14344 + }, + { + "epoch": 5.73112265281662, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14345 + }, + { + "epoch": 5.73152217339193, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14346 + }, + { + "epoch": 5.731921693967239, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14347 + }, + { + "epoch": 5.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14348 + }, + { + "epoch": 5.732720735117859, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14349 + }, + { + "epoch": 5.733120255693168, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14350 + }, + { + "epoch": 5.733519776268478, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14351 + }, + { + "epoch": 5.733919296843787, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 14352 + }, + { + "epoch": 5.734318817419097, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14353 + }, + { + "epoch": 5.734718337994407, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14354 + }, + { + "epoch": 5.7351178585697165, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14355 + }, + { + "epoch": 5.735517379145026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14356 + }, + { + "epoch": 5.735916899720335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14357 + }, + { + "epoch": 5.736316420295645, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14358 + }, + { + "epoch": 5.736715940870955, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14359 + }, + { + "epoch": 5.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14360 + }, + { + "epoch": 5.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14361 + }, + { + "epoch": 5.737914502596884, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14362 + }, + { + "epoch": 5.738314023172194, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14363 + }, + { + "epoch": 5.738713543747503, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14364 + }, + { + "epoch": 5.7391130643228125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14365 + }, + { + "epoch": 5.739512584898122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14366 + }, + { + "epoch": 5.739912105473432, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 14367 + }, + { + "epoch": 5.740311626048742, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14368 + }, + { + "epoch": 5.740711146624051, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14369 + }, + { + "epoch": 5.741110667199361, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14370 + }, + { + "epoch": 5.74151018777467, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14371 + }, + { + "epoch": 5.74190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14372 + }, + { + "epoch": 5.74230922892529, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14373 + }, + { + "epoch": 5.742708749500599, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14374 + }, + { + "epoch": 5.743108270075909, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 14375 + }, + { + "epoch": 5.743507790651218, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14376 + }, + { + "epoch": 5.743907311226528, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14377 + }, + { + "epoch": 5.744306831801838, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14378 + }, + { + "epoch": 5.744706352377148, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14379 + }, + { + "epoch": 5.745105872952457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14380 + }, + { + "epoch": 5.7455053935277665, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14381 + }, + { + "epoch": 5.745904914103076, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14382 + }, + { + "epoch": 5.746304434678386, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 14383 + }, + { + "epoch": 5.746703955253696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14384 + }, + { + "epoch": 5.747103475829006, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14385 + }, + { + "epoch": 5.7475029964043145, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14386 + }, + { + "epoch": 5.747902516979624, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14387 + }, + { + "epoch": 5.748302037554934, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14388 + }, + { + "epoch": 5.748701558130244, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14389 + }, + { + "epoch": 5.749101078705554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14390 + }, + { + "epoch": 5.7495005992808625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14391 + }, + { + "epoch": 5.749900119856172, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14392 + }, + { + "epoch": 5.750299640431482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14393 + }, + { + "epoch": 5.750699161006792, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14394 + }, + { + "epoch": 5.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14395 + }, + { + "epoch": 5.7514982021574115, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14396 + }, + { + "epoch": 5.75189772273272, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14397 + }, + { + "epoch": 5.75229724330803, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14398 + }, + { + "epoch": 5.75269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14399 + }, + { + "epoch": 5.75309628445865, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14400 + }, + { + "epoch": 5.7534958050339595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14401 + }, + { + "epoch": 5.753895325609269, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14402 + }, + { + "epoch": 5.754294846184578, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14403 + }, + { + "epoch": 5.754694366759888, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14404 + }, + { + "epoch": 5.755093887335198, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14405 + }, + { + "epoch": 5.755493407910508, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14406 + }, + { + "epoch": 5.755892928485817, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14407 + }, + { + "epoch": 5.756292449061126, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14408 + }, + { + "epoch": 5.756691969636436, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14409 + }, + { + "epoch": 5.757091490211746, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14410 + }, + { + "epoch": 5.757491010787056, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14411 + }, + { + "epoch": 5.757890531362365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14412 + }, + { + "epoch": 5.758290051937674, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14413 + }, + { + "epoch": 5.758689572512984, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14414 + }, + { + "epoch": 5.759089093088294, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14415 + }, + { + "epoch": 5.759488613663604, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14416 + }, + { + "epoch": 5.759888134238913, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14417 + }, + { + "epoch": 5.760287654814223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14418 + }, + { + "epoch": 5.760687175389533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14419 + }, + { + "epoch": 5.761086695964842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 14420 + }, + { + "epoch": 5.761486216540152, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14421 + }, + { + "epoch": 5.7618857371154615, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14422 + }, + { + "epoch": 5.762285257690771, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14423 + }, + { + "epoch": 5.762684778266081, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14424 + }, + { + "epoch": 5.76308429884139, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14425 + }, + { + "epoch": 5.7634838194167, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14426 + }, + { + "epoch": 5.7638833399920095, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14427 + }, + { + "epoch": 5.764282860567319, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14428 + }, + { + "epoch": 5.764682381142629, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14429 + }, + { + "epoch": 5.765081901717938, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14430 + }, + { + "epoch": 5.765481422293248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14431 + }, + { + "epoch": 5.765880942868558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14432 + }, + { + "epoch": 5.766280463443867, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14433 + }, + { + "epoch": 5.766679984019177, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14434 + }, + { + "epoch": 5.767079504594487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14435 + }, + { + "epoch": 5.767479025169797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14436 + }, + { + "epoch": 5.767878545745106, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14437 + }, + { + "epoch": 5.768278066320415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14438 + }, + { + "epoch": 5.768677586895725, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14439 + }, + { + "epoch": 5.769077107471035, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14440 + }, + { + "epoch": 5.769476628046345, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14441 + }, + { + "epoch": 5.769876148621654, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14442 + }, + { + "epoch": 5.770275669196963, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14443 + }, + { + "epoch": 5.770675189772273, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14444 + }, + { + "epoch": 5.771074710347583, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14445 + }, + { + "epoch": 5.771474230922893, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14446 + }, + { + "epoch": 5.771873751498202, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14447 + }, + { + "epoch": 5.7722732720735115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14448 + }, + { + "epoch": 5.772672792648821, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14449 + }, + { + "epoch": 5.773072313224131, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14450 + }, + { + "epoch": 5.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14451 + }, + { + "epoch": 5.773871354374751, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14452 + }, + { + "epoch": 5.77427087495006, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14453 + }, + { + "epoch": 5.774670395525369, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14454 + }, + { + "epoch": 5.775069916100679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14455 + }, + { + "epoch": 5.775469436675989, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14456 + }, + { + "epoch": 5.775868957251299, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14457 + }, + { + "epoch": 5.7762684778266085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14458 + }, + { + "epoch": 5.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14459 + }, + { + "epoch": 5.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14460 + }, + { + "epoch": 5.777467039552537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 14461 + }, + { + "epoch": 5.777866560127847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14462 + }, + { + "epoch": 5.7782660807031565, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14463 + }, + { + "epoch": 5.778665601278465, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14464 + }, + { + "epoch": 5.779065121853775, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 14465 + }, + { + "epoch": 5.779464642429085, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14466 + }, + { + "epoch": 5.779864163004395, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14467 + }, + { + "epoch": 5.7802636835797045, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14468 + }, + { + "epoch": 5.780663204155014, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14469 + }, + { + "epoch": 5.781062724730323, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14470 + }, + { + "epoch": 5.781462245305633, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14471 + }, + { + "epoch": 5.781861765880943, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14472 + }, + { + "epoch": 5.782261286456253, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14473 + }, + { + "epoch": 5.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14474 + }, + { + "epoch": 5.783060327606872, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14475 + }, + { + "epoch": 5.783459848182181, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14476 + }, + { + "epoch": 5.783859368757491, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14477 + }, + { + "epoch": 5.784258889332801, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14478 + }, + { + "epoch": 5.78465840990811, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14479 + }, + { + "epoch": 5.78505793048342, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14480 + }, + { + "epoch": 5.785457451058729, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14481 + }, + { + "epoch": 5.785856971634039, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 14482 + }, + { + "epoch": 5.786256492209349, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14483 + }, + { + "epoch": 5.7866560127846585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14484 + }, + { + "epoch": 5.787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 14485 + }, + { + "epoch": 5.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 14486 + }, + { + "epoch": 5.787854574510587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14487 + }, + { + "epoch": 5.788254095085897, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14488 + }, + { + "epoch": 5.7886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14489 + }, + { + "epoch": 5.789053136236516, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14490 + }, + { + "epoch": 5.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14491 + }, + { + "epoch": 5.789852177387136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14492 + }, + { + "epoch": 5.790251697962445, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14493 + }, + { + "epoch": 5.7906512185377546, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14494 + }, + { + "epoch": 5.791050739113064, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14495 + }, + { + "epoch": 5.791450259688374, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 14496 + }, + { + "epoch": 5.791849780263684, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14497 + }, + { + "epoch": 5.792249300838993, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14498 + }, + { + "epoch": 5.792648821414303, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14499 + }, + { + "epoch": 5.793048341989612, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 14500 + }, + { + "epoch": 5.793447862564922, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14501 + }, + { + "epoch": 5.793847383140232, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14502 + }, + { + "epoch": 5.794246903715541, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14503 + }, + { + "epoch": 5.794646424290851, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.2985, + "step": 14504 + }, + { + "epoch": 5.79504594486616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14505 + }, + { + "epoch": 5.79544546544147, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14506 + }, + { + "epoch": 5.79584498601678, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14507 + }, + { + "epoch": 5.79624450659209, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14508 + }, + { + "epoch": 5.7966440271674, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14509 + }, + { + "epoch": 5.7970435477427085, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14510 + }, + { + "epoch": 5.797443068318018, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14511 + }, + { + "epoch": 5.797842588893328, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14512 + }, + { + "epoch": 5.798242109468638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14513 + }, + { + "epoch": 5.798641630043948, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14514 + }, + { + "epoch": 5.7990411506192565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14515 + }, + { + "epoch": 5.799440671194566, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14516 + }, + { + "epoch": 5.799840191769876, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14517 + }, + { + "epoch": 5.800239712345186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14518 + }, + { + "epoch": 5.800639232920496, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14519 + }, + { + "epoch": 5.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14520 + }, + { + "epoch": 5.801438274071114, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14521 + }, + { + "epoch": 5.801837794646424, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14522 + }, + { + "epoch": 5.802237315221734, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14523 + }, + { + "epoch": 5.802636835797044, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14524 + }, + { + "epoch": 5.8030363563723535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14525 + }, + { + "epoch": 5.803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14526 + }, + { + "epoch": 5.803835397522972, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14527 + }, + { + "epoch": 5.804234918098282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14528 + }, + { + "epoch": 5.804634438673592, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14529 + }, + { + "epoch": 5.8050339592489015, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14530 + }, + { + "epoch": 5.805433479824211, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14531 + }, + { + "epoch": 5.80583300039952, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 14532 + }, + { + "epoch": 5.80623252097483, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14533 + }, + { + "epoch": 5.80663204155014, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14534 + }, + { + "epoch": 5.80703156212545, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14535 + }, + { + "epoch": 5.807431082700759, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 14536 + }, + { + "epoch": 5.807830603276068, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14537 + }, + { + "epoch": 5.808230123851378, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14538 + }, + { + "epoch": 5.808629644426688, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14539 + }, + { + "epoch": 5.809029165001998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14540 + }, + { + "epoch": 5.809428685577307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14541 + }, + { + "epoch": 5.809828206152617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14542 + }, + { + "epoch": 5.810227726727927, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14543 + }, + { + "epoch": 5.810627247303236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14544 + }, + { + "epoch": 5.811026767878546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14545 + }, + { + "epoch": 5.8114262884538554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14546 + }, + { + "epoch": 5.811825809029165, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14547 + }, + { + "epoch": 5.812225329604475, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14548 + }, + { + "epoch": 5.812624850179784, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14549 + }, + { + "epoch": 5.813024370755094, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14550 + }, + { + "epoch": 5.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 14551 + }, + { + "epoch": 5.813823411905713, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14552 + }, + { + "epoch": 5.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14553 + }, + { + "epoch": 5.814622453056332, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14554 + }, + { + "epoch": 5.815021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14555 + }, + { + "epoch": 5.8154214942069515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14556 + }, + { + "epoch": 5.815821014782261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14557 + }, + { + "epoch": 5.816220535357571, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14558 + }, + { + "epoch": 5.816620055932881, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14559 + }, + { + "epoch": 5.81701957650819, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14560 + }, + { + "epoch": 5.8174190970835, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14561 + }, + { + "epoch": 5.817818617658809, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14562 + }, + { + "epoch": 5.818218138234119, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14563 + }, + { + "epoch": 5.818617658809429, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 14564 + }, + { + "epoch": 5.819017179384739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14565 + }, + { + "epoch": 5.819416699960048, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14566 + }, + { + "epoch": 5.819816220535357, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14567 + }, + { + "epoch": 5.820215741110667, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14568 + }, + { + "epoch": 5.820615261685977, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14569 + }, + { + "epoch": 5.821014782261287, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14570 + }, + { + "epoch": 5.821414302836596, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14571 + }, + { + "epoch": 5.8218138234119055, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14572 + }, + { + "epoch": 5.822213343987215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14573 + }, + { + "epoch": 5.822612864562525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14574 + }, + { + "epoch": 5.823012385137835, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14575 + }, + { + "epoch": 5.823411905713145, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14576 + }, + { + "epoch": 5.8238114262884535, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14577 + }, + { + "epoch": 5.824210946863763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14578 + }, + { + "epoch": 5.824610467439073, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14579 + }, + { + "epoch": 5.825009988014383, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14580 + }, + { + "epoch": 5.825409508589693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14581 + }, + { + "epoch": 5.825809029165002, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14582 + }, + { + "epoch": 5.826208549740311, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14583 + }, + { + "epoch": 5.826608070315621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14584 + }, + { + "epoch": 5.827007590890931, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14585 + }, + { + "epoch": 5.827407111466241, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14586 + }, + { + "epoch": 5.8278066320415505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14587 + }, + { + "epoch": 5.828206152616859, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 14588 + }, + { + "epoch": 5.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14589 + }, + { + "epoch": 5.829005193767479, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14590 + }, + { + "epoch": 5.829404714342789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14591 + }, + { + "epoch": 5.8298042349180985, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14592 + }, + { + "epoch": 5.830203755493407, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14593 + }, + { + "epoch": 5.830603276068717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14594 + }, + { + "epoch": 5.831002796644027, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14595 + }, + { + "epoch": 5.831402317219337, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14596 + }, + { + "epoch": 5.831801837794647, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 14597 + }, + { + "epoch": 5.832201358369956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14598 + }, + { + "epoch": 5.832600878945266, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14599 + }, + { + "epoch": 5.833000399520575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 14600 + }, + { + "epoch": 5.833399920095885, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14601 + }, + { + "epoch": 5.833799440671195, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14602 + }, + { + "epoch": 5.834198961246504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14603 + }, + { + "epoch": 5.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14604 + }, + { + "epoch": 5.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14605 + }, + { + "epoch": 5.835397522972433, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 14606 + }, + { + "epoch": 5.835797043547743, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14607 + }, + { + "epoch": 5.836196564123052, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14608 + }, + { + "epoch": 5.836596084698362, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14609 + }, + { + "epoch": 5.836995605273671, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14610 + }, + { + "epoch": 5.837395125848981, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14611 + }, + { + "epoch": 5.837794646424291, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14612 + }, + { + "epoch": 5.8381941669996005, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14613 + }, + { + "epoch": 5.83859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14614 + }, + { + "epoch": 5.83899320815022, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14615 + }, + { + "epoch": 5.83939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14616 + }, + { + "epoch": 5.839792249300839, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14617 + }, + { + "epoch": 5.8401917698761485, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 14618 + }, + { + "epoch": 5.840591290451458, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14619 + }, + { + "epoch": 5.840990811026768, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14620 + }, + { + "epoch": 5.841390331602078, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14621 + }, + { + "epoch": 5.841789852177387, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14622 + }, + { + "epoch": 5.842189372752697, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14623 + }, + { + "epoch": 5.842588893328006, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14624 + }, + { + "epoch": 5.842988413903316, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14625 + }, + { + "epoch": 5.843387934478626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14626 + }, + { + "epoch": 5.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14627 + }, + { + "epoch": 5.844186975629245, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14628 + }, + { + "epoch": 5.844586496204554, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14629 + }, + { + "epoch": 5.844986016779864, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14630 + }, + { + "epoch": 5.845385537355174, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14631 + }, + { + "epoch": 5.845785057930484, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14632 + }, + { + "epoch": 5.8461845785057935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14633 + }, + { + "epoch": 5.846584099081102, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14634 + }, + { + "epoch": 5.846983619656412, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14635 + }, + { + "epoch": 5.847383140231722, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14636 + }, + { + "epoch": 5.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14637 + }, + { + "epoch": 5.848182181382342, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14638 + }, + { + "epoch": 5.8485817019576505, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14639 + }, + { + "epoch": 5.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14640 + }, + { + "epoch": 5.84938074310827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14641 + }, + { + "epoch": 5.84978026368358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 14642 + }, + { + "epoch": 5.85017978425889, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14643 + }, + { + "epoch": 5.8505793048341985, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14644 + }, + { + "epoch": 5.850978825409508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14645 + }, + { + "epoch": 5.851378345984818, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14646 + }, + { + "epoch": 5.851777866560128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14647 + }, + { + "epoch": 5.852177387135438, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14648 + }, + { + "epoch": 5.8525769077107475, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 14649 + }, + { + "epoch": 5.852976428286056, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14650 + }, + { + "epoch": 5.853375948861366, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 14651 + }, + { + "epoch": 5.853775469436676, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14652 + }, + { + "epoch": 5.854174990011986, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14653 + }, + { + "epoch": 5.8545745105872955, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14654 + }, + { + "epoch": 5.854974031162605, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14655 + }, + { + "epoch": 5.855373551737914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14656 + }, + { + "epoch": 5.855773072313224, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 14657 + }, + { + "epoch": 5.856172592888534, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14658 + }, + { + "epoch": 5.8565721134638435, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14659 + }, + { + "epoch": 5.856971634039153, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14660 + }, + { + "epoch": 5.857371154614462, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14661 + }, + { + "epoch": 5.857770675189772, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14662 + }, + { + "epoch": 5.858170195765082, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14663 + }, + { + "epoch": 5.858569716340392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14664 + }, + { + "epoch": 5.858969236915701, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14665 + }, + { + "epoch": 5.859368757491011, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14666 + }, + { + "epoch": 5.85976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14667 + }, + { + "epoch": 5.86016779864163, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14668 + }, + { + "epoch": 5.86056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14669 + }, + { + "epoch": 5.860966839792249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14670 + }, + { + "epoch": 5.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14671 + }, + { + "epoch": 5.861765880942869, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14672 + }, + { + "epoch": 5.862165401518178, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14673 + }, + { + "epoch": 5.862564922093488, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14674 + }, + { + "epoch": 5.8629644426687975, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14675 + }, + { + "epoch": 5.863363963244107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14676 + }, + { + "epoch": 5.863763483819417, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 14677 + }, + { + "epoch": 5.864163004394726, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14678 + }, + { + "epoch": 5.864562524970036, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14679 + }, + { + "epoch": 5.8649620455453455, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 14680 + }, + { + "epoch": 5.865361566120655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14681 + }, + { + "epoch": 5.865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14682 + }, + { + "epoch": 5.866160607271274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14683 + }, + { + "epoch": 5.866560127846584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14684 + }, + { + "epoch": 5.8669596484218935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14685 + }, + { + "epoch": 5.867359168997203, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14686 + }, + { + "epoch": 5.867758689572513, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14687 + }, + { + "epoch": 5.868158210147823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14688 + }, + { + "epoch": 5.868557730723133, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14689 + }, + { + "epoch": 5.868957251298442, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 14690 + }, + { + "epoch": 5.869356771873751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14691 + }, + { + "epoch": 5.869756292449061, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14692 + }, + { + "epoch": 5.870155813024371, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14693 + }, + { + "epoch": 5.870555333599681, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14694 + }, + { + "epoch": 5.87095485417499, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14695 + }, + { + "epoch": 5.871354374750299, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14696 + }, + { + "epoch": 5.871753895325609, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14697 + }, + { + "epoch": 5.872153415900919, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14698 + }, + { + "epoch": 5.872552936476229, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14699 + }, + { + "epoch": 5.872952457051538, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14700 + }, + { + "epoch": 5.8733519776268475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14701 + }, + { + "epoch": 5.873751498202157, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 14702 + }, + { + "epoch": 5.874151018777467, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 14703 + }, + { + "epoch": 5.874550539352777, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14704 + }, + { + "epoch": 5.874950059928087, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14705 + }, + { + "epoch": 5.875349580503396, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14706 + }, + { + "epoch": 5.875749101078705, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14707 + }, + { + "epoch": 5.876148621654015, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14708 + }, + { + "epoch": 5.876548142229325, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 14709 + }, + { + "epoch": 5.876947662804635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 14710 + }, + { + "epoch": 5.877347183379944, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14711 + }, + { + "epoch": 5.877746703955253, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14712 + }, + { + "epoch": 5.878146224530563, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14713 + }, + { + "epoch": 5.878545745105873, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14714 + }, + { + "epoch": 5.878945265681183, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14715 + }, + { + "epoch": 5.8793447862564925, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14716 + }, + { + "epoch": 5.879744306831801, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14717 + }, + { + "epoch": 5.880143827407111, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14718 + }, + { + "epoch": 5.880543347982421, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14719 + }, + { + "epoch": 5.880942868557731, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14720 + }, + { + "epoch": 5.8813423891330405, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14721 + }, + { + "epoch": 5.88174190970835, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14722 + }, + { + "epoch": 5.88214143028366, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 14723 + }, + { + "epoch": 5.882540950858969, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14724 + }, + { + "epoch": 5.882940471434279, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14725 + }, + { + "epoch": 5.883339992009589, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 14726 + }, + { + "epoch": 5.883739512584898, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14727 + }, + { + "epoch": 5.884139033160208, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14728 + }, + { + "epoch": 5.884538553735517, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14729 + }, + { + "epoch": 5.884938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14730 + }, + { + "epoch": 5.885337594886137, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14731 + }, + { + "epoch": 5.885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14732 + }, + { + "epoch": 5.886136636036756, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14733 + }, + { + "epoch": 5.886536156612065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 14734 + }, + { + "epoch": 5.886935677187375, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14735 + }, + { + "epoch": 5.887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14736 + }, + { + "epoch": 5.887734718337994, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14737 + }, + { + "epoch": 5.888134238913304, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14738 + }, + { + "epoch": 5.888533759488614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 14739 + }, + { + "epoch": 5.888933280063923, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14740 + }, + { + "epoch": 5.889332800639233, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 14741 + }, + { + "epoch": 5.8897323212145425, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14742 + }, + { + "epoch": 5.890131841789852, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 14743 + }, + { + "epoch": 5.890531362365162, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14744 + }, + { + "epoch": 5.890930882940472, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14745 + }, + { + "epoch": 5.891330403515781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14746 + }, + { + "epoch": 5.8917299240910905, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14747 + }, + { + "epoch": 5.8921294446664, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14748 + }, + { + "epoch": 5.89252896524171, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 14749 + }, + { + "epoch": 5.89292848581702, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14750 + }, + { + "epoch": 5.893328006392329, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14751 + }, + { + "epoch": 5.893727526967639, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14752 + }, + { + "epoch": 5.894127047542948, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14753 + }, + { + "epoch": 5.894526568118258, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14754 + }, + { + "epoch": 5.894926088693568, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14755 + }, + { + "epoch": 5.895325609268878, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14756 + }, + { + "epoch": 5.895725129844187, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14757 + }, + { + "epoch": 5.896124650419496, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 14758 + }, + { + "epoch": 5.896524170994806, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14759 + }, + { + "epoch": 5.896923691570116, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14760 + }, + { + "epoch": 5.897323212145426, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14761 + }, + { + "epoch": 5.8977227327207356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14762 + }, + { + "epoch": 5.8981222532960444, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14763 + }, + { + "epoch": 5.898521773871354, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14764 + }, + { + "epoch": 5.898921294446664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14765 + }, + { + "epoch": 5.899320815021974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14766 + }, + { + "epoch": 5.899720335597284, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14767 + }, + { + "epoch": 5.9001198561725925, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14768 + }, + { + "epoch": 5.900519376747902, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14769 + }, + { + "epoch": 5.900918897323212, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14770 + }, + { + "epoch": 5.901318417898522, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 14771 + }, + { + "epoch": 5.901717938473832, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14772 + }, + { + "epoch": 5.9021174590491405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14773 + }, + { + "epoch": 5.90251697962445, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14774 + }, + { + "epoch": 5.90291650019976, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14775 + }, + { + "epoch": 5.90331602077507, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 14776 + }, + { + "epoch": 5.90371554135038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14777 + }, + { + "epoch": 5.9041150619256895, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14778 + }, + { + "epoch": 5.904514582500999, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14779 + }, + { + "epoch": 5.904914103076308, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14780 + }, + { + "epoch": 5.905313623651618, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14781 + }, + { + "epoch": 5.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 14782 + }, + { + "epoch": 5.9061126648022375, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14783 + }, + { + "epoch": 5.906512185377547, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14784 + }, + { + "epoch": 5.906911705952856, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14785 + }, + { + "epoch": 5.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14786 + }, + { + "epoch": 5.907710747103476, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14787 + }, + { + "epoch": 5.9081102676787856, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14788 + }, + { + "epoch": 5.908509788254095, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14789 + }, + { + "epoch": 5.908909308829404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14790 + }, + { + "epoch": 5.909308829404714, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14791 + }, + { + "epoch": 5.909708349980024, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14792 + }, + { + "epoch": 5.910107870555334, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14793 + }, + { + "epoch": 5.910507391130643, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14794 + }, + { + "epoch": 5.910906911705953, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14795 + }, + { + "epoch": 5.911306432281263, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 14796 + }, + { + "epoch": 5.911705952856572, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14797 + }, + { + "epoch": 5.912105473431882, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14798 + }, + { + "epoch": 5.912504994007191, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14799 + }, + { + "epoch": 5.912904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 14800 + }, + { + "epoch": 5.913304035157811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14801 + }, + { + "epoch": 5.91370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14802 + }, + { + "epoch": 5.91410307630843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14803 + }, + { + "epoch": 5.9145025968837395, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14804 + }, + { + "epoch": 5.914902117459049, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 14805 + }, + { + "epoch": 5.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14806 + }, + { + "epoch": 5.915701158609668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14807 + }, + { + "epoch": 5.916100679184978, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14808 + }, + { + "epoch": 5.9165001997602875, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14809 + }, + { + "epoch": 5.916899720335597, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14810 + }, + { + "epoch": 5.917299240910907, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14811 + }, + { + "epoch": 5.917698761486217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14812 + }, + { + "epoch": 5.918098282061527, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14813 + }, + { + "epoch": 5.918497802636836, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14814 + }, + { + "epoch": 5.918897323212145, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14815 + }, + { + "epoch": 5.919296843787455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14816 + }, + { + "epoch": 5.919696364362765, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14817 + }, + { + "epoch": 5.920095884938075, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14818 + }, + { + "epoch": 5.920495405513384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14819 + }, + { + "epoch": 5.920894926088693, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14820 + }, + { + "epoch": 5.921294446664003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14821 + }, + { + "epoch": 5.921693967239313, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14822 + }, + { + "epoch": 5.922093487814623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14823 + }, + { + "epoch": 5.922493008389932, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14824 + }, + { + "epoch": 5.922892528965241, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14825 + }, + { + "epoch": 5.923292049540551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14826 + }, + { + "epoch": 5.923691570115861, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14827 + }, + { + "epoch": 5.924091090691171, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14828 + }, + { + "epoch": 5.924490611266481, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14829 + }, + { + "epoch": 5.9248901318417895, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14830 + }, + { + "epoch": 5.925289652417099, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14831 + }, + { + "epoch": 5.925689172992409, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14832 + }, + { + "epoch": 5.926088693567719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14833 + }, + { + "epoch": 5.926488214143029, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14834 + }, + { + "epoch": 5.926887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14835 + }, + { + "epoch": 5.927287255293647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14836 + }, + { + "epoch": 5.927686775868957, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 14837 + }, + { + "epoch": 5.928086296444267, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14838 + }, + { + "epoch": 5.928485817019577, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14839 + }, + { + "epoch": 5.9288853375948865, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14840 + }, + { + "epoch": 5.929284858170195, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 14841 + }, + { + "epoch": 5.929684378745505, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14842 + }, + { + "epoch": 5.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14843 + }, + { + "epoch": 5.930483419896125, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 14844 + }, + { + "epoch": 5.9308829404714345, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 14845 + }, + { + "epoch": 5.931282461046744, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14846 + }, + { + "epoch": 5.931681981622053, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14847 + }, + { + "epoch": 5.932081502197363, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14848 + }, + { + "epoch": 5.932481022772673, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14849 + }, + { + "epoch": 5.9328805433479825, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14850 + }, + { + "epoch": 5.933280063923292, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14851 + }, + { + "epoch": 5.933679584498602, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14852 + }, + { + "epoch": 5.934079105073911, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14853 + }, + { + "epoch": 5.934478625649221, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14854 + }, + { + "epoch": 5.934878146224531, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14855 + }, + { + "epoch": 5.93527766679984, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14856 + }, + { + "epoch": 5.93567718737515, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14857 + }, + { + "epoch": 5.936076707950459, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14858 + }, + { + "epoch": 5.936476228525769, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14859 + }, + { + "epoch": 5.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 14860 + }, + { + "epoch": 5.937275269676388, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 14861 + }, + { + "epoch": 5.937674790251698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14862 + }, + { + "epoch": 5.938074310827007, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14863 + }, + { + "epoch": 5.938473831402317, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14864 + }, + { + "epoch": 5.938873351977627, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14865 + }, + { + "epoch": 5.9392728725529365, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 14866 + }, + { + "epoch": 5.939672393128246, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14867 + }, + { + "epoch": 5.940071913703556, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14868 + }, + { + "epoch": 5.940471434278866, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14869 + }, + { + "epoch": 5.940870954854175, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14870 + }, + { + "epoch": 5.9412704754294845, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14871 + }, + { + "epoch": 5.941669996004794, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14872 + }, + { + "epoch": 5.942069516580104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14873 + }, + { + "epoch": 5.942469037155414, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14874 + }, + { + "epoch": 5.942868557730723, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14875 + }, + { + "epoch": 5.9432680783060325, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14876 + }, + { + "epoch": 5.943667598881342, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14877 + }, + { + "epoch": 5.944067119456652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14878 + }, + { + "epoch": 5.944466640031962, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14879 + }, + { + "epoch": 5.944866160607271, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 14880 + }, + { + "epoch": 5.945265681182581, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14881 + }, + { + "epoch": 5.94566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14882 + }, + { + "epoch": 5.9460647223332, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14883 + }, + { + "epoch": 5.94646424290851, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14884 + }, + { + "epoch": 5.94686376348382, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14885 + }, + { + "epoch": 5.9472632840591295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14886 + }, + { + "epoch": 5.947662804634438, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 14887 + }, + { + "epoch": 5.948062325209748, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14888 + }, + { + "epoch": 5.948461845785058, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14889 + }, + { + "epoch": 5.948861366360368, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14890 + }, + { + "epoch": 5.949260886935678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14891 + }, + { + "epoch": 5.9496604075109865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14892 + }, + { + "epoch": 5.950059928086296, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14893 + }, + { + "epoch": 5.950459448661606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14894 + }, + { + "epoch": 5.950858969236916, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14895 + }, + { + "epoch": 5.951258489812226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 14896 + }, + { + "epoch": 5.9516580103875345, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 14897 + }, + { + "epoch": 5.952057530962844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14898 + }, + { + "epoch": 5.952457051538154, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14899 + }, + { + "epoch": 5.952856572113464, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14900 + }, + { + "epoch": 5.953256092688774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14901 + }, + { + "epoch": 5.953655613264083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14902 + }, + { + "epoch": 5.954055133839393, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14903 + }, + { + "epoch": 5.954454654414702, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14904 + }, + { + "epoch": 5.954854174990012, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14905 + }, + { + "epoch": 5.955253695565322, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14906 + }, + { + "epoch": 5.9556532161406315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14907 + }, + { + "epoch": 5.956052736715941, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14908 + }, + { + "epoch": 5.95645225729125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14909 + }, + { + "epoch": 5.95685177786656, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14910 + }, + { + "epoch": 5.95725129844187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14911 + }, + { + "epoch": 5.9576508190171795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14912 + }, + { + "epoch": 5.958050339592489, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14913 + }, + { + "epoch": 5.958449860167798, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14914 + }, + { + "epoch": 5.958849380743108, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14915 + }, + { + "epoch": 5.959248901318418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14916 + }, + { + "epoch": 5.959648421893728, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14917 + }, + { + "epoch": 5.960047942469037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14918 + }, + { + "epoch": 5.960447463044347, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14919 + }, + { + "epoch": 5.960846983619656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14920 + }, + { + "epoch": 5.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14921 + }, + { + "epoch": 5.961646024770276, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14922 + }, + { + "epoch": 5.962045545345585, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14923 + }, + { + "epoch": 5.962445065920895, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14924 + }, + { + "epoch": 5.962844586496205, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14925 + }, + { + "epoch": 5.963244107071514, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14926 + }, + { + "epoch": 5.963643627646824, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14927 + }, + { + "epoch": 5.964043148222133, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14928 + }, + { + "epoch": 5.964442668797443, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14929 + }, + { + "epoch": 5.964842189372753, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14930 + }, + { + "epoch": 5.965241709948062, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14931 + }, + { + "epoch": 5.965641230523372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14932 + }, + { + "epoch": 5.9660407510986815, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14933 + }, + { + "epoch": 5.966440271673991, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14934 + }, + { + "epoch": 5.966839792249301, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14935 + }, + { + "epoch": 5.967239312824611, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14936 + }, + { + "epoch": 5.96763883339992, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14937 + }, + { + "epoch": 5.9680383539752295, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14938 + }, + { + "epoch": 5.968437874550539, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14939 + }, + { + "epoch": 5.968837395125849, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14940 + }, + { + "epoch": 5.969236915701159, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14941 + }, + { + "epoch": 5.969636436276469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14942 + }, + { + "epoch": 5.970035956851778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14943 + }, + { + "epoch": 5.970435477427087, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14944 + }, + { + "epoch": 5.970834998002397, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 14945 + }, + { + "epoch": 5.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14946 + }, + { + "epoch": 5.971634039153017, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14947 + }, + { + "epoch": 5.972033559728326, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14948 + }, + { + "epoch": 5.972433080303635, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14949 + }, + { + "epoch": 5.972832600878945, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14950 + }, + { + "epoch": 5.973232121454255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14951 + }, + { + "epoch": 5.973631642029565, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14952 + }, + { + "epoch": 5.974031162604874, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14953 + }, + { + "epoch": 5.974430683180183, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14954 + }, + { + "epoch": 5.974830203755493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14955 + }, + { + "epoch": 5.975229724330803, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14956 + }, + { + "epoch": 5.975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14957 + }, + { + "epoch": 5.976028765481423, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14958 + }, + { + "epoch": 5.976428286056732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14959 + }, + { + "epoch": 5.976827806632041, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 14960 + }, + { + "epoch": 5.977227327207351, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14961 + }, + { + "epoch": 5.977626847782661, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14962 + }, + { + "epoch": 5.978026368357971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14963 + }, + { + "epoch": 5.97842588893328, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14964 + }, + { + "epoch": 5.978825409508589, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14965 + }, + { + "epoch": 5.979224930083899, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14966 + }, + { + "epoch": 5.979624450659209, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14967 + }, + { + "epoch": 5.980023971234519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14968 + }, + { + "epoch": 5.9804234918098285, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14969 + }, + { + "epoch": 5.980823012385137, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14970 + }, + { + "epoch": 5.981222532960447, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14971 + }, + { + "epoch": 5.981622053535757, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 14972 + }, + { + "epoch": 5.982021574111067, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14973 + }, + { + "epoch": 5.9824210946863765, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14974 + }, + { + "epoch": 5.982820615261686, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14975 + }, + { + "epoch": 5.983220135836996, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14976 + }, + { + "epoch": 5.983619656412305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14977 + }, + { + "epoch": 5.984019176987615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14978 + }, + { + "epoch": 5.9844186975629245, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14979 + }, + { + "epoch": 5.984818218138234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14980 + }, + { + "epoch": 5.985217738713544, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14981 + }, + { + "epoch": 5.985617259288853, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14982 + }, + { + "epoch": 5.986016779864163, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14983 + }, + { + "epoch": 5.986416300439473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14984 + }, + { + "epoch": 5.986815821014782, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14985 + }, + { + "epoch": 5.987215341590092, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14986 + }, + { + "epoch": 5.987614862165401, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14987 + }, + { + "epoch": 5.988014382740711, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14988 + }, + { + "epoch": 5.988413903316021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14989 + }, + { + "epoch": 5.98881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14990 + }, + { + "epoch": 5.98921294446664, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14991 + }, + { + "epoch": 5.98961246504195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14992 + }, + { + "epoch": 5.99001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14993 + }, + { + "epoch": 5.990411506192569, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14994 + }, + { + "epoch": 5.9908110267678785, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14995 + }, + { + "epoch": 5.991210547343188, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14996 + }, + { + "epoch": 5.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14997 + }, + { + "epoch": 5.992009588493808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14998 + }, + { + "epoch": 5.992409109069117, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14999 + }, + { + "epoch": 5.9928086296444265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15000 + }, + { + "epoch": 5.993208150219736, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15001 + }, + { + "epoch": 5.993607670795046, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 15002 + }, + { + "epoch": 5.994007191370356, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15003 + }, + { + "epoch": 5.994406711945665, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15004 + }, + { + "epoch": 5.9948062325209746, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15005 + }, + { + "epoch": 5.995205753096284, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 15006 + }, + { + "epoch": 5.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15007 + }, + { + "epoch": 5.996004794246904, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 15008 + }, + { + "epoch": 5.996404314822214, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15009 + }, + { + "epoch": 5.996803835397523, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15010 + }, + { + "epoch": 5.997203355972832, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15011 + }, + { + "epoch": 5.997602876548142, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15012 + }, + { + "epoch": 5.998002397123452, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15013 + }, + { + "epoch": 5.998401917698762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15014 + }, + { + "epoch": 5.9988014382740715, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15015 + }, + { + "epoch": 5.99920095884938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15016 + }, + { + "epoch": 5.99960047942469, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 15017 + }, + { + "epoch": 6.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15018 + }, + { + "epoch": 6.00039952057531, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15019 + }, + { + "epoch": 6.00079904115062, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15020 + }, + { + "epoch": 6.0011985617259285, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15021 + }, + { + "epoch": 6.001598082301238, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15022 + }, + { + "epoch": 6.001997602876548, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15023 + }, + { + "epoch": 6.002397123451858, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15024 + }, + { + "epoch": 6.002796644027168, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 15025 + }, + { + "epoch": 6.003196164602477, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15026 + }, + { + "epoch": 6.003595685177786, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15027 + }, + { + "epoch": 6.003995205753096, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15028 + }, + { + "epoch": 6.004394726328406, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15029 + }, + { + "epoch": 6.004794246903716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15030 + }, + { + "epoch": 6.0051937674790254, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15031 + }, + { + "epoch": 6.005593288054335, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15032 + }, + { + "epoch": 6.005992808629644, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15033 + }, + { + "epoch": 6.006392329204954, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15034 + }, + { + "epoch": 6.006791849780264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15035 + }, + { + "epoch": 6.0071913703555735, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 15036 + }, + { + "epoch": 6.007590890930883, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15037 + }, + { + "epoch": 6.007990411506192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15038 + }, + { + "epoch": 6.008389932081502, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15039 + }, + { + "epoch": 6.008789452656812, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15040 + }, + { + "epoch": 6.0091889732321215, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15041 + }, + { + "epoch": 6.009588493807431, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15042 + }, + { + "epoch": 6.009988014382741, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15043 + }, + { + "epoch": 6.01038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15044 + }, + { + "epoch": 6.01078705553336, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15045 + }, + { + "epoch": 6.01118657610867, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15046 + }, + { + "epoch": 6.011586096683979, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15047 + }, + { + "epoch": 6.011985617259289, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15048 + }, + { + "epoch": 6.012385137834598, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15049 + }, + { + "epoch": 6.012784658409908, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15050 + }, + { + "epoch": 6.013184178985218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15051 + }, + { + "epoch": 6.013583699560527, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15052 + }, + { + "epoch": 6.013983220135837, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15053 + }, + { + "epoch": 6.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15054 + }, + { + "epoch": 6.014782261286456, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15055 + }, + { + "epoch": 6.015181781861766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15056 + }, + { + "epoch": 6.0155813024370755, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15057 + }, + { + "epoch": 6.015980823012385, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 15058 + }, + { + "epoch": 6.016380343587695, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15059 + }, + { + "epoch": 6.016779864163005, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 15060 + }, + { + "epoch": 6.017179384738314, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15061 + }, + { + "epoch": 6.0175789053136235, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15062 + }, + { + "epoch": 6.017978425888933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15063 + }, + { + "epoch": 6.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15064 + }, + { + "epoch": 6.018777467039553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15065 + }, + { + "epoch": 6.019176987614862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15066 + }, + { + "epoch": 6.0195765081901715, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15067 + }, + { + "epoch": 6.019976028765481, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15068 + }, + { + "epoch": 6.020375549340791, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15069 + }, + { + "epoch": 6.020775069916101, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 15070 + }, + { + "epoch": 6.021174590491411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15071 + }, + { + "epoch": 6.02157411106672, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15072 + }, + { + "epoch": 6.021973631642029, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15073 + }, + { + "epoch": 6.022373152217339, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15074 + }, + { + "epoch": 6.022772672792649, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 15075 + }, + { + "epoch": 6.023172193367959, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15076 + }, + { + "epoch": 6.0235717139432685, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15077 + }, + { + "epoch": 6.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15078 + }, + { + "epoch": 6.024370755093887, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15079 + }, + { + "epoch": 6.024770275669197, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15080 + }, + { + "epoch": 6.025169796244507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15081 + }, + { + "epoch": 6.025569316819817, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15082 + }, + { + "epoch": 6.0259688373951255, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15083 + }, + { + "epoch": 6.026368357970435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15084 + }, + { + "epoch": 6.026767878545745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15085 + }, + { + "epoch": 6.027167399121055, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15086 + }, + { + "epoch": 6.027566919696365, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15087 + }, + { + "epoch": 6.027966440271674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15088 + }, + { + "epoch": 6.028365960846983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15089 + }, + { + "epoch": 6.028765481422293, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15090 + }, + { + "epoch": 6.029165001997603, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15091 + }, + { + "epoch": 6.029564522572913, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15092 + }, + { + "epoch": 6.029964043148222, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15093 + }, + { + "epoch": 6.030363563723531, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15094 + }, + { + "epoch": 6.030763084298841, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15095 + }, + { + "epoch": 6.031162604874151, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15096 + }, + { + "epoch": 6.031562125449461, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15097 + }, + { + "epoch": 6.0319616460247705, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15098 + }, + { + "epoch": 6.03236116660008, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15099 + }, + { + "epoch": 6.032760687175389, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15100 + }, + { + "epoch": 6.033160207750699, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15101 + }, + { + "epoch": 6.033559728326009, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15102 + }, + { + "epoch": 6.0339592489013185, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15103 + }, + { + "epoch": 6.034358769476628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15104 + }, + { + "epoch": 6.034758290051938, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15105 + }, + { + "epoch": 6.035157810627247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15106 + }, + { + "epoch": 6.035557331202557, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15107 + }, + { + "epoch": 6.035956851777867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 15108 + }, + { + "epoch": 6.036356372353176, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15109 + }, + { + "epoch": 6.036755892928486, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15110 + }, + { + "epoch": 6.037155413503795, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15111 + }, + { + "epoch": 6.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15112 + }, + { + "epoch": 6.037954454654415, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15113 + }, + { + "epoch": 6.038353975229724, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15114 + }, + { + "epoch": 6.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15115 + }, + { + "epoch": 6.039153016380344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15116 + }, + { + "epoch": 6.039552536955653, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15117 + }, + { + "epoch": 6.039952057530963, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15118 + }, + { + "epoch": 6.040351578106272, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15119 + }, + { + "epoch": 6.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15120 + }, + { + "epoch": 6.041150619256892, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15121 + }, + { + "epoch": 6.041550139832202, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15122 + }, + { + "epoch": 6.041949660407511, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15123 + }, + { + "epoch": 6.0423491809828205, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 15124 + }, + { + "epoch": 6.04274870155813, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15125 + }, + { + "epoch": 6.04314822213344, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15126 + }, + { + "epoch": 6.04354774270875, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15127 + }, + { + "epoch": 6.043947263284059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15128 + }, + { + "epoch": 6.0443467838593685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15129 + }, + { + "epoch": 6.044746304434678, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15130 + }, + { + "epoch": 6.045145825009988, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15131 + }, + { + "epoch": 6.045545345585298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15132 + }, + { + "epoch": 6.045944866160608, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15133 + }, + { + "epoch": 6.046344386735917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15134 + }, + { + "epoch": 6.046743907311226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15135 + }, + { + "epoch": 6.047143427886536, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15136 + }, + { + "epoch": 6.047542948461846, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15137 + }, + { + "epoch": 6.047942469037156, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15138 + }, + { + "epoch": 6.048341989612465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15139 + }, + { + "epoch": 6.048741510187774, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15140 + }, + { + "epoch": 6.049141030763084, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15141 + }, + { + "epoch": 6.049540551338394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15142 + }, + { + "epoch": 6.049940071913704, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15143 + }, + { + "epoch": 6.0503395924890135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15144 + }, + { + "epoch": 6.050739113064322, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15145 + }, + { + "epoch": 6.051138633639632, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15146 + }, + { + "epoch": 6.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15147 + }, + { + "epoch": 6.051937674790252, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15148 + }, + { + "epoch": 6.052337195365562, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15149 + }, + { + "epoch": 6.052736715940871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15150 + }, + { + "epoch": 6.05313623651618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15151 + }, + { + "epoch": 6.05353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15152 + }, + { + "epoch": 6.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 15153 + }, + { + "epoch": 6.05433479824211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 15154 + }, + { + "epoch": 6.054734318817419, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15155 + }, + { + "epoch": 6.055133839392728, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15156 + }, + { + "epoch": 6.055533359968038, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15157 + }, + { + "epoch": 6.055932880543348, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15158 + }, + { + "epoch": 6.056332401118658, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15159 + }, + { + "epoch": 6.0567319216939675, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15160 + }, + { + "epoch": 6.057131442269277, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15161 + }, + { + "epoch": 6.057530962844586, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15162 + }, + { + "epoch": 6.057930483419896, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15163 + }, + { + "epoch": 6.058330003995206, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15164 + }, + { + "epoch": 6.0587295245705155, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15165 + }, + { + "epoch": 6.059129045145825, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15166 + }, + { + "epoch": 6.059528565721135, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15167 + }, + { + "epoch": 6.059928086296444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15168 + }, + { + "epoch": 6.060327606871754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15169 + }, + { + "epoch": 6.0607271274470635, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15170 + }, + { + "epoch": 6.061126648022373, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15171 + }, + { + "epoch": 6.061526168597683, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15172 + }, + { + "epoch": 6.061925689172992, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15173 + }, + { + "epoch": 6.062325209748302, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15174 + }, + { + "epoch": 6.062724730323612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 15175 + }, + { + "epoch": 6.063124250898921, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15176 + }, + { + "epoch": 6.063523771474231, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15177 + }, + { + "epoch": 6.063923292049541, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15178 + }, + { + "epoch": 6.06432281262485, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 15179 + }, + { + "epoch": 6.06472233320016, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15180 + }, + { + "epoch": 6.065121853775469, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15181 + }, + { + "epoch": 6.065521374350779, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15182 + }, + { + "epoch": 6.065920894926089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15183 + }, + { + "epoch": 6.066320415501398, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15184 + }, + { + "epoch": 6.066719936076708, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15185 + }, + { + "epoch": 6.0671194566520175, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15186 + }, + { + "epoch": 6.067518977227327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15187 + }, + { + "epoch": 6.067918497802637, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15188 + }, + { + "epoch": 6.068318018377947, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15189 + }, + { + "epoch": 6.068717538953256, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15190 + }, + { + "epoch": 6.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15191 + }, + { + "epoch": 6.069516580103875, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15192 + }, + { + "epoch": 6.069916100679185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15193 + }, + { + "epoch": 6.070315621254495, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15194 + }, + { + "epoch": 6.070715141829805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15195 + }, + { + "epoch": 6.0711146624051135, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15196 + }, + { + "epoch": 6.071514182980423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15197 + }, + { + "epoch": 6.071913703555733, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15198 + }, + { + "epoch": 6.072313224131043, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15199 + }, + { + "epoch": 6.072712744706353, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15200 + }, + { + "epoch": 6.073112265281662, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15201 + }, + { + "epoch": 6.073511785856971, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15202 + }, + { + "epoch": 6.073911306432281, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15203 + }, + { + "epoch": 6.074310827007591, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15204 + }, + { + "epoch": 6.074710347582901, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15205 + }, + { + "epoch": 6.0751098681582105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15206 + }, + { + "epoch": 6.075509388733519, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15207 + }, + { + "epoch": 6.075908909308829, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15208 + }, + { + "epoch": 6.076308429884139, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15209 + }, + { + "epoch": 6.076707950459449, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15210 + }, + { + "epoch": 6.077107471034759, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15211 + }, + { + "epoch": 6.077506991610068, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15212 + }, + { + "epoch": 6.077906512185377, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15213 + }, + { + "epoch": 6.078306032760687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15214 + }, + { + "epoch": 6.078705553335997, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15215 + }, + { + "epoch": 6.079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15216 + }, + { + "epoch": 6.079504594486616, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15217 + }, + { + "epoch": 6.079904115061925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15218 + }, + { + "epoch": 6.080303635637235, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15219 + }, + { + "epoch": 6.080703156212545, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 15220 + }, + { + "epoch": 6.081102676787855, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15221 + }, + { + "epoch": 6.081502197363164, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15222 + }, + { + "epoch": 6.081901717938474, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15223 + }, + { + "epoch": 6.082301238513783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15224 + }, + { + "epoch": 6.082700759089093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15225 + }, + { + "epoch": 6.083100279664403, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15226 + }, + { + "epoch": 6.0834998002397125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15227 + }, + { + "epoch": 6.083899320815022, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15228 + }, + { + "epoch": 6.084298841390331, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15229 + }, + { + "epoch": 6.084698361965641, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15230 + }, + { + "epoch": 6.085097882540951, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15231 + }, + { + "epoch": 6.0854974031162605, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15232 + }, + { + "epoch": 6.08589692369157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15233 + }, + { + "epoch": 6.08629644426688, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15234 + }, + { + "epoch": 6.086695964842189, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15235 + }, + { + "epoch": 6.087095485417499, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15236 + }, + { + "epoch": 6.087495005992809, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15237 + }, + { + "epoch": 6.087894526568118, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15238 + }, + { + "epoch": 6.088294047143428, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15239 + }, + { + "epoch": 6.088693567718738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15240 + }, + { + "epoch": 6.089093088294047, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15241 + }, + { + "epoch": 6.089492608869357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15242 + }, + { + "epoch": 6.089892129444666, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15243 + }, + { + "epoch": 6.090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15244 + }, + { + "epoch": 6.090691170595286, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15245 + }, + { + "epoch": 6.091090691170595, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15246 + }, + { + "epoch": 6.091490211745905, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15247 + }, + { + "epoch": 6.0918897323212144, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15248 + }, + { + "epoch": 6.092289252896524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15249 + }, + { + "epoch": 6.092688773471834, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15250 + }, + { + "epoch": 6.093088294047144, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15251 + }, + { + "epoch": 6.093487814622453, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15252 + }, + { + "epoch": 6.0938873351977625, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15253 + }, + { + "epoch": 6.094286855773072, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15254 + }, + { + "epoch": 6.094686376348382, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15255 + }, + { + "epoch": 6.095085896923692, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15256 + }, + { + "epoch": 6.095485417499002, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15257 + }, + { + "epoch": 6.0958849380743105, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15258 + }, + { + "epoch": 6.09628445864962, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15259 + }, + { + "epoch": 6.09668397922493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15260 + }, + { + "epoch": 6.09708349980024, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15261 + }, + { + "epoch": 6.09748302037555, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15262 + }, + { + "epoch": 6.097882540950859, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15263 + }, + { + "epoch": 6.098282061526168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15264 + }, + { + "epoch": 6.098681582101478, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15265 + }, + { + "epoch": 6.099081102676788, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15266 + }, + { + "epoch": 6.099480623252098, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15267 + }, + { + "epoch": 6.0998801438274075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15268 + }, + { + "epoch": 6.100279664402716, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15269 + }, + { + "epoch": 6.100679184978026, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15270 + }, + { + "epoch": 6.101078705553336, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15271 + }, + { + "epoch": 6.101478226128646, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15272 + }, + { + "epoch": 6.1018777467039556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15273 + }, + { + "epoch": 6.1022772672792644, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15274 + }, + { + "epoch": 6.102676787854574, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15275 + }, + { + "epoch": 6.103076308429884, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15276 + }, + { + "epoch": 6.103475829005194, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15277 + }, + { + "epoch": 6.103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15278 + }, + { + "epoch": 6.104274870155813, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15279 + }, + { + "epoch": 6.104674390731122, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15280 + }, + { + "epoch": 6.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15281 + }, + { + "epoch": 6.105473431881742, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15282 + }, + { + "epoch": 6.105872952457052, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15283 + }, + { + "epoch": 6.106272473032361, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15284 + }, + { + "epoch": 6.106671993607671, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15285 + }, + { + "epoch": 6.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15286 + }, + { + "epoch": 6.10747103475829, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15287 + }, + { + "epoch": 6.1078705553336, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15288 + }, + { + "epoch": 6.1082700759089095, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15289 + }, + { + "epoch": 6.108669596484219, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15290 + }, + { + "epoch": 6.109069117059528, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15291 + }, + { + "epoch": 6.109468637634838, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15292 + }, + { + "epoch": 6.109868158210148, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15293 + }, + { + "epoch": 6.1102676787854575, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15294 + }, + { + "epoch": 6.110667199360767, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15295 + }, + { + "epoch": 6.111066719936077, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15296 + }, + { + "epoch": 6.111466240511386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15297 + }, + { + "epoch": 6.111865761086696, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15298 + }, + { + "epoch": 6.112265281662006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15299 + }, + { + "epoch": 6.112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15300 + }, + { + "epoch": 6.113064322812625, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15301 + }, + { + "epoch": 6.113463843387935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15302 + }, + { + "epoch": 6.113863363963244, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15303 + }, + { + "epoch": 6.114262884538554, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15304 + }, + { + "epoch": 6.114662405113863, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15305 + }, + { + "epoch": 6.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15306 + }, + { + "epoch": 6.115461446264483, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15307 + }, + { + "epoch": 6.115860966839792, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15308 + }, + { + "epoch": 6.116260487415102, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15309 + }, + { + "epoch": 6.116660007990411, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15310 + }, + { + "epoch": 6.117059528565721, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15311 + }, + { + "epoch": 6.117459049141031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 15312 + }, + { + "epoch": 6.117858569716341, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15313 + }, + { + "epoch": 6.11825809029165, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15314 + }, + { + "epoch": 6.1186576108669595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15315 + }, + { + "epoch": 6.119057131442269, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 15316 + }, + { + "epoch": 6.119456652017579, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15317 + }, + { + "epoch": 6.119856172592889, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15318 + }, + { + "epoch": 6.120255693168198, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15319 + }, + { + "epoch": 6.1206552137435075, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15320 + }, + { + "epoch": 6.121054734318817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15321 + }, + { + "epoch": 6.121454254894127, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15322 + }, + { + "epoch": 6.121853775469437, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15323 + }, + { + "epoch": 6.122253296044747, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15324 + }, + { + "epoch": 6.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15325 + }, + { + "epoch": 6.123052337195365, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15326 + }, + { + "epoch": 6.123451857770675, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15327 + }, + { + "epoch": 6.123851378345985, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15328 + }, + { + "epoch": 6.124250898921295, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15329 + }, + { + "epoch": 6.1246504194966045, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15330 + }, + { + "epoch": 6.125049940071913, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 15331 + }, + { + "epoch": 6.125449460647223, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15332 + }, + { + "epoch": 6.125848981222533, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15333 + }, + { + "epoch": 6.126248501797843, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15334 + }, + { + "epoch": 6.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15335 + }, + { + "epoch": 6.127047542948461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15336 + }, + { + "epoch": 6.127447063523771, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15337 + }, + { + "epoch": 6.127846584099081, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15338 + }, + { + "epoch": 6.128246104674391, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 15339 + }, + { + "epoch": 6.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15340 + }, + { + "epoch": 6.12904514582501, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15341 + }, + { + "epoch": 6.129444666400319, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15342 + }, + { + "epoch": 6.129844186975629, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15343 + }, + { + "epoch": 6.130243707550939, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15344 + }, + { + "epoch": 6.130643228126249, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15345 + }, + { + "epoch": 6.131042748701558, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15346 + }, + { + "epoch": 6.131442269276867, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15347 + }, + { + "epoch": 6.131841789852177, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15348 + }, + { + "epoch": 6.132241310427487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15349 + }, + { + "epoch": 6.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15350 + }, + { + "epoch": 6.1330403515781065, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15351 + }, + { + "epoch": 6.133439872153416, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15352 + }, + { + "epoch": 6.133839392728725, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15353 + }, + { + "epoch": 6.134238913304035, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15354 + }, + { + "epoch": 6.134638433879345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15355 + }, + { + "epoch": 6.1350379544546545, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15356 + }, + { + "epoch": 6.135437475029964, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15357 + }, + { + "epoch": 6.135836995605274, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15358 + }, + { + "epoch": 6.136236516180583, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15359 + }, + { + "epoch": 6.136636036755893, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15360 + }, + { + "epoch": 6.1370355573312025, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15361 + }, + { + "epoch": 6.137435077906512, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15362 + }, + { + "epoch": 6.137834598481822, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15363 + }, + { + "epoch": 6.138234119057131, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15364 + }, + { + "epoch": 6.138633639632441, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 15365 + }, + { + "epoch": 6.139033160207751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15366 + }, + { + "epoch": 6.13943268078306, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15367 + }, + { + "epoch": 6.13983220135837, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15368 + }, + { + "epoch": 6.14023172193368, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15369 + }, + { + "epoch": 6.140631242508989, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15370 + }, + { + "epoch": 6.141030763084299, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15371 + }, + { + "epoch": 6.141430283659608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15372 + }, + { + "epoch": 6.141829804234918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15373 + }, + { + "epoch": 6.142229324810228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15374 + }, + { + "epoch": 6.142628845385538, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15375 + }, + { + "epoch": 6.143028365960847, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15376 + }, + { + "epoch": 6.1434278865361565, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15377 + }, + { + "epoch": 6.143827407111466, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15378 + }, + { + "epoch": 6.144226927686776, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15379 + }, + { + "epoch": 6.144626448262086, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15380 + }, + { + "epoch": 6.145025968837395, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15381 + }, + { + "epoch": 6.1454254894127045, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 15382 + }, + { + "epoch": 6.145825009988014, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15383 + }, + { + "epoch": 6.146224530563324, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15384 + }, + { + "epoch": 6.146624051138634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15385 + }, + { + "epoch": 6.147023571713944, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15386 + }, + { + "epoch": 6.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15387 + }, + { + "epoch": 6.147822612864562, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15388 + }, + { + "epoch": 6.148222133439872, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15389 + }, + { + "epoch": 6.148621654015182, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15390 + }, + { + "epoch": 6.149021174590492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15391 + }, + { + "epoch": 6.1494206951658015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15392 + }, + { + "epoch": 6.14982021574111, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15393 + }, + { + "epoch": 6.15021973631642, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15394 + }, + { + "epoch": 6.15061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15395 + }, + { + "epoch": 6.15101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15396 + }, + { + "epoch": 6.1514182980423495, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15397 + }, + { + "epoch": 6.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 15398 + }, + { + "epoch": 6.152217339192968, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15399 + }, + { + "epoch": 6.152616859768278, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15400 + }, + { + "epoch": 6.153016380343588, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15401 + }, + { + "epoch": 6.153415900918898, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15402 + }, + { + "epoch": 6.153815421494207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15403 + }, + { + "epoch": 6.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2963, + "step": 15404 + }, + { + "epoch": 6.154614462644826, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15405 + }, + { + "epoch": 6.155013983220136, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15406 + }, + { + "epoch": 6.155413503795446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15407 + }, + { + "epoch": 6.155813024370755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15408 + }, + { + "epoch": 6.156212544946065, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15409 + }, + { + "epoch": 6.156612065521374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15410 + }, + { + "epoch": 6.157011586096684, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15411 + }, + { + "epoch": 6.157411106671994, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15412 + }, + { + "epoch": 6.157810627247303, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15413 + }, + { + "epoch": 6.158210147822613, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15414 + }, + { + "epoch": 6.158609668397922, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15415 + }, + { + "epoch": 6.159009188973232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15416 + }, + { + "epoch": 6.159408709548542, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15417 + }, + { + "epoch": 6.1598082301238515, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15418 + }, + { + "epoch": 6.160207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15419 + }, + { + "epoch": 6.160607271274471, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.2983, + "step": 15420 + }, + { + "epoch": 6.16100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 15421 + }, + { + "epoch": 6.16140631242509, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15422 + }, + { + "epoch": 6.1618058330003995, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15423 + }, + { + "epoch": 6.162205353575709, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15424 + }, + { + "epoch": 6.162604874151019, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15425 + }, + { + "epoch": 6.163004394726328, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15426 + }, + { + "epoch": 6.163403915301638, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15427 + }, + { + "epoch": 6.163803435876948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15428 + }, + { + "epoch": 6.164202956452257, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15429 + }, + { + "epoch": 6.164602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15430 + }, + { + "epoch": 6.165001997602877, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15431 + }, + { + "epoch": 6.165401518178186, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15432 + }, + { + "epoch": 6.165801038753496, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15433 + }, + { + "epoch": 6.166200559328805, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15434 + }, + { + "epoch": 6.166600079904115, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15435 + }, + { + "epoch": 6.166999600479425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15436 + }, + { + "epoch": 6.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 15437 + }, + { + "epoch": 6.167798641630044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15438 + }, + { + "epoch": 6.168198162205353, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15439 + }, + { + "epoch": 6.168597682780663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15440 + }, + { + "epoch": 6.168997203355973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15441 + }, + { + "epoch": 6.169396723931283, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15442 + }, + { + "epoch": 6.169796244506592, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15443 + }, + { + "epoch": 6.1701957650819015, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 15444 + }, + { + "epoch": 6.170595285657211, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15445 + }, + { + "epoch": 6.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15446 + }, + { + "epoch": 6.171394326807831, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15447 + }, + { + "epoch": 6.171793847383141, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15448 + }, + { + "epoch": 6.1721933679584495, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15449 + }, + { + "epoch": 6.172592888533759, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15450 + }, + { + "epoch": 6.172992409109069, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 15451 + }, + { + "epoch": 6.173391929684379, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15452 + }, + { + "epoch": 6.173791450259689, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15453 + }, + { + "epoch": 6.174190970834998, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15454 + }, + { + "epoch": 6.174590491410307, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15455 + }, + { + "epoch": 6.174990011985617, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15456 + }, + { + "epoch": 6.175389532560927, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15457 + }, + { + "epoch": 6.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15458 + }, + { + "epoch": 6.1761885737115465, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15459 + }, + { + "epoch": 6.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15460 + }, + { + "epoch": 6.176987614862165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15461 + }, + { + "epoch": 6.177387135437475, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 15462 + }, + { + "epoch": 6.177786656012785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15463 + }, + { + "epoch": 6.1781861765880945, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15464 + }, + { + "epoch": 6.178585697163404, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15465 + }, + { + "epoch": 6.178985217738713, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15466 + }, + { + "epoch": 6.179384738314023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15467 + }, + { + "epoch": 6.179784258889333, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15468 + }, + { + "epoch": 6.180183779464643, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15469 + }, + { + "epoch": 6.180583300039952, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15470 + }, + { + "epoch": 6.180982820615261, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15471 + }, + { + "epoch": 6.181382341190571, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15472 + }, + { + "epoch": 6.181781861765881, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15473 + }, + { + "epoch": 6.182181382341191, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15474 + }, + { + "epoch": 6.1825809029165, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15475 + }, + { + "epoch": 6.18298042349181, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15476 + }, + { + "epoch": 6.183379944067119, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15477 + }, + { + "epoch": 6.183779464642429, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 15478 + }, + { + "epoch": 6.184178985217739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15479 + }, + { + "epoch": 6.1845785057930485, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15480 + }, + { + "epoch": 6.184978026368358, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15481 + }, + { + "epoch": 6.185377546943668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15482 + }, + { + "epoch": 6.185777067518977, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15483 + }, + { + "epoch": 6.186176588094287, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15484 + }, + { + "epoch": 6.1865761086695965, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15485 + }, + { + "epoch": 6.186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15486 + }, + { + "epoch": 6.187375149820216, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15487 + }, + { + "epoch": 6.187774670395525, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15488 + }, + { + "epoch": 6.188174190970835, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15489 + }, + { + "epoch": 6.1885737115461446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15490 + }, + { + "epoch": 6.188973232121454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15491 + }, + { + "epoch": 6.189372752696764, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15492 + }, + { + "epoch": 6.189772273272074, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15493 + }, + { + "epoch": 6.190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15494 + }, + { + "epoch": 6.190571314422693, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15495 + }, + { + "epoch": 6.190970834998002, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15496 + }, + { + "epoch": 6.191370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 15497 + }, + { + "epoch": 6.191769876148622, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15498 + }, + { + "epoch": 6.192169396723932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15499 + }, + { + "epoch": 6.192568917299241, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15500 + }, + { + "epoch": 6.19296843787455, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15501 + }, + { + "epoch": 6.19336795844986, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15502 + }, + { + "epoch": 6.19376747902517, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15503 + }, + { + "epoch": 6.19416699960048, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15504 + }, + { + "epoch": 6.194566520175789, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15505 + }, + { + "epoch": 6.1949660407510985, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15506 + }, + { + "epoch": 6.195365561326408, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15507 + }, + { + "epoch": 6.195765081901718, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15508 + }, + { + "epoch": 6.196164602477028, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15509 + }, + { + "epoch": 6.196564123052338, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15510 + }, + { + "epoch": 6.1969636436276465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15511 + }, + { + "epoch": 6.197363164202956, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15512 + }, + { + "epoch": 6.197762684778266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15513 + }, + { + "epoch": 6.198162205353576, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15514 + }, + { + "epoch": 6.198561725928886, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15515 + }, + { + "epoch": 6.1989612465041946, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15516 + }, + { + "epoch": 6.199360767079504, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15517 + }, + { + "epoch": 6.199760287654814, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15518 + }, + { + "epoch": 6.200159808230124, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15519 + }, + { + "epoch": 6.200559328805434, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15520 + }, + { + "epoch": 6.2009588493807435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 15521 + }, + { + "epoch": 6.201358369956052, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15522 + }, + { + "epoch": 6.201757890531362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15523 + }, + { + "epoch": 6.202157411106672, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15524 + }, + { + "epoch": 6.202556931681982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 15525 + }, + { + "epoch": 6.2029564522572915, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15526 + }, + { + "epoch": 6.2033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15527 + }, + { + "epoch": 6.20375549340791, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15528 + }, + { + "epoch": 6.20415501398322, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15529 + }, + { + "epoch": 6.20455453455853, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15530 + }, + { + "epoch": 6.20495405513384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15531 + }, + { + "epoch": 6.205353575709149, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15532 + }, + { + "epoch": 6.205753096284458, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15533 + }, + { + "epoch": 6.206152616859768, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15534 + }, + { + "epoch": 6.206552137435078, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15535 + }, + { + "epoch": 6.206951658010388, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15536 + }, + { + "epoch": 6.207351178585697, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15537 + }, + { + "epoch": 6.207750699161007, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15538 + }, + { + "epoch": 6.208150219736316, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15539 + }, + { + "epoch": 6.208549740311626, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15540 + }, + { + "epoch": 6.208949260886936, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15541 + }, + { + "epoch": 6.2093487814622454, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15542 + }, + { + "epoch": 6.209748302037555, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15543 + }, + { + "epoch": 6.210147822612864, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15544 + }, + { + "epoch": 6.210547343188174, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15545 + }, + { + "epoch": 6.210946863763484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15546 + }, + { + "epoch": 6.2113463843387935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15547 + }, + { + "epoch": 6.211745904914103, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15548 + }, + { + "epoch": 6.212145425489413, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15549 + }, + { + "epoch": 6.212544946064722, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15550 + }, + { + "epoch": 6.212944466640032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15551 + }, + { + "epoch": 6.2133439872153415, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15552 + }, + { + "epoch": 6.213743507790651, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15553 + }, + { + "epoch": 6.214143028365961, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15554 + }, + { + "epoch": 6.214542548941271, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15555 + }, + { + "epoch": 6.21494206951658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15556 + }, + { + "epoch": 6.21534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 15557 + }, + { + "epoch": 6.215741110667199, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 15558 + }, + { + "epoch": 6.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 15559 + }, + { + "epoch": 6.216540151817819, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15560 + }, + { + "epoch": 6.216939672393128, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15561 + }, + { + "epoch": 6.217339192968438, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15562 + }, + { + "epoch": 6.217738713543747, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15563 + }, + { + "epoch": 6.218138234119057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15564 + }, + { + "epoch": 6.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15565 + }, + { + "epoch": 6.218937275269677, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15566 + }, + { + "epoch": 6.219336795844986, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15567 + }, + { + "epoch": 6.2197363164202955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15568 + }, + { + "epoch": 6.220135836995605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15569 + }, + { + "epoch": 6.220535357570915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15570 + }, + { + "epoch": 6.220934878146225, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15571 + }, + { + "epoch": 6.221334398721535, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15572 + }, + { + "epoch": 6.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15573 + }, + { + "epoch": 6.222133439872153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15574 + }, + { + "epoch": 6.222532960447463, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 15575 + }, + { + "epoch": 6.222932481022773, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15576 + }, + { + "epoch": 6.223332001598083, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15577 + }, + { + "epoch": 6.2237315221733915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15578 + }, + { + "epoch": 6.224131042748701, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15579 + }, + { + "epoch": 6.224530563324011, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15580 + }, + { + "epoch": 6.224930083899321, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15581 + }, + { + "epoch": 6.225329604474631, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15582 + }, + { + "epoch": 6.2257291250499405, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15583 + }, + { + "epoch": 6.226128645625249, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15584 + }, + { + "epoch": 6.226528166200559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15585 + }, + { + "epoch": 6.226927686775869, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15586 + }, + { + "epoch": 6.227327207351179, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15587 + }, + { + "epoch": 6.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15588 + }, + { + "epoch": 6.228126248501797, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15589 + }, + { + "epoch": 6.228525769077107, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15590 + }, + { + "epoch": 6.228925289652417, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15591 + }, + { + "epoch": 6.229324810227727, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 15592 + }, + { + "epoch": 6.229724330803037, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15593 + }, + { + "epoch": 6.230123851378346, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15594 + }, + { + "epoch": 6.230523371953655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15595 + }, + { + "epoch": 6.230922892528965, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15596 + }, + { + "epoch": 6.231322413104275, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15597 + }, + { + "epoch": 6.231721933679585, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15598 + }, + { + "epoch": 6.232121454254894, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15599 + }, + { + "epoch": 6.232520974830204, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15600 + }, + { + "epoch": 6.232920495405513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15601 + }, + { + "epoch": 6.233320015980823, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15602 + }, + { + "epoch": 6.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15603 + }, + { + "epoch": 6.234119057131442, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15604 + }, + { + "epoch": 6.234518577706752, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 15605 + }, + { + "epoch": 6.234918098282061, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15606 + }, + { + "epoch": 6.235317618857371, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 15607 + }, + { + "epoch": 6.235717139432681, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15608 + }, + { + "epoch": 6.2361166600079905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15609 + }, + { + "epoch": 6.2365161805833, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15610 + }, + { + "epoch": 6.23691570115861, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15611 + }, + { + "epoch": 6.237315221733919, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15612 + }, + { + "epoch": 6.237714742309229, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15613 + }, + { + "epoch": 6.2381142628845385, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 15614 + }, + { + "epoch": 6.238513783459848, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15615 + }, + { + "epoch": 6.238913304035158, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15616 + }, + { + "epoch": 6.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15617 + }, + { + "epoch": 6.239712345185777, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15618 + }, + { + "epoch": 6.240111865761087, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15619 + }, + { + "epoch": 6.240511386336396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15620 + }, + { + "epoch": 6.240910906911706, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15621 + }, + { + "epoch": 6.241310427487016, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15622 + }, + { + "epoch": 6.241709948062325, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15623 + }, + { + "epoch": 6.242109468637635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15624 + }, + { + "epoch": 6.242508989212944, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15625 + }, + { + "epoch": 6.242908509788254, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15626 + }, + { + "epoch": 6.243308030363564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15627 + }, + { + "epoch": 6.243707550938874, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15628 + }, + { + "epoch": 6.244107071514183, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15629 + }, + { + "epoch": 6.244506592089492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15630 + }, + { + "epoch": 6.244906112664802, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15631 + }, + { + "epoch": 6.245305633240112, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15632 + }, + { + "epoch": 6.245705153815422, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15633 + }, + { + "epoch": 6.246104674390731, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15634 + }, + { + "epoch": 6.2465041949660405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 15635 + }, + { + "epoch": 6.24690371554135, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15636 + }, + { + "epoch": 6.24730323611666, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15637 + }, + { + "epoch": 6.24770275669197, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15638 + }, + { + "epoch": 6.24810227726728, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15639 + }, + { + "epoch": 6.2485017978425885, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15640 + }, + { + "epoch": 6.248901318417898, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15641 + }, + { + "epoch": 6.249300838993208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15642 + }, + { + "epoch": 6.249700359568518, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15643 + }, + { + "epoch": 6.250099880143828, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15644 + }, + { + "epoch": 6.2504994007191375, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15645 + }, + { + "epoch": 6.250898921294446, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15646 + }, + { + "epoch": 6.251298441869756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15647 + }, + { + "epoch": 6.251697962445066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15648 + }, + { + "epoch": 6.252097483020376, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15649 + }, + { + "epoch": 6.2524970035956855, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15650 + }, + { + "epoch": 6.252896524170994, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15651 + }, + { + "epoch": 6.253296044746304, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15652 + }, + { + "epoch": 6.253695565321614, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15653 + }, + { + "epoch": 6.254095085896924, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15654 + }, + { + "epoch": 6.2544946064722335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15655 + }, + { + "epoch": 6.254894127047543, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15656 + }, + { + "epoch": 6.255293647622852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15657 + }, + { + "epoch": 6.255693168198162, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15658 + }, + { + "epoch": 6.256092688773472, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15659 + }, + { + "epoch": 6.256492209348782, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15660 + }, + { + "epoch": 6.256891729924091, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15661 + }, + { + "epoch": 6.257291250499401, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15662 + }, + { + "epoch": 6.25769077107471, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15663 + }, + { + "epoch": 6.25809029165002, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15664 + }, + { + "epoch": 6.25848981222533, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15665 + }, + { + "epoch": 6.258889332800639, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15666 + }, + { + "epoch": 6.259288853375949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 15667 + }, + { + "epoch": 6.259688373951258, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15668 + }, + { + "epoch": 6.260087894526568, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15669 + }, + { + "epoch": 6.260487415101878, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15670 + }, + { + "epoch": 6.2608869356771875, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 15671 + }, + { + "epoch": 6.261286456252497, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15672 + }, + { + "epoch": 6.261685976827807, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15673 + }, + { + "epoch": 6.262085497403116, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15674 + }, + { + "epoch": 6.262485017978426, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15675 + }, + { + "epoch": 6.2628845385537355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15676 + }, + { + "epoch": 6.263284059129045, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15677 + }, + { + "epoch": 6.263683579704355, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 15678 + }, + { + "epoch": 6.264083100279665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 15679 + }, + { + "epoch": 6.264482620854974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15680 + }, + { + "epoch": 6.2648821414302835, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15681 + }, + { + "epoch": 6.265281662005593, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15682 + }, + { + "epoch": 6.265681182580903, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15683 + }, + { + "epoch": 6.266080703156213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15684 + }, + { + "epoch": 6.266480223731522, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15685 + }, + { + "epoch": 6.266879744306832, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15686 + }, + { + "epoch": 6.267279264882141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15687 + }, + { + "epoch": 6.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15688 + }, + { + "epoch": 6.268078306032761, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15689 + }, + { + "epoch": 6.26847782660807, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15690 + }, + { + "epoch": 6.26887734718338, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15691 + }, + { + "epoch": 6.269276867758689, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15692 + }, + { + "epoch": 6.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15693 + }, + { + "epoch": 6.270075908909309, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15694 + }, + { + "epoch": 6.270475429484619, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15695 + }, + { + "epoch": 6.270874950059928, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15696 + }, + { + "epoch": 6.2712744706352375, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15697 + }, + { + "epoch": 6.271673991210547, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15698 + }, + { + "epoch": 6.272073511785857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15699 + }, + { + "epoch": 6.272473032361167, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15700 + }, + { + "epoch": 6.272872552936477, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15701 + }, + { + "epoch": 6.2732720735117855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15702 + }, + { + "epoch": 6.273671594087095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 15703 + }, + { + "epoch": 6.274071114662405, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15704 + }, + { + "epoch": 6.274470635237715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15705 + }, + { + "epoch": 6.274870155813025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15706 + }, + { + "epoch": 6.2752696763883336, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 15707 + }, + { + "epoch": 6.275669196963643, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15708 + }, + { + "epoch": 6.276068717538953, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15709 + }, + { + "epoch": 6.276468238114263, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15710 + }, + { + "epoch": 6.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15711 + }, + { + "epoch": 6.2772672792648825, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15712 + }, + { + "epoch": 6.277666799840191, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15713 + }, + { + "epoch": 6.278066320415501, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15714 + }, + { + "epoch": 6.278465840990811, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15715 + }, + { + "epoch": 6.278865361566121, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15716 + }, + { + "epoch": 6.2792648821414305, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15717 + }, + { + "epoch": 6.27966440271674, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15718 + }, + { + "epoch": 6.280063923292049, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15719 + }, + { + "epoch": 6.280463443867359, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15720 + }, + { + "epoch": 6.280862964442669, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15721 + }, + { + "epoch": 6.281262485017979, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15722 + }, + { + "epoch": 6.281662005593288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15723 + }, + { + "epoch": 6.282061526168597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 15724 + }, + { + "epoch": 6.282461046743907, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15725 + }, + { + "epoch": 6.282860567319217, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15726 + }, + { + "epoch": 6.283260087894527, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15727 + }, + { + "epoch": 6.283659608469836, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15728 + }, + { + "epoch": 6.284059129045146, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15729 + }, + { + "epoch": 6.284458649620455, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15730 + }, + { + "epoch": 6.284858170195765, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 15731 + }, + { + "epoch": 6.285257690771075, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 15732 + }, + { + "epoch": 6.285657211346384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15733 + }, + { + "epoch": 6.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15734 + }, + { + "epoch": 6.286456252497004, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15735 + }, + { + "epoch": 6.286855773072313, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15736 + }, + { + "epoch": 6.287255293647623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15737 + }, + { + "epoch": 6.2876548142229325, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 15738 + }, + { + "epoch": 6.288054334798242, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15739 + }, + { + "epoch": 6.288453855373552, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15740 + }, + { + "epoch": 6.288853375948861, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15741 + }, + { + "epoch": 6.289252896524171, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15742 + }, + { + "epoch": 6.2896524170994805, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15743 + }, + { + "epoch": 6.29005193767479, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15744 + }, + { + "epoch": 6.2904514582501, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15745 + }, + { + "epoch": 6.29085097882541, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15746 + }, + { + "epoch": 6.291250499400719, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15747 + }, + { + "epoch": 6.291650019976029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15748 + }, + { + "epoch": 6.292049540551338, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15749 + }, + { + "epoch": 6.292449061126648, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15750 + }, + { + "epoch": 6.292848581701958, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15751 + }, + { + "epoch": 6.293248102277268, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15752 + }, + { + "epoch": 6.293647622852577, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15753 + }, + { + "epoch": 6.294047143427886, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15754 + }, + { + "epoch": 6.294446664003196, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15755 + }, + { + "epoch": 6.294846184578506, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15756 + }, + { + "epoch": 6.295245705153816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15757 + }, + { + "epoch": 6.295645225729125, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 15758 + }, + { + "epoch": 6.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15759 + }, + { + "epoch": 6.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15760 + }, + { + "epoch": 6.296843787455054, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15761 + }, + { + "epoch": 6.297243308030364, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15762 + }, + { + "epoch": 6.297642828605674, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15763 + }, + { + "epoch": 6.2980423491809825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15764 + }, + { + "epoch": 6.298441869756292, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15765 + }, + { + "epoch": 6.298841390331602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15766 + }, + { + "epoch": 6.299240910906912, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15767 + }, + { + "epoch": 6.299640431482222, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 15768 + }, + { + "epoch": 6.300039952057531, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15769 + }, + { + "epoch": 6.30043947263284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15770 + }, + { + "epoch": 6.30083899320815, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15771 + }, + { + "epoch": 6.30123851378346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15772 + }, + { + "epoch": 6.30163803435877, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15773 + }, + { + "epoch": 6.3020375549340795, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15774 + }, + { + "epoch": 6.302437075509388, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15775 + }, + { + "epoch": 6.302836596084698, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15776 + }, + { + "epoch": 6.303236116660008, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15777 + }, + { + "epoch": 6.303635637235318, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15778 + }, + { + "epoch": 6.3040351578106275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15779 + }, + { + "epoch": 6.304434678385936, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15780 + }, + { + "epoch": 6.304834198961246, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15781 + }, + { + "epoch": 6.305233719536556, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15782 + }, + { + "epoch": 6.305633240111866, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15783 + }, + { + "epoch": 6.3060327606871756, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15784 + }, + { + "epoch": 6.306432281262485, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 15785 + }, + { + "epoch": 6.306831801837794, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15786 + }, + { + "epoch": 6.307231322413104, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15787 + }, + { + "epoch": 6.307630842988414, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15788 + }, + { + "epoch": 6.308030363563724, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15789 + }, + { + "epoch": 6.308429884139033, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15790 + }, + { + "epoch": 6.308829404714343, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15791 + }, + { + "epoch": 6.309228925289652, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15792 + }, + { + "epoch": 6.309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15793 + }, + { + "epoch": 6.310027966440272, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15794 + }, + { + "epoch": 6.310427487015581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15795 + }, + { + "epoch": 6.310827007590891, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15796 + }, + { + "epoch": 6.3112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15797 + }, + { + "epoch": 6.31162604874151, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15798 + }, + { + "epoch": 6.31202556931682, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15799 + }, + { + "epoch": 6.3124250898921295, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 15800 + }, + { + "epoch": 6.312824610467439, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15801 + }, + { + "epoch": 6.313224131042749, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 15802 + }, + { + "epoch": 6.313623651618058, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15803 + }, + { + "epoch": 6.314023172193368, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15804 + }, + { + "epoch": 6.3144226927686775, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15805 + }, + { + "epoch": 6.314822213343987, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15806 + }, + { + "epoch": 6.315221733919297, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15807 + }, + { + "epoch": 6.315621254494607, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15808 + }, + { + "epoch": 6.316020775069916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15809 + }, + { + "epoch": 6.316420295645226, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15810 + }, + { + "epoch": 6.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15811 + }, + { + "epoch": 6.317219336795845, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 15812 + }, + { + "epoch": 6.317618857371155, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15813 + }, + { + "epoch": 6.318018377946464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15814 + }, + { + "epoch": 6.318417898521774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 15815 + }, + { + "epoch": 6.318817419097083, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15816 + }, + { + "epoch": 6.319216939672393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15817 + }, + { + "epoch": 6.319616460247703, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15818 + }, + { + "epoch": 6.320015980823013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15819 + }, + { + "epoch": 6.320415501398322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15820 + }, + { + "epoch": 6.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15821 + }, + { + "epoch": 6.321214542548941, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15822 + }, + { + "epoch": 6.321614063124251, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15823 + }, + { + "epoch": 6.322013583699561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15824 + }, + { + "epoch": 6.322413104274871, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15825 + }, + { + "epoch": 6.3228126248501795, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15826 + }, + { + "epoch": 6.323212145425489, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15827 + }, + { + "epoch": 6.323611666000799, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15828 + }, + { + "epoch": 6.324011186576109, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15829 + }, + { + "epoch": 6.324410707151419, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15830 + }, + { + "epoch": 6.3248102277267275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15831 + }, + { + "epoch": 6.325209748302037, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15832 + }, + { + "epoch": 6.325609268877347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15833 + }, + { + "epoch": 6.326008789452657, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15834 + }, + { + "epoch": 6.326408310027967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15835 + }, + { + "epoch": 6.3268078306032765, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15836 + }, + { + "epoch": 6.327207351178585, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15837 + }, + { + "epoch": 6.327606871753895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15838 + }, + { + "epoch": 6.328006392329205, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15839 + }, + { + "epoch": 6.328405912904515, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15840 + }, + { + "epoch": 6.3288054334798245, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15841 + }, + { + "epoch": 6.329204954055134, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15842 + }, + { + "epoch": 6.329604474630443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15843 + }, + { + "epoch": 6.330003995205753, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 15844 + }, + { + "epoch": 6.330403515781063, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15845 + }, + { + "epoch": 6.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 15846 + }, + { + "epoch": 6.331202556931682, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15847 + }, + { + "epoch": 6.331602077506991, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15848 + }, + { + "epoch": 6.332001598082301, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15849 + }, + { + "epoch": 6.332401118657611, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15850 + }, + { + "epoch": 6.332800639232921, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15851 + }, + { + "epoch": 6.33320015980823, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15852 + }, + { + "epoch": 6.33359968038354, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15853 + }, + { + "epoch": 6.333999200958849, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15854 + }, + { + "epoch": 6.334398721534159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15855 + }, + { + "epoch": 6.334798242109469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15856 + }, + { + "epoch": 6.335197762684778, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15857 + }, + { + "epoch": 6.335597283260088, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15858 + }, + { + "epoch": 6.335996803835398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15859 + }, + { + "epoch": 6.336396324410707, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15860 + }, + { + "epoch": 6.336795844986017, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15861 + }, + { + "epoch": 6.3371953655613265, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15862 + }, + { + "epoch": 6.337594886136636, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15863 + }, + { + "epoch": 6.337994406711946, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15864 + }, + { + "epoch": 6.338393927287255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15865 + }, + { + "epoch": 6.338793447862565, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15866 + }, + { + "epoch": 6.3391929684378745, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15867 + }, + { + "epoch": 6.339592489013184, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15868 + }, + { + "epoch": 6.339992009588494, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15869 + }, + { + "epoch": 6.340391530163803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15870 + }, + { + "epoch": 6.340791050739113, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15871 + }, + { + "epoch": 6.3411905713144225, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15872 + }, + { + "epoch": 6.341590091889732, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15873 + }, + { + "epoch": 6.341989612465042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15874 + }, + { + "epoch": 6.342389133040352, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15875 + }, + { + "epoch": 6.342788653615661, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15876 + }, + { + "epoch": 6.343188174190971, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15877 + }, + { + "epoch": 6.34358769476628, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15878 + }, + { + "epoch": 6.34398721534159, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15879 + }, + { + "epoch": 6.3443867359169, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15880 + }, + { + "epoch": 6.34478625649221, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15881 + }, + { + "epoch": 6.345185777067519, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15882 + }, + { + "epoch": 6.345585297642828, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15883 + }, + { + "epoch": 6.345984818218138, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15884 + }, + { + "epoch": 6.346384338793448, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15885 + }, + { + "epoch": 6.346783859368758, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15886 + }, + { + "epoch": 6.347183379944067, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15887 + }, + { + "epoch": 6.3475829005193765, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15888 + }, + { + "epoch": 6.347982421094686, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15889 + }, + { + "epoch": 6.348381941669996, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15890 + }, + { + "epoch": 6.348781462245306, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 15891 + }, + { + "epoch": 6.349180982820616, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15892 + }, + { + "epoch": 6.3495805033959245, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15893 + }, + { + "epoch": 6.349980023971234, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15894 + }, + { + "epoch": 6.350379544546544, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15895 + }, + { + "epoch": 6.350779065121854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15896 + }, + { + "epoch": 6.351178585697164, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15897 + }, + { + "epoch": 6.351578106272473, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15898 + }, + { + "epoch": 6.351977626847782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15899 + }, + { + "epoch": 6.352377147423092, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15900 + }, + { + "epoch": 6.352776667998402, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15901 + }, + { + "epoch": 6.353176188573712, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 15902 + }, + { + "epoch": 6.3535757091490215, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15903 + }, + { + "epoch": 6.35397522972433, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 15904 + }, + { + "epoch": 6.35437475029964, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15905 + }, + { + "epoch": 6.35477427087495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15906 + }, + { + "epoch": 6.35517379145026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15907 + }, + { + "epoch": 6.3555733120255695, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15908 + }, + { + "epoch": 6.355972832600879, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15909 + }, + { + "epoch": 6.356372353176188, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15910 + }, + { + "epoch": 6.356771873751498, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 15911 + }, + { + "epoch": 6.357171394326808, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15912 + }, + { + "epoch": 6.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15913 + }, + { + "epoch": 6.357970435477427, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15914 + }, + { + "epoch": 6.358369956052737, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15915 + }, + { + "epoch": 6.358769476628046, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15916 + }, + { + "epoch": 6.359168997203356, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15917 + }, + { + "epoch": 6.359568517778666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15918 + }, + { + "epoch": 6.359968038353975, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15919 + }, + { + "epoch": 6.360367558929285, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 15920 + }, + { + "epoch": 6.360767079504594, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15921 + }, + { + "epoch": 6.361166600079904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15922 + }, + { + "epoch": 6.361566120655214, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15923 + }, + { + "epoch": 6.361965641230523, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15924 + }, + { + "epoch": 6.362365161805833, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15925 + }, + { + "epoch": 6.362764682381143, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15926 + }, + { + "epoch": 6.363164202956452, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15927 + }, + { + "epoch": 6.363563723531762, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15928 + }, + { + "epoch": 6.3639632441070715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15929 + }, + { + "epoch": 6.364362764682381, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15930 + }, + { + "epoch": 6.364762285257691, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15931 + }, + { + "epoch": 6.365161805833001, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15932 + }, + { + "epoch": 6.36556132640831, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15933 + }, + { + "epoch": 6.3659608469836195, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15934 + }, + { + "epoch": 6.366360367558929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15935 + }, + { + "epoch": 6.366759888134239, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.297, + "step": 15936 + }, + { + "epoch": 6.367159408709549, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15937 + }, + { + "epoch": 6.367558929284858, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15938 + }, + { + "epoch": 6.367958449860168, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15939 + }, + { + "epoch": 6.368357970435477, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15940 + }, + { + "epoch": 6.368757491010787, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15941 + }, + { + "epoch": 6.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 15942 + }, + { + "epoch": 6.369556532161407, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15943 + }, + { + "epoch": 6.369956052736716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15944 + }, + { + "epoch": 6.370355573312025, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15945 + }, + { + "epoch": 6.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15946 + }, + { + "epoch": 6.371154614462645, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15947 + }, + { + "epoch": 6.371554135037955, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15948 + }, + { + "epoch": 6.3719536556132645, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15949 + }, + { + "epoch": 6.372353176188573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15950 + }, + { + "epoch": 6.372752696763883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15951 + }, + { + "epoch": 6.373152217339193, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15952 + }, + { + "epoch": 6.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 15953 + }, + { + "epoch": 6.373951258489813, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15954 + }, + { + "epoch": 6.3743507790651215, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15955 + }, + { + "epoch": 6.374750299640431, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15956 + }, + { + "epoch": 6.375149820215741, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15957 + }, + { + "epoch": 6.375549340791051, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15958 + }, + { + "epoch": 6.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15959 + }, + { + "epoch": 6.3763483819416695, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15960 + }, + { + "epoch": 6.376747902516979, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 15961 + }, + { + "epoch": 6.377147423092289, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15962 + }, + { + "epoch": 6.377546943667599, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15963 + }, + { + "epoch": 6.377946464242909, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15964 + }, + { + "epoch": 6.3783459848182185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15965 + }, + { + "epoch": 6.378745505393527, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15966 + }, + { + "epoch": 6.379145025968837, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 15967 + }, + { + "epoch": 6.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15968 + }, + { + "epoch": 6.379944067119457, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15969 + }, + { + "epoch": 6.3803435876947665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15970 + }, + { + "epoch": 6.380743108270076, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15971 + }, + { + "epoch": 6.381142628845385, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15972 + }, + { + "epoch": 6.381542149420695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15973 + }, + { + "epoch": 6.381941669996005, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15974 + }, + { + "epoch": 6.3823411905713145, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15975 + }, + { + "epoch": 6.382740711146624, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15976 + }, + { + "epoch": 6.383140231721933, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15977 + }, + { + "epoch": 6.383539752297243, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15978 + }, + { + "epoch": 6.383939272872553, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15979 + }, + { + "epoch": 6.384338793447863, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15980 + }, + { + "epoch": 6.384738314023172, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15981 + }, + { + "epoch": 6.385137834598482, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15982 + }, + { + "epoch": 6.385537355173791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15983 + }, + { + "epoch": 6.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15984 + }, + { + "epoch": 6.386336396324411, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15985 + }, + { + "epoch": 6.38673591689972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15986 + }, + { + "epoch": 6.38713543747503, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15987 + }, + { + "epoch": 6.38753495805034, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 15988 + }, + { + "epoch": 6.387934478625649, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15989 + }, + { + "epoch": 6.388333999200959, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15990 + }, + { + "epoch": 6.3887335197762685, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15991 + }, + { + "epoch": 6.389133040351578, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15992 + }, + { + "epoch": 6.389532560926888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15993 + }, + { + "epoch": 6.389932081502197, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15994 + }, + { + "epoch": 6.390331602077507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15995 + }, + { + "epoch": 6.3907311226528165, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15996 + }, + { + "epoch": 6.391130643228126, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15997 + }, + { + "epoch": 6.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15998 + }, + { + "epoch": 6.391929684378746, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15999 + }, + { + "epoch": 6.392329204954055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16000 + }, + { + "epoch": 6.3927287255293646, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16001 + }, + { + "epoch": 6.393128246104674, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 16002 + }, + { + "epoch": 6.393527766679984, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16003 + }, + { + "epoch": 6.393927287255294, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16004 + }, + { + "epoch": 6.394326807830604, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16005 + }, + { + "epoch": 6.394726328405913, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16006 + }, + { + "epoch": 6.395125848981222, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16007 + }, + { + "epoch": 6.395525369556532, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16008 + }, + { + "epoch": 6.395924890131842, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16009 + }, + { + "epoch": 6.396324410707152, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 16010 + }, + { + "epoch": 6.396723931282461, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16011 + }, + { + "epoch": 6.39712345185777, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16012 + }, + { + "epoch": 6.39752297243308, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16013 + }, + { + "epoch": 6.39792249300839, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16014 + }, + { + "epoch": 6.3983220135837, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16015 + }, + { + "epoch": 6.39872153415901, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16016 + }, + { + "epoch": 6.3991210547343185, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16017 + }, + { + "epoch": 6.399520575309628, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16018 + }, + { + "epoch": 6.399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16019 + }, + { + "epoch": 6.400319616460248, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16020 + }, + { + "epoch": 6.400719137035558, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16021 + }, + { + "epoch": 6.401118657610867, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16022 + }, + { + "epoch": 6.401518178186176, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16023 + }, + { + "epoch": 6.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16024 + }, + { + "epoch": 6.402317219336796, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16025 + }, + { + "epoch": 6.402716739912106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16026 + }, + { + "epoch": 6.4031162604874154, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16027 + }, + { + "epoch": 6.403515781062724, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16028 + }, + { + "epoch": 6.403915301638034, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16029 + }, + { + "epoch": 6.404314822213344, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16030 + }, + { + "epoch": 6.404714342788654, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16031 + }, + { + "epoch": 6.4051138633639635, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16032 + }, + { + "epoch": 6.405513383939273, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16033 + }, + { + "epoch": 6.405912904514582, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16034 + }, + { + "epoch": 6.406312425089892, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16035 + }, + { + "epoch": 6.406711945665202, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16036 + }, + { + "epoch": 6.4071114662405115, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16037 + }, + { + "epoch": 6.407510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16038 + }, + { + "epoch": 6.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16039 + }, + { + "epoch": 6.40831002796644, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16040 + }, + { + "epoch": 6.40870954854175, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16041 + }, + { + "epoch": 6.40910906911706, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16042 + }, + { + "epoch": 6.409508589692369, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16043 + }, + { + "epoch": 6.409908110267679, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16044 + }, + { + "epoch": 6.410307630842988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16045 + }, + { + "epoch": 6.410707151418298, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16046 + }, + { + "epoch": 6.411106671993608, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16047 + }, + { + "epoch": 6.411506192568917, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16048 + }, + { + "epoch": 6.411905713144227, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16049 + }, + { + "epoch": 6.412305233719536, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16050 + }, + { + "epoch": 6.412704754294846, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16051 + }, + { + "epoch": 6.413104274870156, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16052 + }, + { + "epoch": 6.4135037954454654, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16053 + }, + { + "epoch": 6.413903316020775, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16054 + }, + { + "epoch": 6.414302836596085, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16055 + }, + { + "epoch": 6.414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16056 + }, + { + "epoch": 6.415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16057 + }, + { + "epoch": 6.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16058 + }, + { + "epoch": 6.415900918897323, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16059 + }, + { + "epoch": 6.416300439472633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16060 + }, + { + "epoch": 6.416699960047943, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16061 + }, + { + "epoch": 6.417099480623252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 16062 + }, + { + "epoch": 6.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16063 + }, + { + "epoch": 6.417898521773871, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 16064 + }, + { + "epoch": 6.418298042349181, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16065 + }, + { + "epoch": 6.418697562924491, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16066 + }, + { + "epoch": 6.4190970834998, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16067 + }, + { + "epoch": 6.41949660407511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16068 + }, + { + "epoch": 6.419896124650419, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16069 + }, + { + "epoch": 6.420295645225729, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16070 + }, + { + "epoch": 6.420695165801039, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16071 + }, + { + "epoch": 6.421094686376349, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16072 + }, + { + "epoch": 6.421494206951658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16073 + }, + { + "epoch": 6.421893727526967, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16074 + }, + { + "epoch": 6.422293248102277, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16075 + }, + { + "epoch": 6.422692768677587, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16076 + }, + { + "epoch": 6.423092289252897, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16077 + }, + { + "epoch": 6.4234918098282066, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16078 + }, + { + "epoch": 6.4238913304035155, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16079 + }, + { + "epoch": 6.424290850978825, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16080 + }, + { + "epoch": 6.424690371554135, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16081 + }, + { + "epoch": 6.425089892129445, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16082 + }, + { + "epoch": 6.425489412704755, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16083 + }, + { + "epoch": 6.4258889332800635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16084 + }, + { + "epoch": 6.426288453855373, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16085 + }, + { + "epoch": 6.426687974430683, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16086 + }, + { + "epoch": 6.427087495005993, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16087 + }, + { + "epoch": 6.427487015581303, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16088 + }, + { + "epoch": 6.427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16089 + }, + { + "epoch": 6.428286056731921, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16090 + }, + { + "epoch": 6.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16091 + }, + { + "epoch": 6.429085097882541, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16092 + }, + { + "epoch": 6.429484618457851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16093 + }, + { + "epoch": 6.4298841390331605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16094 + }, + { + "epoch": 6.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16095 + }, + { + "epoch": 6.430683180183779, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16096 + }, + { + "epoch": 6.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16097 + }, + { + "epoch": 6.431482221334399, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16098 + }, + { + "epoch": 6.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16099 + }, + { + "epoch": 6.432281262485018, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16100 + }, + { + "epoch": 6.432680783060327, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16101 + }, + { + "epoch": 6.433080303635637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16102 + }, + { + "epoch": 6.433479824210947, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 16103 + }, + { + "epoch": 6.433879344786257, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16104 + }, + { + "epoch": 6.434278865361566, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16105 + }, + { + "epoch": 6.434678385936876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16106 + }, + { + "epoch": 6.435077906512185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16107 + }, + { + "epoch": 6.435477427087495, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16108 + }, + { + "epoch": 6.435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16109 + }, + { + "epoch": 6.436276468238114, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16110 + }, + { + "epoch": 6.436675988813424, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16111 + }, + { + "epoch": 6.437075509388734, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16112 + }, + { + "epoch": 6.437475029964043, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16113 + }, + { + "epoch": 6.437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16114 + }, + { + "epoch": 6.438274071114662, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 16115 + }, + { + "epoch": 6.438673591689972, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16116 + }, + { + "epoch": 6.439073112265282, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 16117 + }, + { + "epoch": 6.439472632840591, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16118 + }, + { + "epoch": 6.439872153415901, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16119 + }, + { + "epoch": 6.4402716739912105, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16120 + }, + { + "epoch": 6.44067119456652, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16121 + }, + { + "epoch": 6.44107071514183, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16122 + }, + { + "epoch": 6.44147023571714, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16123 + }, + { + "epoch": 6.441869756292449, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16124 + }, + { + "epoch": 6.4422692768677585, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16125 + }, + { + "epoch": 6.442668797443068, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16126 + }, + { + "epoch": 6.443068318018378, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16127 + }, + { + "epoch": 6.443467838593688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16128 + }, + { + "epoch": 6.443867359168998, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16129 + }, + { + "epoch": 6.444266879744307, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16130 + }, + { + "epoch": 6.444666400319616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16131 + }, + { + "epoch": 6.445065920894926, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16132 + }, + { + "epoch": 6.445465441470236, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16133 + }, + { + "epoch": 6.445864962045546, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16134 + }, + { + "epoch": 6.446264482620855, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 16135 + }, + { + "epoch": 6.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16136 + }, + { + "epoch": 6.447063523771474, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16137 + }, + { + "epoch": 6.447463044346784, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16138 + }, + { + "epoch": 6.447862564922094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16139 + }, + { + "epoch": 6.448262085497403, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16140 + }, + { + "epoch": 6.448661606072712, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16141 + }, + { + "epoch": 6.449061126648022, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16142 + }, + { + "epoch": 6.449460647223332, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16143 + }, + { + "epoch": 6.449860167798642, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16144 + }, + { + "epoch": 6.450259688373952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16145 + }, + { + "epoch": 6.4506592089492605, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16146 + }, + { + "epoch": 6.45105872952457, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16147 + }, + { + "epoch": 6.45145825009988, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16148 + }, + { + "epoch": 6.45185777067519, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16149 + }, + { + "epoch": 6.4522572912505, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16150 + }, + { + "epoch": 6.452656811825809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16151 + }, + { + "epoch": 6.453056332401118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16152 + }, + { + "epoch": 6.453455852976428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16153 + }, + { + "epoch": 6.453855373551738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16154 + }, + { + "epoch": 6.454254894127048, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16155 + }, + { + "epoch": 6.4546544147023575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16156 + }, + { + "epoch": 6.455053935277666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16157 + }, + { + "epoch": 6.455453455852976, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16158 + }, + { + "epoch": 6.455852976428286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16159 + }, + { + "epoch": 6.456252497003596, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16160 + }, + { + "epoch": 6.4566520175789055, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16161 + }, + { + "epoch": 6.457051538154215, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16162 + }, + { + "epoch": 6.457451058729524, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16163 + }, + { + "epoch": 6.457850579304834, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16164 + }, + { + "epoch": 6.458250099880144, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16165 + }, + { + "epoch": 6.4586496204554535, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16166 + }, + { + "epoch": 6.459049141030763, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16167 + }, + { + "epoch": 6.459448661606073, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16168 + }, + { + "epoch": 6.459848182181382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16169 + }, + { + "epoch": 6.460247702756692, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 16170 + }, + { + "epoch": 6.460647223332002, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16171 + }, + { + "epoch": 6.461046743907311, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16172 + }, + { + "epoch": 6.461446264482621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16173 + }, + { + "epoch": 6.46184578505793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16174 + }, + { + "epoch": 6.46224530563324, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16175 + }, + { + "epoch": 6.46264482620855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16176 + }, + { + "epoch": 6.463044346783859, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16177 + }, + { + "epoch": 6.463443867359169, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16178 + }, + { + "epoch": 6.463843387934479, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16179 + }, + { + "epoch": 6.464242908509788, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16180 + }, + { + "epoch": 6.464642429085098, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16181 + }, + { + "epoch": 6.4650419496604075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16182 + }, + { + "epoch": 6.465441470235717, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16183 + }, + { + "epoch": 6.465840990811027, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16184 + }, + { + "epoch": 6.466240511386337, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16185 + }, + { + "epoch": 6.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16186 + }, + { + "epoch": 6.4670395525369555, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16187 + }, + { + "epoch": 6.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 16188 + }, + { + "epoch": 6.467838593687575, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16189 + }, + { + "epoch": 6.468238114262885, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16190 + }, + { + "epoch": 6.468637634838194, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16191 + }, + { + "epoch": 6.4690371554135035, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16192 + }, + { + "epoch": 6.469436675988813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16193 + }, + { + "epoch": 6.469836196564123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16194 + }, + { + "epoch": 6.470235717139433, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16195 + }, + { + "epoch": 6.470635237714743, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16196 + }, + { + "epoch": 6.471034758290052, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16197 + }, + { + "epoch": 6.471434278865361, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16198 + }, + { + "epoch": 6.471833799440671, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16199 + }, + { + "epoch": 6.472233320015981, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16200 + }, + { + "epoch": 6.472632840591291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.2986, + "step": 16201 + }, + { + "epoch": 6.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16202 + }, + { + "epoch": 6.473431881741909, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16203 + }, + { + "epoch": 6.473831402317219, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16204 + }, + { + "epoch": 6.474230922892529, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16205 + }, + { + "epoch": 6.474630443467839, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16206 + }, + { + "epoch": 6.475029964043149, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16207 + }, + { + "epoch": 6.4754294846184575, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16208 + }, + { + "epoch": 6.475829005193767, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16209 + }, + { + "epoch": 6.476228525769077, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16210 + }, + { + "epoch": 6.476628046344387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16211 + }, + { + "epoch": 6.477027566919697, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16212 + }, + { + "epoch": 6.477427087495006, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16213 + }, + { + "epoch": 6.477826608070315, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16214 + }, + { + "epoch": 6.478226128645625, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16215 + }, + { + "epoch": 6.478625649220935, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16216 + }, + { + "epoch": 6.479025169796245, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16217 + }, + { + "epoch": 6.479424690371554, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16218 + }, + { + "epoch": 6.479824210946864, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16219 + }, + { + "epoch": 6.480223731522173, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16220 + }, + { + "epoch": 6.480623252097483, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16221 + }, + { + "epoch": 6.481022772672793, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16222 + }, + { + "epoch": 6.4814222932481025, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16223 + }, + { + "epoch": 6.481821813823412, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16224 + }, + { + "epoch": 6.482221334398721, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16225 + }, + { + "epoch": 6.482620854974031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16226 + }, + { + "epoch": 6.483020375549341, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16227 + }, + { + "epoch": 6.4834198961246505, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16228 + }, + { + "epoch": 6.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16229 + }, + { + "epoch": 6.484218937275269, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16230 + }, + { + "epoch": 6.484618457850579, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16231 + }, + { + "epoch": 6.485017978425889, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16232 + }, + { + "epoch": 6.485417499001199, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16233 + }, + { + "epoch": 6.485817019576508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16234 + }, + { + "epoch": 6.486216540151818, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16235 + }, + { + "epoch": 6.486616060727127, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16236 + }, + { + "epoch": 6.487015581302437, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16237 + }, + { + "epoch": 6.487415101877747, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16238 + }, + { + "epoch": 6.487814622453056, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16239 + }, + { + "epoch": 6.488214143028366, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16240 + }, + { + "epoch": 6.488613663603676, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16241 + }, + { + "epoch": 6.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16242 + }, + { + "epoch": 6.489412704754295, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16243 + }, + { + "epoch": 6.4898122253296044, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16244 + }, + { + "epoch": 6.490211745904914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16245 + }, + { + "epoch": 6.490611266480224, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16246 + }, + { + "epoch": 6.491010787055533, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16247 + }, + { + "epoch": 6.491410307630843, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16248 + }, + { + "epoch": 6.4918098282061525, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16249 + }, + { + "epoch": 6.492209348781462, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16250 + }, + { + "epoch": 6.492608869356772, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 16251 + }, + { + "epoch": 6.493008389932082, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16252 + }, + { + "epoch": 6.493407910507391, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16253 + }, + { + "epoch": 6.4938074310827005, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16254 + }, + { + "epoch": 6.49420695165801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16255 + }, + { + "epoch": 6.49460647223332, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16256 + }, + { + "epoch": 6.49500599280863, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16257 + }, + { + "epoch": 6.49540551338394, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16258 + }, + { + "epoch": 6.495805033959249, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16259 + }, + { + "epoch": 6.496204554534558, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16260 + }, + { + "epoch": 6.496604075109868, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16261 + }, + { + "epoch": 6.497003595685178, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16262 + }, + { + "epoch": 6.497403116260488, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16263 + }, + { + "epoch": 6.497802636835797, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16264 + }, + { + "epoch": 6.498202157411106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16265 + }, + { + "epoch": 6.498601677986416, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16266 + }, + { + "epoch": 6.499001198561726, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16267 + }, + { + "epoch": 6.499400719137036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 16268 + }, + { + "epoch": 6.4998002397123456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16269 + }, + { + "epoch": 6.5001997602876544, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16270 + }, + { + "epoch": 6.500599280862964, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16271 + }, + { + "epoch": 6.500998801438274, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16272 + }, + { + "epoch": 6.501398322013584, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16273 + }, + { + "epoch": 6.501797842588894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16274 + }, + { + "epoch": 6.502197363164203, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16275 + }, + { + "epoch": 6.502596883739512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16276 + }, + { + "epoch": 6.502996404314822, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16277 + }, + { + "epoch": 6.503395924890132, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16278 + }, + { + "epoch": 6.503795445465442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16279 + }, + { + "epoch": 6.504194966040751, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16280 + }, + { + "epoch": 6.50459448661606, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.2957, + "step": 16281 + }, + { + "epoch": 6.50499400719137, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16282 + }, + { + "epoch": 6.50539352776668, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16283 + }, + { + "epoch": 6.50579304834199, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 16284 + }, + { + "epoch": 6.5061925689172995, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16285 + }, + { + "epoch": 6.506592089492609, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16286 + }, + { + "epoch": 6.506991610067918, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16287 + }, + { + "epoch": 6.507391130643228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16288 + }, + { + "epoch": 6.507790651218538, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16289 + }, + { + "epoch": 6.5081901717938475, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 16290 + }, + { + "epoch": 6.508589692369157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16291 + }, + { + "epoch": 6.508989212944467, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16292 + }, + { + "epoch": 6.509388733519776, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16293 + }, + { + "epoch": 6.509788254095086, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16294 + }, + { + "epoch": 6.5101877746703956, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16295 + }, + { + "epoch": 6.510587295245705, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 16296 + }, + { + "epoch": 6.510986815821015, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16297 + }, + { + "epoch": 6.511386336396324, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16298 + }, + { + "epoch": 6.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16299 + }, + { + "epoch": 6.512185377546944, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16300 + }, + { + "epoch": 6.512584898122253, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16301 + }, + { + "epoch": 6.512984418697563, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16302 + }, + { + "epoch": 6.513383939272872, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16303 + }, + { + "epoch": 6.513783459848182, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16304 + }, + { + "epoch": 6.514182980423492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16305 + }, + { + "epoch": 6.514582500998801, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16306 + }, + { + "epoch": 6.514982021574111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16307 + }, + { + "epoch": 6.515381542149421, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16308 + }, + { + "epoch": 6.515781062724731, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16309 + }, + { + "epoch": 6.51618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16310 + }, + { + "epoch": 6.5165801038753495, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16311 + }, + { + "epoch": 6.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16312 + }, + { + "epoch": 6.517379145025969, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16313 + }, + { + "epoch": 6.517778665601279, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16314 + }, + { + "epoch": 6.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16315 + }, + { + "epoch": 6.5185777067518975, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16316 + }, + { + "epoch": 6.518977227327207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16317 + }, + { + "epoch": 6.519376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 16318 + }, + { + "epoch": 6.519776268477827, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16319 + }, + { + "epoch": 6.520175789053136, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16320 + }, + { + "epoch": 6.520575309628446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16321 + }, + { + "epoch": 6.520974830203755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16322 + }, + { + "epoch": 6.521374350779065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16323 + }, + { + "epoch": 6.521773871354375, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16324 + }, + { + "epoch": 6.522173391929685, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16325 + }, + { + "epoch": 6.5225729125049945, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16326 + }, + { + "epoch": 6.522972433080303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16327 + }, + { + "epoch": 6.523371953655613, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16328 + }, + { + "epoch": 6.523771474230923, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 16329 + }, + { + "epoch": 6.524170994806233, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16330 + }, + { + "epoch": 6.5245705153815425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16331 + }, + { + "epoch": 6.524970035956851, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16332 + }, + { + "epoch": 6.525369556532161, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 16333 + }, + { + "epoch": 6.525769077107471, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16334 + }, + { + "epoch": 6.526168597682781, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16335 + }, + { + "epoch": 6.526568118258091, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16336 + }, + { + "epoch": 6.5269676388333995, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16337 + }, + { + "epoch": 6.527367159408709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16338 + }, + { + "epoch": 6.527766679984019, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16339 + }, + { + "epoch": 6.528166200559329, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16340 + }, + { + "epoch": 6.528565721134639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16341 + }, + { + "epoch": 6.528965241709948, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16342 + }, + { + "epoch": 6.529364762285257, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16343 + }, + { + "epoch": 6.529764282860567, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16344 + }, + { + "epoch": 6.530163803435877, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16345 + }, + { + "epoch": 6.530563324011187, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 16346 + }, + { + "epoch": 6.5309628445864965, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 16347 + }, + { + "epoch": 6.531362365161806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16348 + }, + { + "epoch": 6.531761885737115, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16349 + }, + { + "epoch": 6.532161406312425, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16350 + }, + { + "epoch": 6.532560926887735, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16351 + }, + { + "epoch": 6.5329604474630445, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16352 + }, + { + "epoch": 6.533359968038354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16353 + }, + { + "epoch": 6.533759488613663, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 16354 + }, + { + "epoch": 6.534159009188973, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16355 + }, + { + "epoch": 6.534558529764283, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16356 + }, + { + "epoch": 6.5349580503395925, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16357 + }, + { + "epoch": 6.535357570914902, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 16358 + }, + { + "epoch": 6.535757091490212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16359 + }, + { + "epoch": 6.536156612065521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16360 + }, + { + "epoch": 6.536556132640831, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16361 + }, + { + "epoch": 6.536955653216141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 16362 + }, + { + "epoch": 6.53735517379145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16363 + }, + { + "epoch": 6.53775469436676, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16364 + }, + { + "epoch": 6.53815421494207, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 16365 + }, + { + "epoch": 6.538553735517379, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16366 + }, + { + "epoch": 6.538953256092689, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16367 + }, + { + "epoch": 6.539352776667998, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16368 + }, + { + "epoch": 6.539752297243308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16369 + }, + { + "epoch": 6.540151817818618, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16370 + }, + { + "epoch": 6.540551338393927, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16371 + }, + { + "epoch": 6.540950858969237, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16372 + }, + { + "epoch": 6.5413503795445465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16373 + }, + { + "epoch": 6.541749900119856, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16374 + }, + { + "epoch": 6.542149420695166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16375 + }, + { + "epoch": 6.542548941270476, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16376 + }, + { + "epoch": 6.542948461845785, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16377 + }, + { + "epoch": 6.5433479824210945, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16378 + }, + { + "epoch": 6.543747502996404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16379 + }, + { + "epoch": 6.544147023571714, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16380 + }, + { + "epoch": 6.544546544147024, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16381 + }, + { + "epoch": 6.544946064722334, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16382 + }, + { + "epoch": 6.5453455852976425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16383 + }, + { + "epoch": 6.545745105872952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16384 + }, + { + "epoch": 6.546144626448262, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16385 + }, + { + "epoch": 6.546544147023572, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16386 + }, + { + "epoch": 6.546943667598882, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16387 + }, + { + "epoch": 6.547343188174191, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 16388 + }, + { + "epoch": 6.5477427087495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16389 + }, + { + "epoch": 6.54814222932481, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16390 + }, + { + "epoch": 6.54854174990012, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16391 + }, + { + "epoch": 6.54894127047543, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16392 + }, + { + "epoch": 6.549340791050739, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16393 + }, + { + "epoch": 6.549740311626048, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16394 + }, + { + "epoch": 6.550139832201358, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16395 + }, + { + "epoch": 6.550539352776668, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16396 + }, + { + "epoch": 6.550938873351978, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 16397 + }, + { + "epoch": 6.551338393927288, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16398 + }, + { + "epoch": 6.551737914502597, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16399 + }, + { + "epoch": 6.552137435077906, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16400 + }, + { + "epoch": 6.552536955653216, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16401 + }, + { + "epoch": 6.552936476228526, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16402 + }, + { + "epoch": 6.553335996803836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16403 + }, + { + "epoch": 6.553735517379145, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 16404 + }, + { + "epoch": 6.554135037954454, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16405 + }, + { + "epoch": 6.554534558529764, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16406 + }, + { + "epoch": 6.554934079105074, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16407 + }, + { + "epoch": 6.555333599680384, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16408 + }, + { + "epoch": 6.555733120255693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16409 + }, + { + "epoch": 6.556132640831002, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16410 + }, + { + "epoch": 6.556532161406312, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16411 + }, + { + "epoch": 6.556931681981622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16412 + }, + { + "epoch": 6.557331202556932, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16413 + }, + { + "epoch": 6.5577307231322415, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16414 + }, + { + "epoch": 6.558130243707551, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16415 + }, + { + "epoch": 6.558529764282861, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16416 + }, + { + "epoch": 6.55892928485817, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16417 + }, + { + "epoch": 6.55932880543348, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16418 + }, + { + "epoch": 6.5597283260087895, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16419 + }, + { + "epoch": 6.560127846584099, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16420 + }, + { + "epoch": 6.560527367159409, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16421 + }, + { + "epoch": 6.560926887734718, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16422 + }, + { + "epoch": 6.561326408310028, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16423 + }, + { + "epoch": 6.561725928885338, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16424 + }, + { + "epoch": 6.562125449460647, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16425 + }, + { + "epoch": 6.562524970035957, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16426 + }, + { + "epoch": 6.562924490611266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16427 + }, + { + "epoch": 6.563324011186576, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16428 + }, + { + "epoch": 6.563723531761886, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16429 + }, + { + "epoch": 6.564123052337195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16430 + }, + { + "epoch": 6.564522572912505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16431 + }, + { + "epoch": 6.564922093487815, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16432 + }, + { + "epoch": 6.565321614063124, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16433 + }, + { + "epoch": 6.565721134638434, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16434 + }, + { + "epoch": 6.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16435 + }, + { + "epoch": 6.566520175789053, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16436 + }, + { + "epoch": 6.566919696364363, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16437 + }, + { + "epoch": 6.567319216939673, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16438 + }, + { + "epoch": 6.567718737514982, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 16439 + }, + { + "epoch": 6.5681182580902915, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16440 + }, + { + "epoch": 6.568517778665601, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16441 + }, + { + "epoch": 6.568917299240911, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16442 + }, + { + "epoch": 6.569316819816221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16443 + }, + { + "epoch": 6.56971634039153, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16444 + }, + { + "epoch": 6.5701158609668395, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16445 + }, + { + "epoch": 6.570515381542149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16446 + }, + { + "epoch": 6.570914902117459, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 16447 + }, + { + "epoch": 6.571314422692769, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16448 + }, + { + "epoch": 6.571713943268079, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16449 + }, + { + "epoch": 6.572113463843388, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16450 + }, + { + "epoch": 6.572512984418697, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16451 + }, + { + "epoch": 6.572912504994007, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16452 + }, + { + "epoch": 6.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16453 + }, + { + "epoch": 6.573711546144627, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16454 + }, + { + "epoch": 6.5741110667199365, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16455 + }, + { + "epoch": 6.574510587295245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16456 + }, + { + "epoch": 6.574910107870555, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16457 + }, + { + "epoch": 6.575309628445865, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16458 + }, + { + "epoch": 6.575709149021175, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16459 + }, + { + "epoch": 6.5761086695964845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16460 + }, + { + "epoch": 6.5765081901717934, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16461 + }, + { + "epoch": 6.576907710747103, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16462 + }, + { + "epoch": 6.577307231322413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16463 + }, + { + "epoch": 6.577706751897723, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16464 + }, + { + "epoch": 6.578106272473033, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16465 + }, + { + "epoch": 6.578505793048342, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16466 + }, + { + "epoch": 6.578905313623651, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16467 + }, + { + "epoch": 6.579304834198961, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16468 + }, + { + "epoch": 6.579704354774271, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16469 + }, + { + "epoch": 6.580103875349581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16470 + }, + { + "epoch": 6.58050339592489, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16471 + }, + { + "epoch": 6.5809029165002, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16472 + }, + { + "epoch": 6.581302437075509, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16473 + }, + { + "epoch": 6.581701957650819, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16474 + }, + { + "epoch": 6.582101478226129, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16475 + }, + { + "epoch": 6.5825009988014385, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16476 + }, + { + "epoch": 6.582900519376748, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16477 + }, + { + "epoch": 6.583300039952057, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16478 + }, + { + "epoch": 6.583699560527367, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16479 + }, + { + "epoch": 6.584099081102677, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16480 + }, + { + "epoch": 6.5844986016779865, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16481 + }, + { + "epoch": 6.584898122253296, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16482 + }, + { + "epoch": 6.585297642828605, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16483 + }, + { + "epoch": 6.585697163403915, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16484 + }, + { + "epoch": 6.586096683979225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16485 + }, + { + "epoch": 6.5864962045545346, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16486 + }, + { + "epoch": 6.586895725129844, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16487 + }, + { + "epoch": 6.587295245705154, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16488 + }, + { + "epoch": 6.587694766280464, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16489 + }, + { + "epoch": 6.588094286855773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16490 + }, + { + "epoch": 6.588493807431083, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16491 + }, + { + "epoch": 6.588893328006392, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16492 + }, + { + "epoch": 6.589292848581702, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16493 + }, + { + "epoch": 6.589692369157012, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16494 + }, + { + "epoch": 6.590091889732321, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16495 + }, + { + "epoch": 6.590491410307631, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16496 + }, + { + "epoch": 6.59089093088294, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16497 + }, + { + "epoch": 6.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16498 + }, + { + "epoch": 6.59168997203356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16499 + }, + { + "epoch": 6.592089492608869, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16500 + }, + { + "epoch": 6.592489013184179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16501 + }, + { + "epoch": 6.5928885337594885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16502 + }, + { + "epoch": 6.593288054334798, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16503 + }, + { + "epoch": 6.593687574910108, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16504 + }, + { + "epoch": 6.594087095485418, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16505 + }, + { + "epoch": 6.594486616060728, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16506 + }, + { + "epoch": 6.5948861366360365, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16507 + }, + { + "epoch": 6.595285657211346, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16508 + }, + { + "epoch": 6.595685177786656, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16509 + }, + { + "epoch": 6.596084698361966, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16510 + }, + { + "epoch": 6.596484218937276, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16511 + }, + { + "epoch": 6.5968837395125846, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16512 + }, + { + "epoch": 6.597283260087894, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16513 + }, + { + "epoch": 6.597682780663204, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16514 + }, + { + "epoch": 6.598082301238514, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16515 + }, + { + "epoch": 6.598481821813824, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 16516 + }, + { + "epoch": 6.598881342389133, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 16517 + }, + { + "epoch": 6.599280862964442, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16518 + }, + { + "epoch": 6.599680383539752, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16519 + }, + { + "epoch": 6.600079904115062, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16520 + }, + { + "epoch": 6.600479424690372, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16521 + }, + { + "epoch": 6.6008789452656815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16522 + }, + { + "epoch": 6.60127846584099, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16523 + }, + { + "epoch": 6.6016779864163, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16524 + }, + { + "epoch": 6.60207750699161, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16525 + }, + { + "epoch": 6.60247702756692, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16526 + }, + { + "epoch": 6.60287654814223, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16527 + }, + { + "epoch": 6.603276068717539, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16528 + }, + { + "epoch": 6.603675589292848, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16529 + }, + { + "epoch": 6.604075109868158, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16530 + }, + { + "epoch": 6.604474630443468, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16531 + }, + { + "epoch": 6.604874151018778, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16532 + }, + { + "epoch": 6.605273671594087, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16533 + }, + { + "epoch": 6.605673192169396, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16534 + }, + { + "epoch": 6.606072712744706, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16535 + }, + { + "epoch": 6.606472233320016, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16536 + }, + { + "epoch": 6.606871753895326, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16537 + }, + { + "epoch": 6.6072712744706354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16538 + }, + { + "epoch": 6.607670795045945, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16539 + }, + { + "epoch": 6.608070315621254, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16540 + }, + { + "epoch": 6.608469836196564, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16541 + }, + { + "epoch": 6.608869356771874, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 16542 + }, + { + "epoch": 6.6092688773471835, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16543 + }, + { + "epoch": 6.609668397922493, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16544 + }, + { + "epoch": 6.610067918497803, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16545 + }, + { + "epoch": 6.610467439073112, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16546 + }, + { + "epoch": 6.610866959648422, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16547 + }, + { + "epoch": 6.6112664802237315, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16548 + }, + { + "epoch": 6.611666000799041, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16549 + }, + { + "epoch": 6.612065521374351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16550 + }, + { + "epoch": 6.61246504194966, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16551 + }, + { + "epoch": 6.61286456252497, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16552 + }, + { + "epoch": 6.61326408310028, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16553 + }, + { + "epoch": 6.613663603675589, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16554 + }, + { + "epoch": 6.614063124250899, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16555 + }, + { + "epoch": 6.614462644826209, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16556 + }, + { + "epoch": 6.614862165401518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16557 + }, + { + "epoch": 6.615261685976828, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16558 + }, + { + "epoch": 6.615661206552137, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16559 + }, + { + "epoch": 6.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16560 + }, + { + "epoch": 6.616460247702757, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16561 + }, + { + "epoch": 6.616859768278067, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16562 + }, + { + "epoch": 6.617259288853376, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16563 + }, + { + "epoch": 6.6176588094286855, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16564 + }, + { + "epoch": 6.618058330003995, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16565 + }, + { + "epoch": 6.618457850579305, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16566 + }, + { + "epoch": 6.618857371154615, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16567 + }, + { + "epoch": 6.619256891729924, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16568 + }, + { + "epoch": 6.6196564123052335, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16569 + }, + { + "epoch": 6.620055932880543, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16570 + }, + { + "epoch": 6.620455453455853, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16571 + }, + { + "epoch": 6.620854974031163, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16572 + }, + { + "epoch": 6.621254494606472, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 16573 + }, + { + "epoch": 6.6216540151817815, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16574 + }, + { + "epoch": 6.622053535757091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 16575 + }, + { + "epoch": 6.622453056332401, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16576 + }, + { + "epoch": 6.622852576907711, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16577 + }, + { + "epoch": 6.623252097483021, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16578 + }, + { + "epoch": 6.6236516180583305, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16579 + }, + { + "epoch": 6.624051138633639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16580 + }, + { + "epoch": 6.624450659208949, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16581 + }, + { + "epoch": 6.624850179784259, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16582 + }, + { + "epoch": 6.625249700359569, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16583 + }, + { + "epoch": 6.6256492209348785, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16584 + }, + { + "epoch": 6.626048741510187, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16585 + }, + { + "epoch": 6.626448262085497, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16586 + }, + { + "epoch": 6.626847782660807, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16587 + }, + { + "epoch": 6.627247303236117, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16588 + }, + { + "epoch": 6.627646823811427, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16589 + }, + { + "epoch": 6.6280463443867355, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16590 + }, + { + "epoch": 6.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16591 + }, + { + "epoch": 6.628845385537355, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16592 + }, + { + "epoch": 6.629244906112665, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16593 + }, + { + "epoch": 6.629644426687975, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16594 + }, + { + "epoch": 6.630043947263284, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16595 + }, + { + "epoch": 6.630443467838594, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16596 + }, + { + "epoch": 6.630842988413903, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16597 + }, + { + "epoch": 6.631242508989213, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16598 + }, + { + "epoch": 6.631642029564523, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 16599 + }, + { + "epoch": 6.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16600 + }, + { + "epoch": 6.632441070715142, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16601 + }, + { + "epoch": 6.632840591290451, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16602 + }, + { + "epoch": 6.633240111865761, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16603 + }, + { + "epoch": 6.633639632441071, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16604 + }, + { + "epoch": 6.6340391530163805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16605 + }, + { + "epoch": 6.63443867359169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16606 + }, + { + "epoch": 6.634838194166999, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16607 + }, + { + "epoch": 6.635237714742309, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16608 + }, + { + "epoch": 6.635637235317619, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 16609 + }, + { + "epoch": 6.6360367558929285, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16610 + }, + { + "epoch": 6.636436276468238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16611 + }, + { + "epoch": 6.636835797043548, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16612 + }, + { + "epoch": 6.637235317618857, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16613 + }, + { + "epoch": 6.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16614 + }, + { + "epoch": 6.638034358769477, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16615 + }, + { + "epoch": 6.638433879344786, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16616 + }, + { + "epoch": 6.638833399920096, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 16617 + }, + { + "epoch": 6.639232920495406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16618 + }, + { + "epoch": 6.639632441070715, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16619 + }, + { + "epoch": 6.640031961646025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16620 + }, + { + "epoch": 6.640431482221334, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16621 + }, + { + "epoch": 6.640831002796644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16622 + }, + { + "epoch": 6.641230523371954, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16623 + }, + { + "epoch": 6.641630043947263, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16624 + }, + { + "epoch": 6.642029564522573, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16625 + }, + { + "epoch": 6.642429085097882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16626 + }, + { + "epoch": 6.642828605673192, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16627 + }, + { + "epoch": 6.643228126248502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16628 + }, + { + "epoch": 6.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16629 + }, + { + "epoch": 6.644027167399121, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16630 + }, + { + "epoch": 6.6444266879744305, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16631 + }, + { + "epoch": 6.64482620854974, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16632 + }, + { + "epoch": 6.64522572912505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16633 + }, + { + "epoch": 6.64562524970036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16634 + }, + { + "epoch": 6.64602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16635 + }, + { + "epoch": 6.6464242908509785, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16636 + }, + { + "epoch": 6.646823811426288, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16637 + }, + { + "epoch": 6.647223332001598, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16638 + }, + { + "epoch": 6.647622852576908, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16639 + }, + { + "epoch": 6.648022373152218, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16640 + }, + { + "epoch": 6.648421893727527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16641 + }, + { + "epoch": 6.648821414302836, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16642 + }, + { + "epoch": 6.649220934878146, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16643 + }, + { + "epoch": 6.649620455453456, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16644 + }, + { + "epoch": 6.650019976028766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16645 + }, + { + "epoch": 6.6504194966040755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16646 + }, + { + "epoch": 6.650819017179384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16647 + }, + { + "epoch": 6.651218537754694, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16648 + }, + { + "epoch": 6.651618058330004, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16649 + }, + { + "epoch": 6.652017578905314, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16650 + }, + { + "epoch": 6.6524170994806235, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16651 + }, + { + "epoch": 6.652816620055933, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16652 + }, + { + "epoch": 6.653216140631242, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16653 + }, + { + "epoch": 6.653615661206552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16654 + }, + { + "epoch": 6.654015181781862, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16655 + }, + { + "epoch": 6.654414702357172, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16656 + }, + { + "epoch": 6.654814222932481, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16657 + }, + { + "epoch": 6.65521374350779, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16658 + }, + { + "epoch": 6.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16659 + }, + { + "epoch": 6.65601278465841, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16660 + }, + { + "epoch": 6.65641230523372, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16661 + }, + { + "epoch": 6.656811825809029, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16662 + }, + { + "epoch": 6.657211346384338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16663 + }, + { + "epoch": 6.657610866959648, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16664 + }, + { + "epoch": 6.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16665 + }, + { + "epoch": 6.658409908110268, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16666 + }, + { + "epoch": 6.6588094286855775, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16667 + }, + { + "epoch": 6.659208949260887, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16668 + }, + { + "epoch": 6.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16669 + }, + { + "epoch": 6.660007990411506, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16670 + }, + { + "epoch": 6.660407510986816, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16671 + }, + { + "epoch": 6.6608070315621255, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16672 + }, + { + "epoch": 6.661206552137435, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16673 + }, + { + "epoch": 6.661606072712745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16674 + }, + { + "epoch": 6.662005593288054, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16675 + }, + { + "epoch": 6.662405113863364, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16676 + }, + { + "epoch": 6.6628046344386735, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16677 + }, + { + "epoch": 6.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16678 + }, + { + "epoch": 6.663603675589293, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16679 + }, + { + "epoch": 6.664003196164602, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16680 + }, + { + "epoch": 6.664402716739912, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16681 + }, + { + "epoch": 6.664802237315222, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16682 + }, + { + "epoch": 6.665201757890531, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16683 + }, + { + "epoch": 6.665601278465841, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16684 + }, + { + "epoch": 6.666000799041151, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16685 + }, + { + "epoch": 6.666400319616461, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16686 + }, + { + "epoch": 6.66679984019177, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16687 + }, + { + "epoch": 6.667199360767079, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16688 + }, + { + "epoch": 6.667598881342389, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16689 + }, + { + "epoch": 6.667998401917699, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16690 + }, + { + "epoch": 6.668397922493009, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16691 + }, + { + "epoch": 6.668797443068318, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16692 + }, + { + "epoch": 6.6691969636436275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16693 + }, + { + "epoch": 6.669596484218937, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16694 + }, + { + "epoch": 6.669996004794247, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16695 + }, + { + "epoch": 6.670395525369557, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16696 + }, + { + "epoch": 6.670795045944866, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16697 + }, + { + "epoch": 6.6711945665201755, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16698 + }, + { + "epoch": 6.671594087095485, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16699 + }, + { + "epoch": 6.671993607670795, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16700 + }, + { + "epoch": 6.672393128246105, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16701 + }, + { + "epoch": 6.672792648821415, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16702 + }, + { + "epoch": 6.6731921693967235, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16703 + }, + { + "epoch": 6.673591689972033, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16704 + }, + { + "epoch": 6.673991210547343, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16705 + }, + { + "epoch": 6.674390731122653, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 16706 + }, + { + "epoch": 6.674790251697963, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16707 + }, + { + "epoch": 6.6751897722732725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16708 + }, + { + "epoch": 6.675589292848581, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16709 + }, + { + "epoch": 6.675988813423891, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16710 + }, + { + "epoch": 6.676388333999201, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16711 + }, + { + "epoch": 6.676787854574511, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16712 + }, + { + "epoch": 6.6771873751498205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16713 + }, + { + "epoch": 6.677586895725129, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16714 + }, + { + "epoch": 6.677986416300439, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16715 + }, + { + "epoch": 6.678385936875749, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16716 + }, + { + "epoch": 6.678785457451059, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16717 + }, + { + "epoch": 6.679184978026369, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16718 + }, + { + "epoch": 6.679584498601678, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16719 + }, + { + "epoch": 6.679984019176987, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 16720 + }, + { + "epoch": 6.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16721 + }, + { + "epoch": 6.680783060327607, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16722 + }, + { + "epoch": 6.681182580902917, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16723 + }, + { + "epoch": 6.681582101478226, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16724 + }, + { + "epoch": 6.681981622053536, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16725 + }, + { + "epoch": 6.682381142628845, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16726 + }, + { + "epoch": 6.682780663204155, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16727 + }, + { + "epoch": 6.683180183779465, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16728 + }, + { + "epoch": 6.683579704354774, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16729 + }, + { + "epoch": 6.683979224930084, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16730 + }, + { + "epoch": 6.684378745505393, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16731 + }, + { + "epoch": 6.684778266080703, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16732 + }, + { + "epoch": 6.685177786656013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16733 + }, + { + "epoch": 6.6855773072313225, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16734 + }, + { + "epoch": 6.685976827806632, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16735 + }, + { + "epoch": 6.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16736 + }, + { + "epoch": 6.686775868957251, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16737 + }, + { + "epoch": 6.687175389532561, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 16738 + }, + { + "epoch": 6.6875749101078705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16739 + }, + { + "epoch": 6.68797443068318, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16740 + }, + { + "epoch": 6.68837395125849, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16741 + }, + { + "epoch": 6.6887734718338, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16742 + }, + { + "epoch": 6.689172992409109, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16743 + }, + { + "epoch": 6.689572512984419, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16744 + }, + { + "epoch": 6.689972033559728, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16745 + }, + { + "epoch": 6.690371554135038, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16746 + }, + { + "epoch": 6.690771074710348, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16747 + }, + { + "epoch": 6.691170595285657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16748 + }, + { + "epoch": 6.691570115860967, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16749 + }, + { + "epoch": 6.691969636436276, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16750 + }, + { + "epoch": 6.692369157011586, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 16751 + }, + { + "epoch": 6.692768677586896, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 16752 + }, + { + "epoch": 6.693168198162205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16753 + }, + { + "epoch": 6.693567718737515, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16754 + }, + { + "epoch": 6.6939672393128244, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16755 + }, + { + "epoch": 6.694366759888134, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16756 + }, + { + "epoch": 6.694766280463444, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16757 + }, + { + "epoch": 6.695165801038754, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16758 + }, + { + "epoch": 6.695565321614064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16759 + }, + { + "epoch": 6.6959648421893725, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16760 + }, + { + "epoch": 6.696364362764682, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16761 + }, + { + "epoch": 6.696763883339992, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16762 + }, + { + "epoch": 6.697163403915302, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16763 + }, + { + "epoch": 6.697562924490612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16764 + }, + { + "epoch": 6.6979624450659205, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16765 + }, + { + "epoch": 6.69836196564123, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16766 + }, + { + "epoch": 6.69876148621654, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16767 + }, + { + "epoch": 6.69916100679185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16768 + }, + { + "epoch": 6.69956052736716, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16769 + }, + { + "epoch": 6.699960047942469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16770 + }, + { + "epoch": 6.700359568517778, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16771 + }, + { + "epoch": 6.700759089093088, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16772 + }, + { + "epoch": 6.701158609668398, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16773 + }, + { + "epoch": 6.701558130243708, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16774 + }, + { + "epoch": 6.7019576508190175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16775 + }, + { + "epoch": 6.702357171394327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16776 + }, + { + "epoch": 6.702756691969636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16777 + }, + { + "epoch": 6.703156212544946, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16778 + }, + { + "epoch": 6.703555733120256, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16779 + }, + { + "epoch": 6.7039552536955656, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16780 + }, + { + "epoch": 6.704354774270875, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16781 + }, + { + "epoch": 6.704754294846184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16782 + }, + { + "epoch": 6.705153815421494, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16783 + }, + { + "epoch": 6.705553335996804, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16784 + }, + { + "epoch": 6.705952856572114, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16785 + }, + { + "epoch": 6.706352377147423, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16786 + }, + { + "epoch": 6.706751897722732, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16787 + }, + { + "epoch": 6.707151418298042, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16788 + }, + { + "epoch": 6.707550938873352, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16789 + }, + { + "epoch": 6.707950459448662, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16790 + }, + { + "epoch": 6.708349980023971, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16791 + }, + { + "epoch": 6.708749500599281, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16792 + }, + { + "epoch": 6.70914902117459, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16793 + }, + { + "epoch": 6.7095485417499, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16794 + }, + { + "epoch": 6.70994806232521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16795 + }, + { + "epoch": 6.7103475829005195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16796 + }, + { + "epoch": 6.710747103475829, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16797 + }, + { + "epoch": 6.711146624051139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16798 + }, + { + "epoch": 6.711546144626448, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16799 + }, + { + "epoch": 6.711945665201758, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16800 + }, + { + "epoch": 6.7123451857770675, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16801 + }, + { + "epoch": 6.712744706352377, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.2968, + "step": 16802 + }, + { + "epoch": 6.713144226927687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16803 + }, + { + "epoch": 6.713543747502996, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16804 + }, + { + "epoch": 6.713943268078306, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16805 + }, + { + "epoch": 6.714342788653616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16806 + }, + { + "epoch": 6.714742309228925, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16807 + }, + { + "epoch": 6.715141829804235, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 16808 + }, + { + "epoch": 6.715541350379545, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16809 + }, + { + "epoch": 6.715940870954854, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 16810 + }, + { + "epoch": 6.716340391530164, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16811 + }, + { + "epoch": 6.716739912105473, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16812 + }, + { + "epoch": 6.717139432680783, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16813 + }, + { + "epoch": 6.717538953256093, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16814 + }, + { + "epoch": 6.717938473831403, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16815 + }, + { + "epoch": 6.718337994406712, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16816 + }, + { + "epoch": 6.718737514982021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16817 + }, + { + "epoch": 6.719137035557331, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16818 + }, + { + "epoch": 6.719536556132641, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16819 + }, + { + "epoch": 6.719936076707951, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16820 + }, + { + "epoch": 6.72033559728326, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16821 + }, + { + "epoch": 6.7207351178585695, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 16822 + }, + { + "epoch": 6.721134638433879, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16823 + }, + { + "epoch": 6.721534159009189, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 16824 + }, + { + "epoch": 6.721933679584499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16825 + }, + { + "epoch": 6.722333200159808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16826 + }, + { + "epoch": 6.7227327207351175, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16827 + }, + { + "epoch": 6.723132241310427, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16828 + }, + { + "epoch": 6.723531761885737, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16829 + }, + { + "epoch": 6.723931282461047, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16830 + }, + { + "epoch": 6.724330803036357, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16831 + }, + { + "epoch": 6.7247303236116664, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16832 + }, + { + "epoch": 6.725129844186975, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16833 + }, + { + "epoch": 6.725529364762285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16834 + }, + { + "epoch": 6.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16835 + }, + { + "epoch": 6.726328405912905, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16836 + }, + { + "epoch": 6.7267279264882145, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16837 + }, + { + "epoch": 6.727127447063523, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16838 + }, + { + "epoch": 6.727526967638833, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16839 + }, + { + "epoch": 6.727926488214143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16840 + }, + { + "epoch": 6.728326008789453, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16841 + }, + { + "epoch": 6.7287255293647625, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16842 + }, + { + "epoch": 6.729125049940071, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16843 + }, + { + "epoch": 6.729524570515381, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16844 + }, + { + "epoch": 6.729924091090691, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16845 + }, + { + "epoch": 6.730323611666001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16846 + }, + { + "epoch": 6.730723132241311, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16847 + }, + { + "epoch": 6.73112265281662, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16848 + }, + { + "epoch": 6.73152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16849 + }, + { + "epoch": 6.731921693967239, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16850 + }, + { + "epoch": 6.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16851 + }, + { + "epoch": 6.732720735117859, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16852 + }, + { + "epoch": 6.733120255693168, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16853 + }, + { + "epoch": 6.733519776268478, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16854 + }, + { + "epoch": 6.733919296843787, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16855 + }, + { + "epoch": 6.734318817419097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16856 + }, + { + "epoch": 6.734718337994407, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16857 + }, + { + "epoch": 6.7351178585697165, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16858 + }, + { + "epoch": 6.735517379145026, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16859 + }, + { + "epoch": 6.735916899720335, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16860 + }, + { + "epoch": 6.736316420295645, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16861 + }, + { + "epoch": 6.736715940870955, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16862 + }, + { + "epoch": 6.7371154614462645, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 16863 + }, + { + "epoch": 6.737514982021574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16864 + }, + { + "epoch": 6.737914502596884, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16865 + }, + { + "epoch": 6.738314023172194, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16866 + }, + { + "epoch": 6.738713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16867 + }, + { + "epoch": 6.7391130643228125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16868 + }, + { + "epoch": 6.739512584898122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16869 + }, + { + "epoch": 6.739912105473432, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16870 + }, + { + "epoch": 6.740311626048742, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16871 + }, + { + "epoch": 6.740711146624051, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16872 + }, + { + "epoch": 6.741110667199361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16873 + }, + { + "epoch": 6.74151018777467, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16874 + }, + { + "epoch": 6.74190970834998, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16875 + }, + { + "epoch": 6.74230922892529, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16876 + }, + { + "epoch": 6.742708749500599, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16877 + }, + { + "epoch": 6.743108270075909, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16878 + }, + { + "epoch": 6.743507790651218, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16879 + }, + { + "epoch": 6.743907311226528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16880 + }, + { + "epoch": 6.744306831801838, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16881 + }, + { + "epoch": 6.744706352377148, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16882 + }, + { + "epoch": 6.745105872952457, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16883 + }, + { + "epoch": 6.7455053935277665, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16884 + }, + { + "epoch": 6.745904914103076, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16885 + }, + { + "epoch": 6.746304434678386, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 16886 + }, + { + "epoch": 6.746703955253696, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16887 + }, + { + "epoch": 6.747103475829006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16888 + }, + { + "epoch": 6.7475029964043145, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16889 + }, + { + "epoch": 6.747902516979624, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16890 + }, + { + "epoch": 6.748302037554934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16891 + }, + { + "epoch": 6.748701558130244, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16892 + }, + { + "epoch": 6.749101078705554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16893 + }, + { + "epoch": 6.7495005992808625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16894 + }, + { + "epoch": 6.749900119856172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16895 + }, + { + "epoch": 6.750299640431482, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16896 + }, + { + "epoch": 6.750699161006792, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16897 + }, + { + "epoch": 6.751098681582102, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16898 + }, + { + "epoch": 6.7514982021574115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.296, + "step": 16899 + }, + { + "epoch": 6.75189772273272, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16900 + }, + { + "epoch": 6.75229724330803, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16901 + }, + { + "epoch": 6.75269676388334, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16902 + }, + { + "epoch": 6.75309628445865, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16903 + }, + { + "epoch": 6.7534958050339595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16904 + }, + { + "epoch": 6.753895325609269, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16905 + }, + { + "epoch": 6.754294846184578, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16906 + }, + { + "epoch": 6.754694366759888, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 16907 + }, + { + "epoch": 6.755093887335198, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16908 + }, + { + "epoch": 6.755493407910508, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16909 + }, + { + "epoch": 6.755892928485817, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 16910 + }, + { + "epoch": 6.756292449061126, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16911 + }, + { + "epoch": 6.756691969636436, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16912 + }, + { + "epoch": 6.757091490211746, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16913 + }, + { + "epoch": 6.757491010787056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16914 + }, + { + "epoch": 6.757890531362365, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16915 + }, + { + "epoch": 6.758290051937674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16916 + }, + { + "epoch": 6.758689572512984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16917 + }, + { + "epoch": 6.759089093088294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16918 + }, + { + "epoch": 6.759488613663604, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16919 + }, + { + "epoch": 6.759888134238913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16920 + }, + { + "epoch": 6.760287654814223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16921 + }, + { + "epoch": 6.760687175389533, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16922 + }, + { + "epoch": 6.761086695964842, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16923 + }, + { + "epoch": 6.761486216540152, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16924 + }, + { + "epoch": 6.7618857371154615, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16925 + }, + { + "epoch": 6.762285257690771, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16926 + }, + { + "epoch": 6.762684778266081, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16927 + }, + { + "epoch": 6.76308429884139, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16928 + }, + { + "epoch": 6.7634838194167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16929 + }, + { + "epoch": 6.7638833399920095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16930 + }, + { + "epoch": 6.764282860567319, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16931 + }, + { + "epoch": 6.764682381142629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16932 + }, + { + "epoch": 6.765081901717938, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16933 + }, + { + "epoch": 6.765481422293248, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16934 + }, + { + "epoch": 6.765880942868558, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16935 + }, + { + "epoch": 6.766280463443867, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16936 + }, + { + "epoch": 6.766679984019177, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16937 + }, + { + "epoch": 6.767079504594487, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16938 + }, + { + "epoch": 6.767479025169797, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16939 + }, + { + "epoch": 6.767878545745106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16940 + }, + { + "epoch": 6.768278066320415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16941 + }, + { + "epoch": 6.768677586895725, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16942 + }, + { + "epoch": 6.769077107471035, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16943 + }, + { + "epoch": 6.769476628046345, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16944 + }, + { + "epoch": 6.769876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16945 + }, + { + "epoch": 6.770275669196963, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16946 + }, + { + "epoch": 6.770675189772273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16947 + }, + { + "epoch": 6.771074710347583, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16948 + }, + { + "epoch": 6.771474230922893, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16949 + }, + { + "epoch": 6.771873751498202, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16950 + }, + { + "epoch": 6.7722732720735115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16951 + }, + { + "epoch": 6.772672792648821, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16952 + }, + { + "epoch": 6.773072313224131, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16953 + }, + { + "epoch": 6.773471833799441, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16954 + }, + { + "epoch": 6.773871354374751, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16955 + }, + { + "epoch": 6.77427087495006, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16956 + }, + { + "epoch": 6.774670395525369, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16957 + }, + { + "epoch": 6.775069916100679, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16958 + }, + { + "epoch": 6.775469436675989, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16959 + }, + { + "epoch": 6.775868957251299, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16960 + }, + { + "epoch": 6.7762684778266085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16961 + }, + { + "epoch": 6.776667998401917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16962 + }, + { + "epoch": 6.777067518977227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16963 + }, + { + "epoch": 6.777467039552537, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 16964 + }, + { + "epoch": 6.777866560127847, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16965 + }, + { + "epoch": 6.7782660807031565, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16966 + }, + { + "epoch": 6.778665601278465, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 16967 + }, + { + "epoch": 6.779065121853775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16968 + }, + { + "epoch": 6.779464642429085, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16969 + }, + { + "epoch": 6.779864163004395, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 16970 + }, + { + "epoch": 6.7802636835797045, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16971 + }, + { + "epoch": 6.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16972 + }, + { + "epoch": 6.781062724730323, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16973 + }, + { + "epoch": 6.781462245305633, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16974 + }, + { + "epoch": 6.781861765880943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16975 + }, + { + "epoch": 6.782261286456253, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16976 + }, + { + "epoch": 6.782660807031562, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16977 + }, + { + "epoch": 6.783060327606872, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16978 + }, + { + "epoch": 6.783459848182181, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16979 + }, + { + "epoch": 6.783859368757491, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16980 + }, + { + "epoch": 6.784258889332801, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16981 + }, + { + "epoch": 6.78465840990811, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16982 + }, + { + "epoch": 6.78505793048342, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16983 + }, + { + "epoch": 6.785457451058729, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16984 + }, + { + "epoch": 6.785856971634039, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16985 + }, + { + "epoch": 6.786256492209349, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16986 + }, + { + "epoch": 6.7866560127846585, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16987 + }, + { + "epoch": 6.787055533359968, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16988 + }, + { + "epoch": 6.787455053935278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16989 + }, + { + "epoch": 6.787854574510587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16990 + }, + { + "epoch": 6.788254095085897, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16991 + }, + { + "epoch": 6.7886536156612065, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16992 + }, + { + "epoch": 6.789053136236516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16993 + }, + { + "epoch": 6.789452656811826, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16994 + }, + { + "epoch": 6.789852177387136, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16995 + }, + { + "epoch": 6.790251697962445, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16996 + }, + { + "epoch": 6.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16997 + }, + { + "epoch": 6.791050739113064, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16998 + }, + { + "epoch": 6.791450259688374, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16999 + }, + { + "epoch": 6.791849780263684, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17000 + }, + { + "epoch": 6.792249300838993, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17001 + }, + { + "epoch": 6.792648821414303, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17002 + }, + { + "epoch": 6.793048341989612, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17003 + }, + { + "epoch": 6.793447862564922, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17004 + }, + { + "epoch": 6.793847383140232, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17005 + }, + { + "epoch": 6.794246903715541, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17006 + }, + { + "epoch": 6.794646424290851, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17007 + }, + { + "epoch": 6.79504594486616, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17008 + }, + { + "epoch": 6.79544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17009 + }, + { + "epoch": 6.79584498601678, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17010 + }, + { + "epoch": 6.79624450659209, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17011 + }, + { + "epoch": 6.7966440271674, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17012 + }, + { + "epoch": 6.7970435477427085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17013 + }, + { + "epoch": 6.797443068318018, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17014 + }, + { + "epoch": 6.797842588893328, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17015 + }, + { + "epoch": 6.798242109468638, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17016 + }, + { + "epoch": 6.798641630043948, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17017 + }, + { + "epoch": 6.7990411506192565, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17018 + }, + { + "epoch": 6.799440671194566, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17019 + }, + { + "epoch": 6.799840191769876, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17020 + }, + { + "epoch": 6.800239712345186, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17021 + }, + { + "epoch": 6.800639232920496, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17022 + }, + { + "epoch": 6.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17023 + }, + { + "epoch": 6.801438274071114, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17024 + }, + { + "epoch": 6.801837794646424, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17025 + }, + { + "epoch": 6.802237315221734, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17026 + }, + { + "epoch": 6.802636835797044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17027 + }, + { + "epoch": 6.8030363563723535, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17028 + }, + { + "epoch": 6.803435876947663, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17029 + }, + { + "epoch": 6.803835397522972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17030 + }, + { + "epoch": 6.804234918098282, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17031 + }, + { + "epoch": 6.804634438673592, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17032 + }, + { + "epoch": 6.8050339592489015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17033 + }, + { + "epoch": 6.805433479824211, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17034 + }, + { + "epoch": 6.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17035 + }, + { + "epoch": 6.80623252097483, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17036 + }, + { + "epoch": 6.80663204155014, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17037 + }, + { + "epoch": 6.80703156212545, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17038 + }, + { + "epoch": 6.807431082700759, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17039 + }, + { + "epoch": 6.807830603276068, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17040 + }, + { + "epoch": 6.808230123851378, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17041 + }, + { + "epoch": 6.808629644426688, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17042 + }, + { + "epoch": 6.809029165001998, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17043 + }, + { + "epoch": 6.809428685577307, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17044 + }, + { + "epoch": 6.809828206152617, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17045 + }, + { + "epoch": 6.810227726727927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17046 + }, + { + "epoch": 6.810627247303236, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17047 + }, + { + "epoch": 6.811026767878546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17048 + }, + { + "epoch": 6.8114262884538554, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17049 + }, + { + "epoch": 6.811825809029165, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17050 + }, + { + "epoch": 6.812225329604475, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17051 + }, + { + "epoch": 6.812624850179784, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17052 + }, + { + "epoch": 6.813024370755094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17053 + }, + { + "epoch": 6.8134238913304035, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17054 + }, + { + "epoch": 6.813823411905713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 17055 + }, + { + "epoch": 6.814222932481023, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17056 + }, + { + "epoch": 6.814622453056332, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 17057 + }, + { + "epoch": 6.815021973631642, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17058 + }, + { + "epoch": 6.8154214942069515, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17059 + }, + { + "epoch": 6.815821014782261, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17060 + }, + { + "epoch": 6.816220535357571, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 17061 + }, + { + "epoch": 6.816620055932881, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 17062 + }, + { + "epoch": 6.81701957650819, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17063 + }, + { + "epoch": 6.8174190970835, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17064 + }, + { + "epoch": 6.817818617658809, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17065 + }, + { + "epoch": 6.818218138234119, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17066 + }, + { + "epoch": 6.818617658809429, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17067 + }, + { + "epoch": 6.819017179384739, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 17068 + }, + { + "epoch": 6.819416699960048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17069 + }, + { + "epoch": 6.819816220535357, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17070 + }, + { + "epoch": 6.820215741110667, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17071 + }, + { + "epoch": 6.820615261685977, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17072 + }, + { + "epoch": 6.821014782261287, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17073 + }, + { + "epoch": 6.821414302836596, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17074 + }, + { + "epoch": 6.8218138234119055, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 17075 + }, + { + "epoch": 6.822213343987215, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17076 + }, + { + "epoch": 6.822612864562525, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 17077 + }, + { + "epoch": 6.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17078 + }, + { + "epoch": 6.823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17079 + }, + { + "epoch": 6.8238114262884535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17080 + }, + { + "epoch": 6.824210946863763, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17081 + }, + { + "epoch": 6.824610467439073, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17082 + }, + { + "epoch": 6.825009988014383, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17083 + }, + { + "epoch": 6.825409508589693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17084 + }, + { + "epoch": 6.825809029165002, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17085 + }, + { + "epoch": 6.826208549740311, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17086 + }, + { + "epoch": 6.826608070315621, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17087 + }, + { + "epoch": 6.827007590890931, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17088 + }, + { + "epoch": 6.827407111466241, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17089 + }, + { + "epoch": 6.8278066320415505, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17090 + }, + { + "epoch": 6.828206152616859, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17091 + }, + { + "epoch": 6.828605673192169, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 17092 + }, + { + "epoch": 6.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17093 + }, + { + "epoch": 6.829404714342789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17094 + }, + { + "epoch": 6.8298042349180985, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17095 + }, + { + "epoch": 6.830203755493407, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17096 + }, + { + "epoch": 6.830603276068717, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17097 + }, + { + "epoch": 6.831002796644027, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17098 + }, + { + "epoch": 6.831402317219337, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17099 + }, + { + "epoch": 6.831801837794647, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17100 + }, + { + "epoch": 6.832201358369956, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17101 + }, + { + "epoch": 6.832600878945266, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17102 + }, + { + "epoch": 6.833000399520575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17103 + }, + { + "epoch": 6.833399920095885, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17104 + }, + { + "epoch": 6.833799440671195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17105 + }, + { + "epoch": 6.834198961246504, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17106 + }, + { + "epoch": 6.834598481821814, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17107 + }, + { + "epoch": 6.834998002397123, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17108 + }, + { + "epoch": 6.835397522972433, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17109 + }, + { + "epoch": 6.835797043547743, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17110 + }, + { + "epoch": 6.836196564123052, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17111 + }, + { + "epoch": 6.836596084698362, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17112 + }, + { + "epoch": 6.836995605273671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17113 + }, + { + "epoch": 6.837395125848981, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17114 + }, + { + "epoch": 6.837794646424291, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17115 + }, + { + "epoch": 6.8381941669996005, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17116 + }, + { + "epoch": 6.83859368757491, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17117 + }, + { + "epoch": 6.83899320815022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17118 + }, + { + "epoch": 6.83939272872553, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17119 + }, + { + "epoch": 6.839792249300839, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17120 + }, + { + "epoch": 6.8401917698761485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17121 + }, + { + "epoch": 6.840591290451458, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17122 + }, + { + "epoch": 6.840990811026768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17123 + }, + { + "epoch": 6.841390331602078, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17124 + }, + { + "epoch": 6.841789852177387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17125 + }, + { + "epoch": 6.842189372752697, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17126 + }, + { + "epoch": 6.842588893328006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17127 + }, + { + "epoch": 6.842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17128 + }, + { + "epoch": 6.843387934478626, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17129 + }, + { + "epoch": 6.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 17130 + }, + { + "epoch": 6.844186975629245, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17131 + }, + { + "epoch": 6.844586496204554, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17132 + }, + { + "epoch": 6.844986016779864, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17133 + }, + { + "epoch": 6.845385537355174, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17134 + }, + { + "epoch": 6.845785057930484, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17135 + }, + { + "epoch": 6.8461845785057935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17136 + }, + { + "epoch": 6.846584099081102, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17137 + }, + { + "epoch": 6.846983619656412, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17138 + }, + { + "epoch": 6.847383140231722, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17139 + }, + { + "epoch": 6.847782660807032, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17140 + }, + { + "epoch": 6.848182181382342, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17141 + }, + { + "epoch": 6.8485817019576505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17142 + }, + { + "epoch": 6.84898122253296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17143 + }, + { + "epoch": 6.84938074310827, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17144 + }, + { + "epoch": 6.84978026368358, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17145 + }, + { + "epoch": 6.85017978425889, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17146 + }, + { + "epoch": 6.8505793048341985, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17147 + }, + { + "epoch": 6.850978825409508, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17148 + }, + { + "epoch": 6.851378345984818, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17149 + }, + { + "epoch": 6.851777866560128, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17150 + }, + { + "epoch": 6.852177387135438, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 17151 + }, + { + "epoch": 6.8525769077107475, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.2944, + "step": 17152 + }, + { + "epoch": 6.852976428286056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17153 + }, + { + "epoch": 6.853375948861366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17154 + }, + { + "epoch": 6.853775469436676, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17155 + }, + { + "epoch": 6.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17156 + }, + { + "epoch": 6.8545745105872955, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17157 + }, + { + "epoch": 6.854974031162605, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17158 + }, + { + "epoch": 6.855373551737914, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17159 + }, + { + "epoch": 6.855773072313224, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17160 + }, + { + "epoch": 6.856172592888534, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17161 + }, + { + "epoch": 6.8565721134638435, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 17162 + }, + { + "epoch": 6.856971634039153, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17163 + }, + { + "epoch": 6.857371154614462, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17164 + }, + { + "epoch": 6.857770675189772, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17165 + }, + { + "epoch": 6.858170195765082, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17166 + }, + { + "epoch": 6.858569716340392, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17167 + }, + { + "epoch": 6.858969236915701, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17168 + }, + { + "epoch": 6.859368757491011, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17169 + }, + { + "epoch": 6.85976827806632, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17170 + }, + { + "epoch": 6.86016779864163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17171 + }, + { + "epoch": 6.86056731921694, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17172 + }, + { + "epoch": 6.860966839792249, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17173 + }, + { + "epoch": 6.861366360367559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17174 + }, + { + "epoch": 6.861765880942869, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17175 + }, + { + "epoch": 6.862165401518178, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17176 + }, + { + "epoch": 6.862564922093488, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17177 + }, + { + "epoch": 6.8629644426687975, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17178 + }, + { + "epoch": 6.863363963244107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17179 + }, + { + "epoch": 6.863763483819417, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17180 + }, + { + "epoch": 6.864163004394726, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17181 + }, + { + "epoch": 6.864562524970036, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17182 + }, + { + "epoch": 6.8649620455453455, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17183 + }, + { + "epoch": 6.865361566120655, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17184 + }, + { + "epoch": 6.865761086695965, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17185 + }, + { + "epoch": 6.866160607271274, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17186 + }, + { + "epoch": 6.866560127846584, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17187 + }, + { + "epoch": 6.8669596484218935, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17188 + }, + { + "epoch": 6.867359168997203, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17189 + }, + { + "epoch": 6.867758689572513, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17190 + }, + { + "epoch": 6.868158210147823, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17191 + }, + { + "epoch": 6.868557730723133, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17192 + }, + { + "epoch": 6.868957251298442, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17193 + }, + { + "epoch": 6.869356771873751, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17194 + }, + { + "epoch": 6.869756292449061, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17195 + }, + { + "epoch": 6.870155813024371, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17196 + }, + { + "epoch": 6.870555333599681, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17197 + }, + { + "epoch": 6.87095485417499, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17198 + }, + { + "epoch": 6.871354374750299, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17199 + }, + { + "epoch": 6.871753895325609, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17200 + }, + { + "epoch": 6.872153415900919, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17201 + }, + { + "epoch": 6.872552936476229, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 17202 + }, + { + "epoch": 6.872952457051538, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17203 + }, + { + "epoch": 6.8733519776268475, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17204 + }, + { + "epoch": 6.873751498202157, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17205 + }, + { + "epoch": 6.874151018777467, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17206 + }, + { + "epoch": 6.874550539352777, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 17207 + }, + { + "epoch": 6.874950059928087, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17208 + }, + { + "epoch": 6.875349580503396, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17209 + }, + { + "epoch": 6.875749101078705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17210 + }, + { + "epoch": 6.876148621654015, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17211 + }, + { + "epoch": 6.876548142229325, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17212 + }, + { + "epoch": 6.876947662804635, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17213 + }, + { + "epoch": 6.877347183379944, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17214 + }, + { + "epoch": 6.877746703955253, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17215 + }, + { + "epoch": 6.878146224530563, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17216 + }, + { + "epoch": 6.878545745105873, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17217 + }, + { + "epoch": 6.878945265681183, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17218 + }, + { + "epoch": 6.8793447862564925, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17219 + }, + { + "epoch": 6.879744306831801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17220 + }, + { + "epoch": 6.880143827407111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17221 + }, + { + "epoch": 6.880543347982421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17222 + }, + { + "epoch": 6.880942868557731, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17223 + }, + { + "epoch": 6.8813423891330405, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17224 + }, + { + "epoch": 6.88174190970835, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17225 + }, + { + "epoch": 6.88214143028366, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17226 + }, + { + "epoch": 6.882540950858969, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17227 + }, + { + "epoch": 6.882940471434279, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17228 + }, + { + "epoch": 6.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17229 + }, + { + "epoch": 6.883739512584898, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17230 + }, + { + "epoch": 6.884139033160208, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17231 + }, + { + "epoch": 6.884538553735517, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17232 + }, + { + "epoch": 6.884938074310827, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17233 + }, + { + "epoch": 6.885337594886137, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17234 + }, + { + "epoch": 6.885737115461446, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17235 + }, + { + "epoch": 6.886136636036756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17236 + }, + { + "epoch": 6.886536156612065, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17237 + }, + { + "epoch": 6.886935677187375, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17238 + }, + { + "epoch": 6.887335197762685, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17239 + }, + { + "epoch": 6.887734718337994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17240 + }, + { + "epoch": 6.888134238913304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17241 + }, + { + "epoch": 6.888533759488614, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17242 + }, + { + "epoch": 6.888933280063923, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17243 + }, + { + "epoch": 6.889332800639233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17244 + }, + { + "epoch": 6.8897323212145425, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17245 + }, + { + "epoch": 6.890131841789852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17246 + }, + { + "epoch": 6.890531362365162, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17247 + }, + { + "epoch": 6.890930882940472, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17248 + }, + { + "epoch": 6.891330403515781, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17249 + }, + { + "epoch": 6.8917299240910905, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17250 + }, + { + "epoch": 6.8921294446664, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17251 + }, + { + "epoch": 6.89252896524171, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 17252 + }, + { + "epoch": 6.89292848581702, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17253 + }, + { + "epoch": 6.893328006392329, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17254 + }, + { + "epoch": 6.893727526967639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17255 + }, + { + "epoch": 6.894127047542948, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17256 + }, + { + "epoch": 6.894526568118258, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17257 + }, + { + "epoch": 6.894926088693568, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17258 + }, + { + "epoch": 6.895325609268878, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17259 + }, + { + "epoch": 6.895725129844187, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17260 + }, + { + "epoch": 6.896124650419496, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17261 + }, + { + "epoch": 6.896524170994806, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17262 + }, + { + "epoch": 6.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17263 + }, + { + "epoch": 6.897323212145426, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17264 + }, + { + "epoch": 6.8977227327207356, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17265 + }, + { + "epoch": 6.8981222532960444, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17266 + }, + { + "epoch": 6.898521773871354, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17267 + }, + { + "epoch": 6.898921294446664, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17268 + }, + { + "epoch": 6.899320815021974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17269 + }, + { + "epoch": 6.899720335597284, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17270 + }, + { + "epoch": 6.9001198561725925, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17271 + }, + { + "epoch": 6.900519376747902, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17272 + }, + { + "epoch": 6.900918897323212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17273 + }, + { + "epoch": 6.901318417898522, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17274 + }, + { + "epoch": 6.901717938473832, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17275 + }, + { + "epoch": 6.9021174590491405, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.2994, + "step": 17276 + }, + { + "epoch": 6.90251697962445, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17277 + }, + { + "epoch": 6.90291650019976, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17278 + }, + { + "epoch": 6.90331602077507, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17279 + }, + { + "epoch": 6.90371554135038, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17280 + }, + { + "epoch": 6.9041150619256895, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17281 + }, + { + "epoch": 6.904514582500999, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17282 + }, + { + "epoch": 6.904914103076308, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17283 + }, + { + "epoch": 6.905313623651618, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17284 + }, + { + "epoch": 6.905713144226928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17285 + }, + { + "epoch": 6.9061126648022375, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17286 + }, + { + "epoch": 6.906512185377547, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17287 + }, + { + "epoch": 6.906911705952856, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17288 + }, + { + "epoch": 6.907311226528166, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17289 + }, + { + "epoch": 6.907710747103476, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17290 + }, + { + "epoch": 6.9081102676787856, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17291 + }, + { + "epoch": 6.908509788254095, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17292 + }, + { + "epoch": 6.908909308829404, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17293 + }, + { + "epoch": 6.909308829404714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17294 + }, + { + "epoch": 6.909708349980024, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17295 + }, + { + "epoch": 6.910107870555334, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17296 + }, + { + "epoch": 6.910507391130643, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17297 + }, + { + "epoch": 6.910906911705953, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17298 + }, + { + "epoch": 6.911306432281263, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17299 + }, + { + "epoch": 6.911705952856572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17300 + }, + { + "epoch": 6.912105473431882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17301 + }, + { + "epoch": 6.912504994007191, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17302 + }, + { + "epoch": 6.912904514582501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17303 + }, + { + "epoch": 6.913304035157811, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17304 + }, + { + "epoch": 6.91370355573312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17305 + }, + { + "epoch": 6.91410307630843, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17306 + }, + { + "epoch": 6.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17307 + }, + { + "epoch": 6.914902117459049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 17308 + }, + { + "epoch": 6.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17309 + }, + { + "epoch": 6.915701158609668, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17310 + }, + { + "epoch": 6.916100679184978, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17311 + }, + { + "epoch": 6.9165001997602875, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17312 + }, + { + "epoch": 6.916899720335597, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17313 + }, + { + "epoch": 6.917299240910907, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17314 + }, + { + "epoch": 6.917698761486217, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 17315 + }, + { + "epoch": 6.918098282061527, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17316 + }, + { + "epoch": 6.918497802636836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17317 + }, + { + "epoch": 6.918897323212145, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17318 + }, + { + "epoch": 6.919296843787455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17319 + }, + { + "epoch": 6.919696364362765, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17320 + }, + { + "epoch": 6.920095884938075, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17321 + }, + { + "epoch": 6.920495405513384, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17322 + }, + { + "epoch": 6.920894926088693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17323 + }, + { + "epoch": 6.921294446664003, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17324 + }, + { + "epoch": 6.921693967239313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 17325 + }, + { + "epoch": 6.922093487814623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17326 + }, + { + "epoch": 6.922493008389932, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17327 + }, + { + "epoch": 6.922892528965241, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17328 + }, + { + "epoch": 6.923292049540551, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17329 + }, + { + "epoch": 6.923691570115861, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 17330 + }, + { + "epoch": 6.924091090691171, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17331 + }, + { + "epoch": 6.924490611266481, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17332 + }, + { + "epoch": 6.9248901318417895, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17333 + }, + { + "epoch": 6.925289652417099, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17334 + }, + { + "epoch": 6.925689172992409, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17335 + }, + { + "epoch": 6.926088693567719, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17336 + }, + { + "epoch": 6.926488214143029, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17337 + }, + { + "epoch": 6.926887734718338, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17338 + }, + { + "epoch": 6.927287255293647, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17339 + }, + { + "epoch": 6.927686775868957, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 17340 + }, + { + "epoch": 6.928086296444267, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17341 + }, + { + "epoch": 6.928485817019577, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17342 + }, + { + "epoch": 6.9288853375948865, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17343 + }, + { + "epoch": 6.929284858170195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17344 + }, + { + "epoch": 6.929684378745505, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17345 + }, + { + "epoch": 6.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 17346 + }, + { + "epoch": 6.930483419896125, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17347 + }, + { + "epoch": 6.9308829404714345, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 17348 + }, + { + "epoch": 6.931282461046744, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17349 + }, + { + "epoch": 6.931681981622053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17350 + }, + { + "epoch": 6.932081502197363, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 17351 + }, + { + "epoch": 6.932481022772673, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17352 + }, + { + "epoch": 6.9328805433479825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17353 + }, + { + "epoch": 6.933280063923292, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17354 + }, + { + "epoch": 6.933679584498602, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17355 + }, + { + "epoch": 6.934079105073911, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17356 + }, + { + "epoch": 6.934478625649221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17357 + }, + { + "epoch": 6.934878146224531, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17358 + }, + { + "epoch": 6.93527766679984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17359 + }, + { + "epoch": 6.93567718737515, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17360 + }, + { + "epoch": 6.936076707950459, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17361 + }, + { + "epoch": 6.936476228525769, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 17362 + }, + { + "epoch": 6.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17363 + }, + { + "epoch": 6.937275269676388, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17364 + }, + { + "epoch": 6.937674790251698, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17365 + }, + { + "epoch": 6.938074310827007, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17366 + }, + { + "epoch": 6.938473831402317, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17367 + }, + { + "epoch": 6.938873351977627, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 17368 + }, + { + "epoch": 6.9392728725529365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17369 + }, + { + "epoch": 6.939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17370 + }, + { + "epoch": 6.940071913703556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17371 + }, + { + "epoch": 6.940471434278866, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17372 + }, + { + "epoch": 6.940870954854175, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17373 + }, + { + "epoch": 6.9412704754294845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17374 + }, + { + "epoch": 6.941669996004794, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17375 + }, + { + "epoch": 6.942069516580104, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17376 + }, + { + "epoch": 6.942469037155414, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17377 + }, + { + "epoch": 6.942868557730723, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17378 + }, + { + "epoch": 6.9432680783060325, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17379 + }, + { + "epoch": 6.943667598881342, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 17380 + }, + { + "epoch": 6.944067119456652, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17381 + }, + { + "epoch": 6.944466640031962, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17382 + }, + { + "epoch": 6.944866160607271, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17383 + }, + { + "epoch": 6.945265681182581, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 17384 + }, + { + "epoch": 6.94566520175789, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17385 + }, + { + "epoch": 6.9460647223332, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17386 + }, + { + "epoch": 6.94646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17387 + }, + { + "epoch": 6.94686376348382, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 17388 + }, + { + "epoch": 6.9472632840591295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17389 + }, + { + "epoch": 6.947662804634438, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17390 + }, + { + "epoch": 6.948062325209748, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17391 + }, + { + "epoch": 6.948461845785058, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17392 + }, + { + "epoch": 6.948861366360368, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17393 + }, + { + "epoch": 6.949260886935678, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17394 + }, + { + "epoch": 6.9496604075109865, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17395 + }, + { + "epoch": 6.950059928086296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 17396 + }, + { + "epoch": 6.950459448661606, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 17397 + }, + { + "epoch": 6.950858969236916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17398 + }, + { + "epoch": 6.951258489812226, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17399 + }, + { + "epoch": 6.9516580103875345, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17400 + }, + { + "epoch": 6.952057530962844, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17401 + }, + { + "epoch": 6.952457051538154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17402 + }, + { + "epoch": 6.952856572113464, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17403 + }, + { + "epoch": 6.953256092688774, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17404 + }, + { + "epoch": 6.953655613264083, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17405 + }, + { + "epoch": 6.954055133839393, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17406 + }, + { + "epoch": 6.954454654414702, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17407 + }, + { + "epoch": 6.954854174990012, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17408 + }, + { + "epoch": 6.955253695565322, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17409 + }, + { + "epoch": 6.9556532161406315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17410 + }, + { + "epoch": 6.956052736715941, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 17411 + }, + { + "epoch": 6.95645225729125, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17412 + }, + { + "epoch": 6.95685177786656, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17413 + }, + { + "epoch": 6.95725129844187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17414 + }, + { + "epoch": 6.9576508190171795, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17415 + }, + { + "epoch": 6.958050339592489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17416 + }, + { + "epoch": 6.958449860167798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17417 + }, + { + "epoch": 6.958849380743108, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17418 + }, + { + "epoch": 6.959248901318418, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17419 + }, + { + "epoch": 6.959648421893728, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17420 + }, + { + "epoch": 6.960047942469037, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17421 + }, + { + "epoch": 6.960447463044347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17422 + }, + { + "epoch": 6.960846983619656, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17423 + }, + { + "epoch": 6.961246504194966, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17424 + }, + { + "epoch": 6.961646024770276, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17425 + }, + { + "epoch": 6.962045545345585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17426 + }, + { + "epoch": 6.962445065920895, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17427 + }, + { + "epoch": 6.962844586496205, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17428 + }, + { + "epoch": 6.963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17429 + }, + { + "epoch": 6.963643627646824, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17430 + }, + { + "epoch": 6.964043148222133, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17431 + }, + { + "epoch": 6.964442668797443, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17432 + }, + { + "epoch": 6.964842189372753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17433 + }, + { + "epoch": 6.965241709948062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17434 + }, + { + "epoch": 6.965641230523372, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17435 + }, + { + "epoch": 6.9660407510986815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17436 + }, + { + "epoch": 6.966440271673991, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17437 + }, + { + "epoch": 6.966839792249301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17438 + }, + { + "epoch": 6.967239312824611, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 17439 + }, + { + "epoch": 6.96763883339992, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17440 + }, + { + "epoch": 6.9680383539752295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17441 + }, + { + "epoch": 6.968437874550539, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17442 + }, + { + "epoch": 6.968837395125849, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17443 + }, + { + "epoch": 6.969236915701159, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17444 + }, + { + "epoch": 6.969636436276469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17445 + }, + { + "epoch": 6.970035956851778, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17446 + }, + { + "epoch": 6.970435477427087, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17447 + }, + { + "epoch": 6.970834998002397, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17448 + }, + { + "epoch": 6.971234518577707, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17449 + }, + { + "epoch": 6.971634039153017, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17450 + }, + { + "epoch": 6.972033559728326, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17451 + }, + { + "epoch": 6.972433080303635, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17452 + }, + { + "epoch": 6.972832600878945, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17453 + }, + { + "epoch": 6.973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17454 + }, + { + "epoch": 6.973631642029565, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17455 + }, + { + "epoch": 6.974031162604874, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17456 + }, + { + "epoch": 6.974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17457 + }, + { + "epoch": 6.974830203755493, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17458 + }, + { + "epoch": 6.975229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17459 + }, + { + "epoch": 6.975629244906113, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17460 + }, + { + "epoch": 6.976028765481423, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17461 + }, + { + "epoch": 6.976428286056732, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 17462 + }, + { + "epoch": 6.976827806632041, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17463 + }, + { + "epoch": 6.977227327207351, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17464 + }, + { + "epoch": 6.977626847782661, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17465 + }, + { + "epoch": 6.978026368357971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17466 + }, + { + "epoch": 6.97842588893328, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17467 + }, + { + "epoch": 6.978825409508589, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17468 + }, + { + "epoch": 6.979224930083899, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17469 + }, + { + "epoch": 6.979624450659209, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17470 + }, + { + "epoch": 6.980023971234519, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17471 + }, + { + "epoch": 6.9804234918098285, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17472 + }, + { + "epoch": 6.980823012385137, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17473 + }, + { + "epoch": 6.981222532960447, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17474 + }, + { + "epoch": 6.981622053535757, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17475 + }, + { + "epoch": 6.982021574111067, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17476 + }, + { + "epoch": 6.9824210946863765, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17477 + }, + { + "epoch": 6.982820615261686, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17478 + }, + { + "epoch": 6.983220135836996, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17479 + }, + { + "epoch": 6.983619656412305, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17480 + }, + { + "epoch": 6.984019176987615, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17481 + }, + { + "epoch": 6.9844186975629245, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17482 + }, + { + "epoch": 6.984818218138234, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17483 + }, + { + "epoch": 6.985217738713544, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17484 + }, + { + "epoch": 6.985617259288853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17485 + }, + { + "epoch": 6.986016779864163, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17486 + }, + { + "epoch": 6.986416300439473, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17487 + }, + { + "epoch": 6.986815821014782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17488 + }, + { + "epoch": 6.987215341590092, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17489 + }, + { + "epoch": 6.987614862165401, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17490 + }, + { + "epoch": 6.988014382740711, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17491 + }, + { + "epoch": 6.988413903316021, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17492 + }, + { + "epoch": 6.98881342389133, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17493 + }, + { + "epoch": 6.98921294446664, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17494 + }, + { + "epoch": 6.98961246504195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17495 + }, + { + "epoch": 6.99001198561726, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17496 + }, + { + "epoch": 6.990411506192569, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17497 + }, + { + "epoch": 6.9908110267678785, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17498 + }, + { + "epoch": 6.991210547343188, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17499 + }, + { + "epoch": 6.991610067918498, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17500 + }, + { + "epoch": 6.992009588493808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17501 + }, + { + "epoch": 6.992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 17502 + }, + { + "epoch": 6.9928086296444265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17503 + }, + { + "epoch": 6.993208150219736, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17504 + }, + { + "epoch": 6.993607670795046, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 17505 + }, + { + "epoch": 6.994007191370356, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17506 + }, + { + "epoch": 6.994406711945665, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17507 + }, + { + "epoch": 6.9948062325209746, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17508 + }, + { + "epoch": 6.995205753096284, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17509 + }, + { + "epoch": 6.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17510 + }, + { + "epoch": 6.996004794246904, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17511 + }, + { + "epoch": 6.996404314822214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17512 + }, + { + "epoch": 6.996803835397523, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 17513 + }, + { + "epoch": 6.997203355972832, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17514 + }, + { + "epoch": 6.997602876548142, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17515 + }, + { + "epoch": 6.998002397123452, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17516 + }, + { + "epoch": 6.998401917698762, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 17517 + }, + { + "epoch": 6.9988014382740715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17518 + }, + { + "epoch": 6.99920095884938, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 17519 + }, + { + "epoch": 6.99960047942469, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17520 + }, + { + "epoch": 7.0, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17521 + }, + { + "epoch": 7.00039952057531, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17522 + }, + { + "epoch": 7.00079904115062, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17523 + }, + { + "epoch": 7.0011985617259285, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17524 + }, + { + "epoch": 7.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17525 + }, + { + "epoch": 7.001997602876548, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17526 + }, + { + "epoch": 7.002397123451858, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17527 + }, + { + "epoch": 7.002796644027168, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17528 + }, + { + "epoch": 7.003196164602477, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17529 + }, + { + "epoch": 7.003595685177786, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17530 + }, + { + "epoch": 7.003995205753096, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17531 + }, + { + "epoch": 7.004394726328406, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17532 + }, + { + "epoch": 7.004794246903716, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17533 + }, + { + "epoch": 7.0051937674790254, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17534 + }, + { + "epoch": 7.005593288054335, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17535 + }, + { + "epoch": 7.005992808629644, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17536 + }, + { + "epoch": 7.006392329204954, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17537 + }, + { + "epoch": 7.006791849780264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17538 + }, + { + "epoch": 7.0071913703555735, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17539 + }, + { + "epoch": 7.007590890930883, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17540 + }, + { + "epoch": 7.007990411506192, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17541 + }, + { + "epoch": 7.008389932081502, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17542 + }, + { + "epoch": 7.008789452656812, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17543 + }, + { + "epoch": 7.0091889732321215, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17544 + }, + { + "epoch": 7.009588493807431, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17545 + }, + { + "epoch": 7.009988014382741, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17546 + }, + { + "epoch": 7.01038753495805, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17547 + }, + { + "epoch": 7.01078705553336, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17548 + }, + { + "epoch": 7.01118657610867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 17549 + }, + { + "epoch": 7.011586096683979, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17550 + }, + { + "epoch": 7.011985617259289, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17551 + }, + { + "epoch": 7.012385137834598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17552 + }, + { + "epoch": 7.012784658409908, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17553 + }, + { + "epoch": 7.013184178985218, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17554 + }, + { + "epoch": 7.013583699560527, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17555 + }, + { + "epoch": 7.013983220135837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17556 + }, + { + "epoch": 7.014382740711147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17557 + }, + { + "epoch": 7.014782261286456, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17558 + }, + { + "epoch": 7.015181781861766, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17559 + }, + { + "epoch": 7.0155813024370755, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17560 + }, + { + "epoch": 7.015980823012385, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17561 + }, + { + "epoch": 7.016380343587695, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17562 + }, + { + "epoch": 7.016779864163005, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17563 + }, + { + "epoch": 7.017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17564 + }, + { + "epoch": 7.0175789053136235, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 17565 + }, + { + "epoch": 7.017978425888933, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17566 + }, + { + "epoch": 7.018377946464243, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 17567 + }, + { + "epoch": 7.018777467039553, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17568 + }, + { + "epoch": 7.019176987614862, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17569 + }, + { + "epoch": 7.0195765081901715, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17570 + }, + { + "epoch": 7.019976028765481, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17571 + }, + { + "epoch": 7.020375549340791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17572 + }, + { + "epoch": 7.020775069916101, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17573 + }, + { + "epoch": 7.021174590491411, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17574 + }, + { + "epoch": 7.02157411106672, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 17575 + }, + { + "epoch": 7.021973631642029, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17576 + }, + { + "epoch": 7.022373152217339, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17577 + }, + { + "epoch": 7.022772672792649, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17578 + }, + { + "epoch": 7.023172193367959, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17579 + }, + { + "epoch": 7.0235717139432685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 17580 + }, + { + "epoch": 7.023971234518577, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17581 + }, + { + "epoch": 7.024370755093887, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17582 + }, + { + "epoch": 7.024770275669197, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17583 + }, + { + "epoch": 7.025169796244507, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17584 + }, + { + "epoch": 7.025569316819817, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17585 + }, + { + "epoch": 7.0259688373951255, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17586 + }, + { + "epoch": 7.026368357970435, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17587 + }, + { + "epoch": 7.026767878545745, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17588 + }, + { + "epoch": 7.027167399121055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17589 + }, + { + "epoch": 7.027566919696365, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 17590 + }, + { + "epoch": 7.027966440271674, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17591 + }, + { + "epoch": 7.028365960846983, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17592 + }, + { + "epoch": 7.028765481422293, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17593 + }, + { + "epoch": 7.029165001997603, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17594 + }, + { + "epoch": 7.029564522572913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17595 + }, + { + "epoch": 7.029964043148222, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17596 + }, + { + "epoch": 7.030363563723531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17597 + }, + { + "epoch": 7.030763084298841, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 17598 + }, + { + "epoch": 7.031162604874151, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17599 + }, + { + "epoch": 7.031562125449461, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17600 + }, + { + "epoch": 7.0319616460247705, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 17601 + }, + { + "epoch": 7.03236116660008, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17602 + }, + { + "epoch": 7.032760687175389, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17603 + }, + { + "epoch": 7.033160207750699, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17604 + }, + { + "epoch": 7.033559728326009, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17605 + }, + { + "epoch": 7.0339592489013185, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17606 + }, + { + "epoch": 7.034358769476628, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17607 + }, + { + "epoch": 7.034758290051938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17608 + }, + { + "epoch": 7.035157810627247, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17609 + }, + { + "epoch": 7.035557331202557, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17610 + }, + { + "epoch": 7.035956851777867, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17611 + }, + { + "epoch": 7.036356372353176, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17612 + }, + { + "epoch": 7.036755892928486, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17613 + }, + { + "epoch": 7.037155413503795, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17614 + }, + { + "epoch": 7.037554934079105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17615 + }, + { + "epoch": 7.037954454654415, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 17616 + }, + { + "epoch": 7.038353975229724, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17617 + }, + { + "epoch": 7.038753495805034, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17618 + }, + { + "epoch": 7.039153016380344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17619 + }, + { + "epoch": 7.039552536955653, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17620 + }, + { + "epoch": 7.039952057530963, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17621 + }, + { + "epoch": 7.040351578106272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 17622 + }, + { + "epoch": 7.040751098681582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17623 + }, + { + "epoch": 7.041150619256892, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17624 + }, + { + "epoch": 7.041550139832202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 17625 + }, + { + "epoch": 7.041949660407511, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17626 + }, + { + "epoch": 7.0423491809828205, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17627 + }, + { + "epoch": 7.04274870155813, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17628 + }, + { + "epoch": 7.04314822213344, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17629 + }, + { + "epoch": 7.04354774270875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17630 + }, + { + "epoch": 7.043947263284059, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17631 + }, + { + "epoch": 7.0443467838593685, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17632 + }, + { + "epoch": 7.044746304434678, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17633 + }, + { + "epoch": 7.045145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17634 + }, + { + "epoch": 7.045545345585298, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17635 + }, + { + "epoch": 7.045944866160608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17636 + }, + { + "epoch": 7.046344386735917, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.2991, + "step": 17637 + }, + { + "epoch": 7.046743907311226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17638 + }, + { + "epoch": 7.047143427886536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17639 + }, + { + "epoch": 7.047542948461846, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17640 + }, + { + "epoch": 7.047942469037156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17641 + }, + { + "epoch": 7.048341989612465, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17642 + }, + { + "epoch": 7.048741510187774, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17643 + }, + { + "epoch": 7.049141030763084, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17644 + }, + { + "epoch": 7.049540551338394, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17645 + }, + { + "epoch": 7.049940071913704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17646 + }, + { + "epoch": 7.0503395924890135, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17647 + }, + { + "epoch": 7.050739113064322, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17648 + }, + { + "epoch": 7.051138633639632, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17649 + }, + { + "epoch": 7.051538154214942, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17650 + }, + { + "epoch": 7.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17651 + }, + { + "epoch": 7.052337195365562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17652 + }, + { + "epoch": 7.052736715940871, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17653 + }, + { + "epoch": 7.05313623651618, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17654 + }, + { + "epoch": 7.05353575709149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17655 + }, + { + "epoch": 7.0539352776668, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17656 + }, + { + "epoch": 7.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17657 + }, + { + "epoch": 7.054734318817419, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17658 + }, + { + "epoch": 7.055133839392728, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17659 + }, + { + "epoch": 7.055533359968038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17660 + }, + { + "epoch": 7.055932880543348, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17661 + }, + { + "epoch": 7.056332401118658, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17662 + }, + { + "epoch": 7.0567319216939675, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17663 + }, + { + "epoch": 7.057131442269277, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17664 + }, + { + "epoch": 7.057530962844586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17665 + }, + { + "epoch": 7.057930483419896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17666 + }, + { + "epoch": 7.058330003995206, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17667 + }, + { + "epoch": 7.0587295245705155, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17668 + }, + { + "epoch": 7.059129045145825, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17669 + }, + { + "epoch": 7.059528565721135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17670 + }, + { + "epoch": 7.059928086296444, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17671 + }, + { + "epoch": 7.060327606871754, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17672 + }, + { + "epoch": 7.0607271274470635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17673 + }, + { + "epoch": 7.061126648022373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17674 + }, + { + "epoch": 7.061526168597683, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17675 + }, + { + "epoch": 7.061925689172992, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17676 + }, + { + "epoch": 7.062325209748302, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17677 + }, + { + "epoch": 7.062724730323612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17678 + }, + { + "epoch": 7.063124250898921, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17679 + }, + { + "epoch": 7.063523771474231, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17680 + }, + { + "epoch": 7.063923292049541, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17681 + }, + { + "epoch": 7.06432281262485, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 17682 + }, + { + "epoch": 7.06472233320016, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17683 + }, + { + "epoch": 7.065121853775469, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17684 + }, + { + "epoch": 7.065521374350779, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17685 + }, + { + "epoch": 7.065920894926089, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17686 + }, + { + "epoch": 7.066320415501398, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17687 + }, + { + "epoch": 7.066719936076708, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17688 + }, + { + "epoch": 7.0671194566520175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 17689 + }, + { + "epoch": 7.067518977227327, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17690 + }, + { + "epoch": 7.067918497802637, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17691 + }, + { + "epoch": 7.068318018377947, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17692 + }, + { + "epoch": 7.068717538953256, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17693 + }, + { + "epoch": 7.0691170595285655, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17694 + }, + { + "epoch": 7.069516580103875, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17695 + }, + { + "epoch": 7.069916100679185, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17696 + }, + { + "epoch": 7.070315621254495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17697 + }, + { + "epoch": 7.070715141829805, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17698 + }, + { + "epoch": 7.0711146624051135, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 17699 + }, + { + "epoch": 7.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17700 + }, + { + "epoch": 7.071913703555733, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17701 + }, + { + "epoch": 7.072313224131043, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 17702 + }, + { + "epoch": 7.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17703 + }, + { + "epoch": 7.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17704 + }, + { + "epoch": 7.073511785856971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17705 + }, + { + "epoch": 7.073911306432281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17706 + }, + { + "epoch": 7.074310827007591, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17707 + }, + { + "epoch": 7.074710347582901, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17708 + }, + { + "epoch": 7.0751098681582105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17709 + }, + { + "epoch": 7.075509388733519, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17710 + }, + { + "epoch": 7.075908909308829, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17711 + }, + { + "epoch": 7.076308429884139, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17712 + }, + { + "epoch": 7.076707950459449, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17713 + }, + { + "epoch": 7.077107471034759, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17714 + }, + { + "epoch": 7.077506991610068, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17715 + }, + { + "epoch": 7.077906512185377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17716 + }, + { + "epoch": 7.078306032760687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17717 + }, + { + "epoch": 7.078705553335997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17718 + }, + { + "epoch": 7.079105073911307, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17719 + }, + { + "epoch": 7.079504594486616, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17720 + }, + { + "epoch": 7.079904115061925, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17721 + }, + { + "epoch": 7.080303635637235, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17722 + }, + { + "epoch": 7.080703156212545, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17723 + }, + { + "epoch": 7.081102676787855, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17724 + }, + { + "epoch": 7.081502197363164, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17725 + }, + { + "epoch": 7.081901717938474, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17726 + }, + { + "epoch": 7.082301238513783, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17727 + }, + { + "epoch": 7.082700759089093, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17728 + }, + { + "epoch": 7.083100279664403, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17729 + }, + { + "epoch": 7.0834998002397125, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17730 + }, + { + "epoch": 7.083899320815022, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 17731 + }, + { + "epoch": 7.084298841390331, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17732 + }, + { + "epoch": 7.084698361965641, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17733 + }, + { + "epoch": 7.085097882540951, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17734 + }, + { + "epoch": 7.0854974031162605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17735 + }, + { + "epoch": 7.08589692369157, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17736 + }, + { + "epoch": 7.08629644426688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17737 + }, + { + "epoch": 7.086695964842189, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17738 + }, + { + "epoch": 7.087095485417499, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17739 + }, + { + "epoch": 7.087495005992809, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17740 + }, + { + "epoch": 7.087894526568118, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17741 + }, + { + "epoch": 7.088294047143428, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17742 + }, + { + "epoch": 7.088693567718738, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17743 + }, + { + "epoch": 7.089093088294047, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17744 + }, + { + "epoch": 7.089492608869357, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17745 + }, + { + "epoch": 7.089892129444666, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17746 + }, + { + "epoch": 7.090291650019976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17747 + }, + { + "epoch": 7.090691170595286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17748 + }, + { + "epoch": 7.091090691170595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 17749 + }, + { + "epoch": 7.091490211745905, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17750 + }, + { + "epoch": 7.0918897323212144, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17751 + }, + { + "epoch": 7.092289252896524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17752 + }, + { + "epoch": 7.092688773471834, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17753 + }, + { + "epoch": 7.093088294047144, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17754 + }, + { + "epoch": 7.093487814622453, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17755 + }, + { + "epoch": 7.0938873351977625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17756 + }, + { + "epoch": 7.094286855773072, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17757 + }, + { + "epoch": 7.094686376348382, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17758 + }, + { + "epoch": 7.095085896923692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17759 + }, + { + "epoch": 7.095485417499002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17760 + }, + { + "epoch": 7.0958849380743105, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17761 + }, + { + "epoch": 7.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17762 + }, + { + "epoch": 7.09668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17763 + }, + { + "epoch": 7.09708349980024, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17764 + }, + { + "epoch": 7.09748302037555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17765 + }, + { + "epoch": 7.097882540950859, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17766 + }, + { + "epoch": 7.098282061526168, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17767 + }, + { + "epoch": 7.098681582101478, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17768 + }, + { + "epoch": 7.099081102676788, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17769 + }, + { + "epoch": 7.099480623252098, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17770 + }, + { + "epoch": 7.0998801438274075, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17771 + }, + { + "epoch": 7.100279664402716, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17772 + }, + { + "epoch": 7.100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17773 + }, + { + "epoch": 7.101078705553336, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17774 + }, + { + "epoch": 7.101478226128646, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17775 + }, + { + "epoch": 7.1018777467039556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17776 + }, + { + "epoch": 7.1022772672792644, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17777 + }, + { + "epoch": 7.102676787854574, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17778 + }, + { + "epoch": 7.103076308429884, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17779 + }, + { + "epoch": 7.103475829005194, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17780 + }, + { + "epoch": 7.103875349580504, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17781 + }, + { + "epoch": 7.104274870155813, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17782 + }, + { + "epoch": 7.104674390731122, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17783 + }, + { + "epoch": 7.105073911306432, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17784 + }, + { + "epoch": 7.105473431881742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 17785 + }, + { + "epoch": 7.105872952457052, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17786 + }, + { + "epoch": 7.106272473032361, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17787 + }, + { + "epoch": 7.106671993607671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17788 + }, + { + "epoch": 7.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17789 + }, + { + "epoch": 7.10747103475829, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17790 + }, + { + "epoch": 7.1078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17791 + }, + { + "epoch": 7.1082700759089095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17792 + }, + { + "epoch": 7.108669596484219, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17793 + }, + { + "epoch": 7.109069117059528, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17794 + }, + { + "epoch": 7.109468637634838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17795 + }, + { + "epoch": 7.109868158210148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17796 + }, + { + "epoch": 7.1102676787854575, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17797 + }, + { + "epoch": 7.110667199360767, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17798 + }, + { + "epoch": 7.111066719936077, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17799 + }, + { + "epoch": 7.111466240511386, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17800 + }, + { + "epoch": 7.111865761086696, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17801 + }, + { + "epoch": 7.112265281662006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 17802 + }, + { + "epoch": 7.112664802237315, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17803 + }, + { + "epoch": 7.113064322812625, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17804 + }, + { + "epoch": 7.113463843387935, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17805 + }, + { + "epoch": 7.113863363963244, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17806 + }, + { + "epoch": 7.114262884538554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17807 + }, + { + "epoch": 7.114662405113863, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17808 + }, + { + "epoch": 7.115061925689173, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17809 + }, + { + "epoch": 7.115461446264483, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17810 + }, + { + "epoch": 7.115860966839792, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17811 + }, + { + "epoch": 7.116260487415102, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 17812 + }, + { + "epoch": 7.116660007990411, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17813 + }, + { + "epoch": 7.117059528565721, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17814 + }, + { + "epoch": 7.117459049141031, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17815 + }, + { + "epoch": 7.117858569716341, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17816 + }, + { + "epoch": 7.11825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17817 + }, + { + "epoch": 7.1186576108669595, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 17818 + }, + { + "epoch": 7.119057131442269, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17819 + }, + { + "epoch": 7.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17820 + }, + { + "epoch": 7.119856172592889, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17821 + }, + { + "epoch": 7.120255693168198, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17822 + }, + { + "epoch": 7.1206552137435075, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17823 + }, + { + "epoch": 7.121054734318817, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17824 + }, + { + "epoch": 7.121454254894127, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17825 + }, + { + "epoch": 7.121853775469437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17826 + }, + { + "epoch": 7.122253296044747, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17827 + }, + { + "epoch": 7.122652816620056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17828 + }, + { + "epoch": 7.123052337195365, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 17829 + }, + { + "epoch": 7.123451857770675, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 17830 + }, + { + "epoch": 7.123851378345985, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17831 + }, + { + "epoch": 7.124250898921295, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17832 + }, + { + "epoch": 7.1246504194966045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17833 + }, + { + "epoch": 7.125049940071913, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17834 + }, + { + "epoch": 7.125449460647223, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17835 + }, + { + "epoch": 7.125848981222533, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17836 + }, + { + "epoch": 7.126248501797843, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17837 + }, + { + "epoch": 7.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17838 + }, + { + "epoch": 7.127047542948461, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17839 + }, + { + "epoch": 7.127447063523771, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17840 + }, + { + "epoch": 7.127846584099081, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17841 + }, + { + "epoch": 7.128246104674391, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17842 + }, + { + "epoch": 7.128645625249701, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17843 + }, + { + "epoch": 7.12904514582501, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17844 + }, + { + "epoch": 7.129444666400319, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 17845 + }, + { + "epoch": 7.129844186975629, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17846 + }, + { + "epoch": 7.130243707550939, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17847 + }, + { + "epoch": 7.130643228126249, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17848 + }, + { + "epoch": 7.131042748701558, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17849 + }, + { + "epoch": 7.131442269276867, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17850 + }, + { + "epoch": 7.131841789852177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17851 + }, + { + "epoch": 7.132241310427487, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17852 + }, + { + "epoch": 7.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17853 + }, + { + "epoch": 7.1330403515781065, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17854 + }, + { + "epoch": 7.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17855 + }, + { + "epoch": 7.133839392728725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17856 + }, + { + "epoch": 7.134238913304035, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17857 + }, + { + "epoch": 7.134638433879345, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17858 + }, + { + "epoch": 7.1350379544546545, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17859 + }, + { + "epoch": 7.135437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17860 + }, + { + "epoch": 7.135836995605274, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17861 + }, + { + "epoch": 7.136236516180583, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17862 + }, + { + "epoch": 7.136636036755893, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17863 + }, + { + "epoch": 7.1370355573312025, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17864 + }, + { + "epoch": 7.137435077906512, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17865 + }, + { + "epoch": 7.137834598481822, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17866 + }, + { + "epoch": 7.138234119057131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 17867 + }, + { + "epoch": 7.138633639632441, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17868 + }, + { + "epoch": 7.139033160207751, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17869 + }, + { + "epoch": 7.13943268078306, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17870 + }, + { + "epoch": 7.13983220135837, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17871 + }, + { + "epoch": 7.14023172193368, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17872 + }, + { + "epoch": 7.140631242508989, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17873 + }, + { + "epoch": 7.141030763084299, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17874 + }, + { + "epoch": 7.141430283659608, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17875 + }, + { + "epoch": 7.141829804234918, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17876 + }, + { + "epoch": 7.142229324810228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17877 + }, + { + "epoch": 7.142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17878 + }, + { + "epoch": 7.143028365960847, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17879 + }, + { + "epoch": 7.1434278865361565, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17880 + }, + { + "epoch": 7.143827407111466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17881 + }, + { + "epoch": 7.144226927686776, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17882 + }, + { + "epoch": 7.144626448262086, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17883 + }, + { + "epoch": 7.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17884 + }, + { + "epoch": 7.1454254894127045, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17885 + }, + { + "epoch": 7.145825009988014, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 17886 + }, + { + "epoch": 7.146224530563324, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17887 + }, + { + "epoch": 7.146624051138634, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17888 + }, + { + "epoch": 7.147023571713944, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17889 + }, + { + "epoch": 7.1474230922892525, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17890 + }, + { + "epoch": 7.147822612864562, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17891 + }, + { + "epoch": 7.148222133439872, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17892 + }, + { + "epoch": 7.148621654015182, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 17893 + }, + { + "epoch": 7.149021174590492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17894 + }, + { + "epoch": 7.1494206951658015, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17895 + }, + { + "epoch": 7.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17896 + }, + { + "epoch": 7.15021973631642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17897 + }, + { + "epoch": 7.15061925689173, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17898 + }, + { + "epoch": 7.15101877746704, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17899 + }, + { + "epoch": 7.1514182980423495, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17900 + }, + { + "epoch": 7.151817818617658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17901 + }, + { + "epoch": 7.152217339192968, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17902 + }, + { + "epoch": 7.152616859768278, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17903 + }, + { + "epoch": 7.153016380343588, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17904 + }, + { + "epoch": 7.153415900918898, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17905 + }, + { + "epoch": 7.153815421494207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17906 + }, + { + "epoch": 7.154214942069516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 17907 + }, + { + "epoch": 7.154614462644826, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17908 + }, + { + "epoch": 7.155013983220136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17909 + }, + { + "epoch": 7.155413503795446, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17910 + }, + { + "epoch": 7.155813024370755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17911 + }, + { + "epoch": 7.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17912 + }, + { + "epoch": 7.156612065521374, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17913 + }, + { + "epoch": 7.157011586096684, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17914 + }, + { + "epoch": 7.157411106671994, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17915 + }, + { + "epoch": 7.157810627247303, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17916 + }, + { + "epoch": 7.158210147822613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17917 + }, + { + "epoch": 7.158609668397922, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17918 + }, + { + "epoch": 7.159009188973232, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17919 + }, + { + "epoch": 7.159408709548542, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17920 + }, + { + "epoch": 7.1598082301238515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17921 + }, + { + "epoch": 7.160207750699161, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17922 + }, + { + "epoch": 7.160607271274471, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17923 + }, + { + "epoch": 7.16100679184978, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17924 + }, + { + "epoch": 7.16140631242509, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17925 + }, + { + "epoch": 7.1618058330003995, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17926 + }, + { + "epoch": 7.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17927 + }, + { + "epoch": 7.162604874151019, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17928 + }, + { + "epoch": 7.163004394726328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17929 + }, + { + "epoch": 7.163403915301638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17930 + }, + { + "epoch": 7.163803435876948, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17931 + }, + { + "epoch": 7.164202956452257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17932 + }, + { + "epoch": 7.164602477027567, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17933 + }, + { + "epoch": 7.165001997602877, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17934 + }, + { + "epoch": 7.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17935 + }, + { + "epoch": 7.165801038753496, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17936 + }, + { + "epoch": 7.166200559328805, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17937 + }, + { + "epoch": 7.166600079904115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17938 + }, + { + "epoch": 7.166999600479425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17939 + }, + { + "epoch": 7.167399121054734, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17940 + }, + { + "epoch": 7.167798641630044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17941 + }, + { + "epoch": 7.168198162205353, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17942 + }, + { + "epoch": 7.168597682780663, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17943 + }, + { + "epoch": 7.168997203355973, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17944 + }, + { + "epoch": 7.169396723931283, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 17945 + }, + { + "epoch": 7.169796244506592, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17946 + }, + { + "epoch": 7.1701957650819015, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17947 + }, + { + "epoch": 7.170595285657211, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17948 + }, + { + "epoch": 7.170994806232521, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17949 + }, + { + "epoch": 7.171394326807831, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17950 + }, + { + "epoch": 7.171793847383141, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17951 + }, + { + "epoch": 7.1721933679584495, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17952 + }, + { + "epoch": 7.172592888533759, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17953 + }, + { + "epoch": 7.172992409109069, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17954 + }, + { + "epoch": 7.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17955 + }, + { + "epoch": 7.173791450259689, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17956 + }, + { + "epoch": 7.174190970834998, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17957 + }, + { + "epoch": 7.174590491410307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17958 + }, + { + "epoch": 7.174990011985617, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17959 + }, + { + "epoch": 7.175389532560927, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17960 + }, + { + "epoch": 7.175789053136237, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17961 + }, + { + "epoch": 7.1761885737115465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17962 + }, + { + "epoch": 7.176588094286855, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17963 + }, + { + "epoch": 7.176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17964 + }, + { + "epoch": 7.177387135437475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17965 + }, + { + "epoch": 7.177786656012785, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17966 + }, + { + "epoch": 7.1781861765880945, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17967 + }, + { + "epoch": 7.178585697163404, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17968 + }, + { + "epoch": 7.178985217738713, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17969 + }, + { + "epoch": 7.179384738314023, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17970 + }, + { + "epoch": 7.179784258889333, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17971 + }, + { + "epoch": 7.180183779464643, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17972 + }, + { + "epoch": 7.180583300039952, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17973 + }, + { + "epoch": 7.180982820615261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17974 + }, + { + "epoch": 7.181382341190571, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17975 + }, + { + "epoch": 7.181781861765881, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17976 + }, + { + "epoch": 7.182181382341191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17977 + }, + { + "epoch": 7.1825809029165, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17978 + }, + { + "epoch": 7.18298042349181, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17979 + }, + { + "epoch": 7.183379944067119, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17980 + }, + { + "epoch": 7.183779464642429, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17981 + }, + { + "epoch": 7.184178985217739, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17982 + }, + { + "epoch": 7.1845785057930485, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17983 + }, + { + "epoch": 7.184978026368358, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17984 + }, + { + "epoch": 7.185377546943668, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17985 + }, + { + "epoch": 7.185777067518977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17986 + }, + { + "epoch": 7.186176588094287, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17987 + }, + { + "epoch": 7.1865761086695965, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17988 + }, + { + "epoch": 7.186975629244906, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17989 + }, + { + "epoch": 7.187375149820216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17990 + }, + { + "epoch": 7.187774670395525, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17991 + }, + { + "epoch": 7.188174190970835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17992 + }, + { + "epoch": 7.1885737115461446, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17993 + }, + { + "epoch": 7.188973232121454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17994 + }, + { + "epoch": 7.189372752696764, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17995 + }, + { + "epoch": 7.189772273272074, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17996 + }, + { + "epoch": 7.190171793847383, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17997 + }, + { + "epoch": 7.190571314422693, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17998 + }, + { + "epoch": 7.190970834998002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17999 + }, + { + "epoch": 7.191370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 18000 + }, + { + "epoch": 7.191769876148622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18001 + }, + { + "epoch": 7.192169396723932, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18002 + }, + { + "epoch": 7.192568917299241, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18003 + }, + { + "epoch": 7.19296843787455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18004 + }, + { + "epoch": 7.19336795844986, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18005 + }, + { + "epoch": 7.19376747902517, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18006 + }, + { + "epoch": 7.19416699960048, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18007 + }, + { + "epoch": 7.194566520175789, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18008 + }, + { + "epoch": 7.1949660407510985, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18009 + }, + { + "epoch": 7.195365561326408, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18010 + }, + { + "epoch": 7.195765081901718, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18011 + }, + { + "epoch": 7.196164602477028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18012 + }, + { + "epoch": 7.196564123052338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18013 + }, + { + "epoch": 7.1969636436276465, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18014 + }, + { + "epoch": 7.197363164202956, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18015 + }, + { + "epoch": 7.197762684778266, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18016 + }, + { + "epoch": 7.198162205353576, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18017 + }, + { + "epoch": 7.198561725928886, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18018 + }, + { + "epoch": 7.1989612465041946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18019 + }, + { + "epoch": 7.199360767079504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18020 + }, + { + "epoch": 7.199760287654814, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18021 + }, + { + "epoch": 7.200159808230124, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18022 + }, + { + "epoch": 7.200559328805434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18023 + }, + { + "epoch": 7.2009588493807435, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18024 + }, + { + "epoch": 7.201358369956052, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18025 + }, + { + "epoch": 7.201757890531362, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18026 + }, + { + "epoch": 7.202157411106672, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18027 + }, + { + "epoch": 7.202556931681982, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18028 + }, + { + "epoch": 7.2029564522572915, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18029 + }, + { + "epoch": 7.2033559728326, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18030 + }, + { + "epoch": 7.20375549340791, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18031 + }, + { + "epoch": 7.20415501398322, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18032 + }, + { + "epoch": 7.20455453455853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18033 + }, + { + "epoch": 7.20495405513384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 18034 + }, + { + "epoch": 7.205353575709149, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18035 + }, + { + "epoch": 7.205753096284458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18036 + }, + { + "epoch": 7.206152616859768, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18037 + }, + { + "epoch": 7.206552137435078, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18038 + }, + { + "epoch": 7.206951658010388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18039 + }, + { + "epoch": 7.207351178585697, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18040 + }, + { + "epoch": 7.207750699161007, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18041 + }, + { + "epoch": 7.208150219736316, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18042 + }, + { + "epoch": 7.208549740311626, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18043 + }, + { + "epoch": 7.208949260886936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18044 + }, + { + "epoch": 7.2093487814622454, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18045 + }, + { + "epoch": 7.209748302037555, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18046 + }, + { + "epoch": 7.210147822612864, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18047 + }, + { + "epoch": 7.210547343188174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18048 + }, + { + "epoch": 7.210946863763484, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18049 + }, + { + "epoch": 7.2113463843387935, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18050 + }, + { + "epoch": 7.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18051 + }, + { + "epoch": 7.212145425489413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18052 + }, + { + "epoch": 7.212544946064722, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18053 + }, + { + "epoch": 7.212944466640032, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18054 + }, + { + "epoch": 7.2133439872153415, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18055 + }, + { + "epoch": 7.213743507790651, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18056 + }, + { + "epoch": 7.214143028365961, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 18057 + }, + { + "epoch": 7.214542548941271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18058 + }, + { + "epoch": 7.21494206951658, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18059 + }, + { + "epoch": 7.21534159009189, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18060 + }, + { + "epoch": 7.215741110667199, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18061 + }, + { + "epoch": 7.216140631242509, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18062 + }, + { + "epoch": 7.216540151817819, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18063 + }, + { + "epoch": 7.216939672393128, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18064 + }, + { + "epoch": 7.217339192968438, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18065 + }, + { + "epoch": 7.217738713543747, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18066 + }, + { + "epoch": 7.218138234119057, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18067 + }, + { + "epoch": 7.218537754694367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18068 + }, + { + "epoch": 7.218937275269677, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18069 + }, + { + "epoch": 7.219336795844986, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18070 + }, + { + "epoch": 7.2197363164202955, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18071 + }, + { + "epoch": 7.220135836995605, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18072 + }, + { + "epoch": 7.220535357570915, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18073 + }, + { + "epoch": 7.220934878146225, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18074 + }, + { + "epoch": 7.221334398721535, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18075 + }, + { + "epoch": 7.2217339192968435, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18076 + }, + { + "epoch": 7.222133439872153, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18077 + }, + { + "epoch": 7.222532960447463, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 18078 + }, + { + "epoch": 7.222932481022773, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18079 + }, + { + "epoch": 7.223332001598083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18080 + }, + { + "epoch": 7.2237315221733915, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18081 + }, + { + "epoch": 7.224131042748701, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 18082 + }, + { + "epoch": 7.224530563324011, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18083 + }, + { + "epoch": 7.224930083899321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18084 + }, + { + "epoch": 7.225329604474631, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18085 + }, + { + "epoch": 7.2257291250499405, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18086 + }, + { + "epoch": 7.226128645625249, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18087 + }, + { + "epoch": 7.226528166200559, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18088 + }, + { + "epoch": 7.226927686775869, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18089 + }, + { + "epoch": 7.227327207351179, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18090 + }, + { + "epoch": 7.2277267279264885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18091 + }, + { + "epoch": 7.228126248501797, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18092 + }, + { + "epoch": 7.228525769077107, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18093 + }, + { + "epoch": 7.228925289652417, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18094 + }, + { + "epoch": 7.229324810227727, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18095 + }, + { + "epoch": 7.229724330803037, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18096 + }, + { + "epoch": 7.230123851378346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18097 + }, + { + "epoch": 7.230523371953655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18098 + }, + { + "epoch": 7.230922892528965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18099 + }, + { + "epoch": 7.231322413104275, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18100 + }, + { + "epoch": 7.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18101 + }, + { + "epoch": 7.232121454254894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18102 + }, + { + "epoch": 7.232520974830204, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18103 + }, + { + "epoch": 7.232920495405513, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18104 + }, + { + "epoch": 7.233320015980823, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18105 + }, + { + "epoch": 7.233719536556133, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18106 + }, + { + "epoch": 7.234119057131442, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18107 + }, + { + "epoch": 7.234518577706752, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18108 + }, + { + "epoch": 7.234918098282061, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18109 + }, + { + "epoch": 7.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18110 + }, + { + "epoch": 7.235717139432681, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18111 + }, + { + "epoch": 7.2361166600079905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18112 + }, + { + "epoch": 7.2365161805833, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18113 + }, + { + "epoch": 7.23691570115861, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18114 + }, + { + "epoch": 7.237315221733919, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18115 + }, + { + "epoch": 7.237714742309229, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 18116 + }, + { + "epoch": 7.2381142628845385, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18117 + }, + { + "epoch": 7.238513783459848, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18118 + }, + { + "epoch": 7.238913304035158, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18119 + }, + { + "epoch": 7.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18120 + }, + { + "epoch": 7.239712345185777, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18121 + }, + { + "epoch": 7.240111865761087, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18122 + }, + { + "epoch": 7.240511386336396, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18123 + }, + { + "epoch": 7.240910906911706, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18124 + }, + { + "epoch": 7.241310427487016, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18125 + }, + { + "epoch": 7.241709948062325, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18126 + }, + { + "epoch": 7.242109468637635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18127 + }, + { + "epoch": 7.242508989212944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18128 + }, + { + "epoch": 7.242908509788254, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18129 + }, + { + "epoch": 7.243308030363564, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 18130 + }, + { + "epoch": 7.243707550938874, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18131 + }, + { + "epoch": 7.244107071514183, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18132 + }, + { + "epoch": 7.244506592089492, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18133 + }, + { + "epoch": 7.244906112664802, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18134 + }, + { + "epoch": 7.245305633240112, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 18135 + }, + { + "epoch": 7.245705153815422, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 18136 + }, + { + "epoch": 7.246104674390731, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18137 + }, + { + "epoch": 7.2465041949660405, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18138 + }, + { + "epoch": 7.24690371554135, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18139 + }, + { + "epoch": 7.24730323611666, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18140 + }, + { + "epoch": 7.24770275669197, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18141 + }, + { + "epoch": 7.24810227726728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18142 + }, + { + "epoch": 7.2485017978425885, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18143 + }, + { + "epoch": 7.248901318417898, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18144 + }, + { + "epoch": 7.249300838993208, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18145 + }, + { + "epoch": 7.249700359568518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18146 + }, + { + "epoch": 7.250099880143828, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18147 + }, + { + "epoch": 7.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18148 + }, + { + "epoch": 7.250898921294446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18149 + }, + { + "epoch": 7.251298441869756, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18150 + }, + { + "epoch": 7.251697962445066, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18151 + }, + { + "epoch": 7.252097483020376, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18152 + }, + { + "epoch": 7.2524970035956855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18153 + }, + { + "epoch": 7.252896524170994, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18154 + }, + { + "epoch": 7.253296044746304, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18155 + }, + { + "epoch": 7.253695565321614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18156 + }, + { + "epoch": 7.254095085896924, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18157 + }, + { + "epoch": 7.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18158 + }, + { + "epoch": 7.254894127047543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18159 + }, + { + "epoch": 7.255293647622852, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18160 + }, + { + "epoch": 7.255693168198162, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18161 + }, + { + "epoch": 7.256092688773472, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18162 + }, + { + "epoch": 7.256492209348782, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18163 + }, + { + "epoch": 7.256891729924091, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18164 + }, + { + "epoch": 7.257291250499401, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 18165 + }, + { + "epoch": 7.25769077107471, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18166 + }, + { + "epoch": 7.25809029165002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18167 + }, + { + "epoch": 7.25848981222533, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18168 + }, + { + "epoch": 7.258889332800639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18169 + }, + { + "epoch": 7.259288853375949, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 18170 + }, + { + "epoch": 7.259688373951258, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18171 + }, + { + "epoch": 7.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18172 + }, + { + "epoch": 7.260487415101878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18173 + }, + { + "epoch": 7.2608869356771875, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 18174 + }, + { + "epoch": 7.261286456252497, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18175 + }, + { + "epoch": 7.261685976827807, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18176 + }, + { + "epoch": 7.262085497403116, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18177 + }, + { + "epoch": 7.262485017978426, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18178 + }, + { + "epoch": 7.2628845385537355, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18179 + }, + { + "epoch": 7.263284059129045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18180 + }, + { + "epoch": 7.263683579704355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18181 + }, + { + "epoch": 7.264083100279665, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18182 + }, + { + "epoch": 7.264482620854974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18183 + }, + { + "epoch": 7.2648821414302835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18184 + }, + { + "epoch": 7.265281662005593, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18185 + }, + { + "epoch": 7.265681182580903, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18186 + }, + { + "epoch": 7.266080703156213, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18187 + }, + { + "epoch": 7.266480223731522, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 18188 + }, + { + "epoch": 7.266879744306832, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18189 + }, + { + "epoch": 7.267279264882141, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18190 + }, + { + "epoch": 7.267678785457451, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18191 + }, + { + "epoch": 7.268078306032761, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18192 + }, + { + "epoch": 7.26847782660807, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 18193 + }, + { + "epoch": 7.26887734718338, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18194 + }, + { + "epoch": 7.269276867758689, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18195 + }, + { + "epoch": 7.269676388333999, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18196 + }, + { + "epoch": 7.270075908909309, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18197 + }, + { + "epoch": 7.270475429484619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18198 + }, + { + "epoch": 7.270874950059928, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18199 + }, + { + "epoch": 7.2712744706352375, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18200 + }, + { + "epoch": 7.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18201 + }, + { + "epoch": 7.272073511785857, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 18202 + }, + { + "epoch": 7.272473032361167, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18203 + }, + { + "epoch": 7.272872552936477, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18204 + }, + { + "epoch": 7.2732720735117855, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18205 + }, + { + "epoch": 7.273671594087095, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18206 + }, + { + "epoch": 7.274071114662405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18207 + }, + { + "epoch": 7.274470635237715, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18208 + }, + { + "epoch": 7.274870155813025, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18209 + }, + { + "epoch": 7.2752696763883336, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18210 + }, + { + "epoch": 7.275669196963643, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18211 + }, + { + "epoch": 7.276068717538953, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2981, + "step": 18212 + }, + { + "epoch": 7.276468238114263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18213 + }, + { + "epoch": 7.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18214 + }, + { + "epoch": 7.2772672792648825, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18215 + }, + { + "epoch": 7.277666799840191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18216 + }, + { + "epoch": 7.278066320415501, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18217 + }, + { + "epoch": 7.278465840990811, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18218 + }, + { + "epoch": 7.278865361566121, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18219 + }, + { + "epoch": 7.2792648821414305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18220 + }, + { + "epoch": 7.27966440271674, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18221 + }, + { + "epoch": 7.280063923292049, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 18222 + }, + { + "epoch": 7.280463443867359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18223 + }, + { + "epoch": 7.280862964442669, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18224 + }, + { + "epoch": 7.281262485017979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18225 + }, + { + "epoch": 7.281662005593288, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18226 + }, + { + "epoch": 7.282061526168597, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18227 + }, + { + "epoch": 7.282461046743907, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18228 + }, + { + "epoch": 7.282860567319217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 18229 + }, + { + "epoch": 7.283260087894527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18230 + }, + { + "epoch": 7.283659608469836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 18231 + }, + { + "epoch": 7.284059129045146, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18232 + }, + { + "epoch": 7.284458649620455, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18233 + }, + { + "epoch": 7.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18234 + }, + { + "epoch": 7.285257690771075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18235 + }, + { + "epoch": 7.285657211346384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18236 + }, + { + "epoch": 7.286056731921694, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18237 + }, + { + "epoch": 7.286456252497004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18238 + }, + { + "epoch": 7.286855773072313, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18239 + }, + { + "epoch": 7.287255293647623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18240 + }, + { + "epoch": 7.2876548142229325, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18241 + }, + { + "epoch": 7.288054334798242, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18242 + }, + { + "epoch": 7.288453855373552, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18243 + }, + { + "epoch": 7.288853375948861, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18244 + }, + { + "epoch": 7.289252896524171, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18245 + }, + { + "epoch": 7.2896524170994805, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18246 + }, + { + "epoch": 7.29005193767479, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18247 + }, + { + "epoch": 7.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18248 + }, + { + "epoch": 7.29085097882541, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18249 + }, + { + "epoch": 7.291250499400719, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18250 + }, + { + "epoch": 7.291650019976029, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18251 + }, + { + "epoch": 7.292049540551338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18252 + }, + { + "epoch": 7.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18253 + }, + { + "epoch": 7.292848581701958, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18254 + }, + { + "epoch": 7.293248102277268, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18255 + }, + { + "epoch": 7.293647622852577, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18256 + }, + { + "epoch": 7.294047143427886, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18257 + }, + { + "epoch": 7.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18258 + }, + { + "epoch": 7.294846184578506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18259 + }, + { + "epoch": 7.295245705153816, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18260 + }, + { + "epoch": 7.295645225729125, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18261 + }, + { + "epoch": 7.2960447463044344, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18262 + }, + { + "epoch": 7.296444266879744, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18263 + }, + { + "epoch": 7.296843787455054, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18264 + }, + { + "epoch": 7.297243308030364, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18265 + }, + { + "epoch": 7.297642828605674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18266 + }, + { + "epoch": 7.2980423491809825, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18267 + }, + { + "epoch": 7.298441869756292, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 18268 + }, + { + "epoch": 7.298841390331602, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18269 + }, + { + "epoch": 7.299240910906912, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18270 + }, + { + "epoch": 7.299640431482222, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18271 + }, + { + "epoch": 7.300039952057531, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18272 + }, + { + "epoch": 7.30043947263284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18273 + }, + { + "epoch": 7.30083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18274 + }, + { + "epoch": 7.30123851378346, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18275 + }, + { + "epoch": 7.30163803435877, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18276 + }, + { + "epoch": 7.3020375549340795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18277 + }, + { + "epoch": 7.302437075509388, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18278 + }, + { + "epoch": 7.302836596084698, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18279 + }, + { + "epoch": 7.303236116660008, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18280 + }, + { + "epoch": 7.303635637235318, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18281 + }, + { + "epoch": 7.3040351578106275, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18282 + }, + { + "epoch": 7.304434678385936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18283 + }, + { + "epoch": 7.304834198961246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 18284 + }, + { + "epoch": 7.305233719536556, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18285 + }, + { + "epoch": 7.305633240111866, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18286 + }, + { + "epoch": 7.3060327606871756, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18287 + }, + { + "epoch": 7.306432281262485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18288 + }, + { + "epoch": 7.306831801837794, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18289 + }, + { + "epoch": 7.307231322413104, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18290 + }, + { + "epoch": 7.307630842988414, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18291 + }, + { + "epoch": 7.308030363563724, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18292 + }, + { + "epoch": 7.308429884139033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18293 + }, + { + "epoch": 7.308829404714343, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18294 + }, + { + "epoch": 7.309228925289652, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18295 + }, + { + "epoch": 7.309628445864962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18296 + }, + { + "epoch": 7.310027966440272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18297 + }, + { + "epoch": 7.310427487015581, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18298 + }, + { + "epoch": 7.310827007590891, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18299 + }, + { + "epoch": 7.3112265281662, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18300 + }, + { + "epoch": 7.31162604874151, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18301 + }, + { + "epoch": 7.31202556931682, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18302 + }, + { + "epoch": 7.3124250898921295, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18303 + }, + { + "epoch": 7.312824610467439, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18304 + }, + { + "epoch": 7.313224131042749, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18305 + }, + { + "epoch": 7.313623651618058, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18306 + }, + { + "epoch": 7.314023172193368, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 18307 + }, + { + "epoch": 7.3144226927686775, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18308 + }, + { + "epoch": 7.314822213343987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18309 + }, + { + "epoch": 7.315221733919297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18310 + }, + { + "epoch": 7.315621254494607, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18311 + }, + { + "epoch": 7.316020775069916, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18312 + }, + { + "epoch": 7.316420295645226, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 18313 + }, + { + "epoch": 7.316819816220535, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18314 + }, + { + "epoch": 7.317219336795845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18315 + }, + { + "epoch": 7.317618857371155, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18316 + }, + { + "epoch": 7.318018377946464, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 18317 + }, + { + "epoch": 7.318417898521774, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18318 + }, + { + "epoch": 7.318817419097083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18319 + }, + { + "epoch": 7.319216939672393, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18320 + }, + { + "epoch": 7.319616460247703, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18321 + }, + { + "epoch": 7.320015980823013, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18322 + }, + { + "epoch": 7.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18323 + }, + { + "epoch": 7.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18324 + }, + { + "epoch": 7.321214542548941, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18325 + }, + { + "epoch": 7.321614063124251, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18326 + }, + { + "epoch": 7.322013583699561, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18327 + }, + { + "epoch": 7.322413104274871, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18328 + }, + { + "epoch": 7.3228126248501795, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18329 + }, + { + "epoch": 7.323212145425489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18330 + }, + { + "epoch": 7.323611666000799, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18331 + }, + { + "epoch": 7.324011186576109, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18332 + }, + { + "epoch": 7.324410707151419, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18333 + }, + { + "epoch": 7.3248102277267275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18334 + }, + { + "epoch": 7.325209748302037, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18335 + }, + { + "epoch": 7.325609268877347, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18336 + }, + { + "epoch": 7.326008789452657, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18337 + }, + { + "epoch": 7.326408310027967, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18338 + }, + { + "epoch": 7.3268078306032765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18339 + }, + { + "epoch": 7.327207351178585, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18340 + }, + { + "epoch": 7.327606871753895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18341 + }, + { + "epoch": 7.328006392329205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18342 + }, + { + "epoch": 7.328405912904515, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18343 + }, + { + "epoch": 7.3288054334798245, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18344 + }, + { + "epoch": 7.329204954055134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18345 + }, + { + "epoch": 7.329604474630443, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18346 + }, + { + "epoch": 7.330003995205753, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18347 + }, + { + "epoch": 7.330403515781063, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 18348 + }, + { + "epoch": 7.3308030363563725, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18349 + }, + { + "epoch": 7.331202556931682, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18350 + }, + { + "epoch": 7.331602077506991, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18351 + }, + { + "epoch": 7.332001598082301, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18352 + }, + { + "epoch": 7.332401118657611, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18353 + }, + { + "epoch": 7.332800639232921, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18354 + }, + { + "epoch": 7.33320015980823, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18355 + }, + { + "epoch": 7.33359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18356 + }, + { + "epoch": 7.333999200958849, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18357 + }, + { + "epoch": 7.334398721534159, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18358 + }, + { + "epoch": 7.334798242109469, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18359 + }, + { + "epoch": 7.335197762684778, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18360 + }, + { + "epoch": 7.335597283260088, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18361 + }, + { + "epoch": 7.335996803835398, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18362 + }, + { + "epoch": 7.336396324410707, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18363 + }, + { + "epoch": 7.336795844986017, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18364 + }, + { + "epoch": 7.3371953655613265, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18365 + }, + { + "epoch": 7.337594886136636, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18366 + }, + { + "epoch": 7.337994406711946, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18367 + }, + { + "epoch": 7.338393927287255, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18368 + }, + { + "epoch": 7.338793447862565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18369 + }, + { + "epoch": 7.3391929684378745, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18370 + }, + { + "epoch": 7.339592489013184, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18371 + }, + { + "epoch": 7.339992009588494, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18372 + }, + { + "epoch": 7.340391530163803, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18373 + }, + { + "epoch": 7.340791050739113, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18374 + }, + { + "epoch": 7.3411905713144225, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18375 + }, + { + "epoch": 7.341590091889732, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18376 + }, + { + "epoch": 7.341989612465042, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18377 + }, + { + "epoch": 7.342389133040352, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18378 + }, + { + "epoch": 7.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18379 + }, + { + "epoch": 7.343188174190971, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18380 + }, + { + "epoch": 7.34358769476628, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18381 + }, + { + "epoch": 7.34398721534159, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18382 + }, + { + "epoch": 7.3443867359169, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 18383 + }, + { + "epoch": 7.34478625649221, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18384 + }, + { + "epoch": 7.345185777067519, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18385 + }, + { + "epoch": 7.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18386 + }, + { + "epoch": 7.345984818218138, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18387 + }, + { + "epoch": 7.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18388 + }, + { + "epoch": 7.346783859368758, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18389 + }, + { + "epoch": 7.347183379944067, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18390 + }, + { + "epoch": 7.3475829005193765, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18391 + }, + { + "epoch": 7.347982421094686, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18392 + }, + { + "epoch": 7.348381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18393 + }, + { + "epoch": 7.348781462245306, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18394 + }, + { + "epoch": 7.349180982820616, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18395 + }, + { + "epoch": 7.3495805033959245, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18396 + }, + { + "epoch": 7.349980023971234, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18397 + }, + { + "epoch": 7.350379544546544, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18398 + }, + { + "epoch": 7.350779065121854, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18399 + }, + { + "epoch": 7.351178585697164, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18400 + }, + { + "epoch": 7.351578106272473, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18401 + }, + { + "epoch": 7.351977626847782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18402 + }, + { + "epoch": 7.352377147423092, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 18403 + }, + { + "epoch": 7.352776667998402, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18404 + }, + { + "epoch": 7.353176188573712, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18405 + }, + { + "epoch": 7.3535757091490215, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18406 + }, + { + "epoch": 7.35397522972433, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18407 + }, + { + "epoch": 7.35437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18408 + }, + { + "epoch": 7.35477427087495, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18409 + }, + { + "epoch": 7.35517379145026, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18410 + }, + { + "epoch": 7.3555733120255695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18411 + }, + { + "epoch": 7.355972832600879, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18412 + }, + { + "epoch": 7.356372353176188, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18413 + }, + { + "epoch": 7.356771873751498, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18414 + }, + { + "epoch": 7.357171394326808, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 18415 + }, + { + "epoch": 7.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18416 + }, + { + "epoch": 7.357970435477427, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18417 + }, + { + "epoch": 7.358369956052737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18418 + }, + { + "epoch": 7.358769476628046, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18419 + }, + { + "epoch": 7.359168997203356, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18420 + }, + { + "epoch": 7.359568517778666, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 18421 + }, + { + "epoch": 7.359968038353975, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18422 + }, + { + "epoch": 7.360367558929285, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 18423 + }, + { + "epoch": 7.360767079504594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18424 + }, + { + "epoch": 7.361166600079904, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18425 + }, + { + "epoch": 7.361566120655214, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18426 + }, + { + "epoch": 7.361965641230523, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18427 + }, + { + "epoch": 7.362365161805833, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18428 + }, + { + "epoch": 7.362764682381143, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18429 + }, + { + "epoch": 7.363164202956452, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18430 + }, + { + "epoch": 7.363563723531762, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18431 + }, + { + "epoch": 7.3639632441070715, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18432 + }, + { + "epoch": 7.364362764682381, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18433 + }, + { + "epoch": 7.364762285257691, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18434 + }, + { + "epoch": 7.365161805833001, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18435 + }, + { + "epoch": 7.36556132640831, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18436 + }, + { + "epoch": 7.3659608469836195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18437 + }, + { + "epoch": 7.366360367558929, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18438 + }, + { + "epoch": 7.366759888134239, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18439 + }, + { + "epoch": 7.367159408709549, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18440 + }, + { + "epoch": 7.367558929284858, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18441 + }, + { + "epoch": 7.367958449860168, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18442 + }, + { + "epoch": 7.368357970435477, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18443 + }, + { + "epoch": 7.368757491010787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18444 + }, + { + "epoch": 7.369157011586097, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18445 + }, + { + "epoch": 7.369556532161407, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18446 + }, + { + "epoch": 7.369956052736716, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18447 + }, + { + "epoch": 7.370355573312025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18448 + }, + { + "epoch": 7.370755093887335, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18449 + }, + { + "epoch": 7.371154614462645, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18450 + }, + { + "epoch": 7.371554135037955, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18451 + }, + { + "epoch": 7.3719536556132645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18452 + }, + { + "epoch": 7.372353176188573, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18453 + }, + { + "epoch": 7.372752696763883, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18454 + }, + { + "epoch": 7.373152217339193, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18455 + }, + { + "epoch": 7.373551737914503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18456 + }, + { + "epoch": 7.373951258489813, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 18457 + }, + { + "epoch": 7.3743507790651215, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18458 + }, + { + "epoch": 7.374750299640431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18459 + }, + { + "epoch": 7.375149820215741, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18460 + }, + { + "epoch": 7.375549340791051, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18461 + }, + { + "epoch": 7.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18462 + }, + { + "epoch": 7.3763483819416695, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18463 + }, + { + "epoch": 7.376747902516979, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18464 + }, + { + "epoch": 7.377147423092289, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18465 + }, + { + "epoch": 7.377546943667599, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 18466 + }, + { + "epoch": 7.377946464242909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18467 + }, + { + "epoch": 7.3783459848182185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18468 + }, + { + "epoch": 7.378745505393527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18469 + }, + { + "epoch": 7.379145025968837, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18470 + }, + { + "epoch": 7.379544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18471 + }, + { + "epoch": 7.379944067119457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18472 + }, + { + "epoch": 7.3803435876947665, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18473 + }, + { + "epoch": 7.380743108270076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18474 + }, + { + "epoch": 7.381142628845385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18475 + }, + { + "epoch": 7.381542149420695, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18476 + }, + { + "epoch": 7.381941669996005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18477 + }, + { + "epoch": 7.3823411905713145, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18478 + }, + { + "epoch": 7.382740711146624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18479 + }, + { + "epoch": 7.383140231721933, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18480 + }, + { + "epoch": 7.383539752297243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18481 + }, + { + "epoch": 7.383939272872553, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18482 + }, + { + "epoch": 7.384338793447863, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18483 + }, + { + "epoch": 7.384738314023172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18484 + }, + { + "epoch": 7.385137834598482, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18485 + }, + { + "epoch": 7.385537355173791, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18486 + }, + { + "epoch": 7.385936875749101, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 18487 + }, + { + "epoch": 7.386336396324411, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18488 + }, + { + "epoch": 7.38673591689972, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18489 + }, + { + "epoch": 7.38713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18490 + }, + { + "epoch": 7.38753495805034, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18491 + }, + { + "epoch": 7.387934478625649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18492 + }, + { + "epoch": 7.388333999200959, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18493 + }, + { + "epoch": 7.3887335197762685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18494 + }, + { + "epoch": 7.389133040351578, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18495 + }, + { + "epoch": 7.389532560926888, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18496 + }, + { + "epoch": 7.389932081502197, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18497 + }, + { + "epoch": 7.390331602077507, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18498 + }, + { + "epoch": 7.3907311226528165, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18499 + }, + { + "epoch": 7.391130643228126, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18500 + }, + { + "epoch": 7.391530163803436, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18501 + }, + { + "epoch": 7.391929684378746, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 18502 + }, + { + "epoch": 7.392329204954055, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18503 + }, + { + "epoch": 7.3927287255293646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18504 + }, + { + "epoch": 7.393128246104674, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18505 + }, + { + "epoch": 7.393527766679984, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18506 + }, + { + "epoch": 7.393927287255294, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18507 + }, + { + "epoch": 7.394326807830604, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18508 + }, + { + "epoch": 7.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18509 + }, + { + "epoch": 7.395125848981222, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18510 + }, + { + "epoch": 7.395525369556532, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18511 + }, + { + "epoch": 7.395924890131842, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18512 + }, + { + "epoch": 7.396324410707152, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18513 + }, + { + "epoch": 7.396723931282461, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 18514 + }, + { + "epoch": 7.39712345185777, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18515 + }, + { + "epoch": 7.39752297243308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18516 + }, + { + "epoch": 7.39792249300839, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18517 + }, + { + "epoch": 7.3983220135837, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18518 + }, + { + "epoch": 7.39872153415901, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18519 + }, + { + "epoch": 7.3991210547343185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 18520 + }, + { + "epoch": 7.399520575309628, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18521 + }, + { + "epoch": 7.399920095884938, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18522 + }, + { + "epoch": 7.400319616460248, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18523 + }, + { + "epoch": 7.400719137035558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18524 + }, + { + "epoch": 7.401118657610867, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18525 + }, + { + "epoch": 7.401518178186176, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18526 + }, + { + "epoch": 7.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18527 + }, + { + "epoch": 7.402317219336796, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18528 + }, + { + "epoch": 7.402716739912106, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18529 + }, + { + "epoch": 7.4031162604874154, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18530 + }, + { + "epoch": 7.403515781062724, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 18531 + }, + { + "epoch": 7.403915301638034, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18532 + }, + { + "epoch": 7.404314822213344, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18533 + }, + { + "epoch": 7.404714342788654, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18534 + }, + { + "epoch": 7.4051138633639635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 18535 + }, + { + "epoch": 7.405513383939273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18536 + }, + { + "epoch": 7.405912904514582, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18537 + }, + { + "epoch": 7.406312425089892, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18538 + }, + { + "epoch": 7.406711945665202, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18539 + }, + { + "epoch": 7.4071114662405115, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18540 + }, + { + "epoch": 7.407510986815821, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18541 + }, + { + "epoch": 7.407910507391131, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18542 + }, + { + "epoch": 7.40831002796644, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18543 + }, + { + "epoch": 7.40870954854175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18544 + }, + { + "epoch": 7.40910906911706, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18545 + }, + { + "epoch": 7.409508589692369, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18546 + }, + { + "epoch": 7.409908110267679, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18547 + }, + { + "epoch": 7.410307630842988, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18548 + }, + { + "epoch": 7.410707151418298, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18549 + }, + { + "epoch": 7.411106671993608, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18550 + }, + { + "epoch": 7.411506192568917, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18551 + }, + { + "epoch": 7.411905713144227, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18552 + }, + { + "epoch": 7.412305233719536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18553 + }, + { + "epoch": 7.412704754294846, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18554 + }, + { + "epoch": 7.413104274870156, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18555 + }, + { + "epoch": 7.4135037954454654, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18556 + }, + { + "epoch": 7.413903316020775, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18557 + }, + { + "epoch": 7.414302836596085, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18558 + }, + { + "epoch": 7.414702357171394, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18559 + }, + { + "epoch": 7.415101877746704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18560 + }, + { + "epoch": 7.4155013983220135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18561 + }, + { + "epoch": 7.415900918897323, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18562 + }, + { + "epoch": 7.416300439472633, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18563 + }, + { + "epoch": 7.416699960047943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18564 + }, + { + "epoch": 7.417099480623252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18565 + }, + { + "epoch": 7.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.2993, + "step": 18566 + }, + { + "epoch": 7.417898521773871, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18567 + }, + { + "epoch": 7.418298042349181, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18568 + }, + { + "epoch": 7.418697562924491, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18569 + }, + { + "epoch": 7.4190970834998, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18570 + }, + { + "epoch": 7.41949660407511, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18571 + }, + { + "epoch": 7.419896124650419, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18572 + }, + { + "epoch": 7.420295645225729, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18573 + }, + { + "epoch": 7.420695165801039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18574 + }, + { + "epoch": 7.421094686376349, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18575 + }, + { + "epoch": 7.421494206951658, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18576 + }, + { + "epoch": 7.421893727526967, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18577 + }, + { + "epoch": 7.422293248102277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18578 + }, + { + "epoch": 7.422692768677587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18579 + }, + { + "epoch": 7.423092289252897, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18580 + }, + { + "epoch": 7.4234918098282066, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 18581 + }, + { + "epoch": 7.4238913304035155, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18582 + }, + { + "epoch": 7.424290850978825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18583 + }, + { + "epoch": 7.424690371554135, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18584 + }, + { + "epoch": 7.425089892129445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18585 + }, + { + "epoch": 7.425489412704755, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18586 + }, + { + "epoch": 7.4258889332800635, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18587 + }, + { + "epoch": 7.426288453855373, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18588 + }, + { + "epoch": 7.426687974430683, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18589 + }, + { + "epoch": 7.427087495005993, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18590 + }, + { + "epoch": 7.427487015581303, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18591 + }, + { + "epoch": 7.427886536156612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18592 + }, + { + "epoch": 7.428286056731921, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18593 + }, + { + "epoch": 7.428685577307231, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18594 + }, + { + "epoch": 7.429085097882541, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18595 + }, + { + "epoch": 7.429484618457851, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18596 + }, + { + "epoch": 7.4298841390331605, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18597 + }, + { + "epoch": 7.43028365960847, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18598 + }, + { + "epoch": 7.430683180183779, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18599 + }, + { + "epoch": 7.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18600 + }, + { + "epoch": 7.431482221334399, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18601 + }, + { + "epoch": 7.4318817419097085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18602 + }, + { + "epoch": 7.432281262485018, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18603 + }, + { + "epoch": 7.432680783060327, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18604 + }, + { + "epoch": 7.433080303635637, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18605 + }, + { + "epoch": 7.433479824210947, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18606 + }, + { + "epoch": 7.433879344786257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18607 + }, + { + "epoch": 7.434278865361566, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18608 + }, + { + "epoch": 7.434678385936876, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18609 + }, + { + "epoch": 7.435077906512185, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 18610 + }, + { + "epoch": 7.435477427087495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18611 + }, + { + "epoch": 7.435876947662805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18612 + }, + { + "epoch": 7.436276468238114, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 18613 + }, + { + "epoch": 7.436675988813424, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18614 + }, + { + "epoch": 7.437075509388734, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18615 + }, + { + "epoch": 7.437475029964043, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18616 + }, + { + "epoch": 7.437874550539353, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18617 + }, + { + "epoch": 7.438274071114662, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18618 + }, + { + "epoch": 7.438673591689972, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18619 + }, + { + "epoch": 7.439073112265282, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18620 + }, + { + "epoch": 7.439472632840591, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18621 + }, + { + "epoch": 7.439872153415901, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18622 + }, + { + "epoch": 7.4402716739912105, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18623 + }, + { + "epoch": 7.44067119456652, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18624 + }, + { + "epoch": 7.44107071514183, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3015, + "step": 18625 + }, + { + "epoch": 7.44147023571714, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18626 + }, + { + "epoch": 7.441869756292449, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18627 + }, + { + "epoch": 7.4422692768677585, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18628 + }, + { + "epoch": 7.442668797443068, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18629 + }, + { + "epoch": 7.443068318018378, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18630 + }, + { + "epoch": 7.443467838593688, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18631 + }, + { + "epoch": 7.443867359168998, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 18632 + }, + { + "epoch": 7.444266879744307, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18633 + }, + { + "epoch": 7.444666400319616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 18634 + }, + { + "epoch": 7.445065920894926, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 18635 + }, + { + "epoch": 7.445465441470236, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18636 + }, + { + "epoch": 7.445864962045546, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18637 + }, + { + "epoch": 7.446264482620855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18638 + }, + { + "epoch": 7.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18639 + }, + { + "epoch": 7.447063523771474, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 18640 + }, + { + "epoch": 7.447463044346784, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18641 + }, + { + "epoch": 7.447862564922094, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18642 + }, + { + "epoch": 7.448262085497403, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18643 + }, + { + "epoch": 7.448661606072712, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18644 + }, + { + "epoch": 7.449061126648022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18645 + }, + { + "epoch": 7.449460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18646 + }, + { + "epoch": 7.449860167798642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18647 + }, + { + "epoch": 7.450259688373952, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18648 + }, + { + "epoch": 7.4506592089492605, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18649 + }, + { + "epoch": 7.45105872952457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18650 + }, + { + "epoch": 7.45145825009988, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18651 + }, + { + "epoch": 7.45185777067519, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18652 + }, + { + "epoch": 7.4522572912505, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18653 + }, + { + "epoch": 7.452656811825809, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18654 + }, + { + "epoch": 7.453056332401118, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18655 + }, + { + "epoch": 7.453455852976428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18656 + }, + { + "epoch": 7.453855373551738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 18657 + }, + { + "epoch": 7.454254894127048, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18658 + }, + { + "epoch": 7.4546544147023575, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18659 + }, + { + "epoch": 7.455053935277666, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18660 + }, + { + "epoch": 7.455453455852976, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18661 + }, + { + "epoch": 7.455852976428286, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18662 + }, + { + "epoch": 7.456252497003596, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18663 + }, + { + "epoch": 7.4566520175789055, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18664 + }, + { + "epoch": 7.457051538154215, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18665 + }, + { + "epoch": 7.457451058729524, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18666 + }, + { + "epoch": 7.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18667 + }, + { + "epoch": 7.458250099880144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18668 + }, + { + "epoch": 7.4586496204554535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18669 + }, + { + "epoch": 7.459049141030763, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18670 + }, + { + "epoch": 7.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18671 + }, + { + "epoch": 7.459848182181382, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18672 + }, + { + "epoch": 7.460247702756692, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18673 + }, + { + "epoch": 7.460647223332002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18674 + }, + { + "epoch": 7.461046743907311, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 18675 + }, + { + "epoch": 7.461446264482621, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18676 + }, + { + "epoch": 7.46184578505793, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18677 + }, + { + "epoch": 7.46224530563324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18678 + }, + { + "epoch": 7.46264482620855, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18679 + }, + { + "epoch": 7.463044346783859, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18680 + }, + { + "epoch": 7.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18681 + }, + { + "epoch": 7.463843387934479, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18682 + }, + { + "epoch": 7.464242908509788, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18683 + }, + { + "epoch": 7.464642429085098, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18684 + }, + { + "epoch": 7.4650419496604075, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18685 + }, + { + "epoch": 7.465441470235717, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18686 + }, + { + "epoch": 7.465840990811027, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18687 + }, + { + "epoch": 7.466240511386337, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18688 + }, + { + "epoch": 7.466640031961646, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18689 + }, + { + "epoch": 7.4670395525369555, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18690 + }, + { + "epoch": 7.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18691 + }, + { + "epoch": 7.467838593687575, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18692 + }, + { + "epoch": 7.468238114262885, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18693 + }, + { + "epoch": 7.468637634838194, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18694 + }, + { + "epoch": 7.4690371554135035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18695 + }, + { + "epoch": 7.469436675988813, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18696 + }, + { + "epoch": 7.469836196564123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18697 + }, + { + "epoch": 7.470235717139433, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18698 + }, + { + "epoch": 7.470635237714743, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18699 + }, + { + "epoch": 7.471034758290052, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18700 + }, + { + "epoch": 7.471434278865361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 18701 + }, + { + "epoch": 7.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18702 + }, + { + "epoch": 7.472233320015981, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.2967, + "step": 18703 + }, + { + "epoch": 7.472632840591291, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18704 + }, + { + "epoch": 7.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18705 + }, + { + "epoch": 7.473431881741909, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18706 + }, + { + "epoch": 7.473831402317219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18707 + }, + { + "epoch": 7.474230922892529, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18708 + }, + { + "epoch": 7.474630443467839, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 18709 + }, + { + "epoch": 7.475029964043149, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 18710 + }, + { + "epoch": 7.4754294846184575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18711 + }, + { + "epoch": 7.475829005193767, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18712 + }, + { + "epoch": 7.476228525769077, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18713 + }, + { + "epoch": 7.476628046344387, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18714 + }, + { + "epoch": 7.477027566919697, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18715 + }, + { + "epoch": 7.477427087495006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 18716 + }, + { + "epoch": 7.477826608070315, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18717 + }, + { + "epoch": 7.478226128645625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18718 + }, + { + "epoch": 7.478625649220935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 18719 + }, + { + "epoch": 7.479025169796245, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18720 + }, + { + "epoch": 7.479424690371554, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18721 + }, + { + "epoch": 7.479824210946864, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18722 + }, + { + "epoch": 7.480223731522173, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18723 + }, + { + "epoch": 7.480623252097483, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18724 + }, + { + "epoch": 7.481022772672793, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18725 + }, + { + "epoch": 7.4814222932481025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18726 + }, + { + "epoch": 7.481821813823412, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18727 + }, + { + "epoch": 7.482221334398721, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18728 + }, + { + "epoch": 7.482620854974031, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18729 + }, + { + "epoch": 7.483020375549341, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18730 + }, + { + "epoch": 7.4834198961246505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18731 + }, + { + "epoch": 7.48381941669996, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18732 + }, + { + "epoch": 7.484218937275269, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18733 + }, + { + "epoch": 7.484618457850579, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18734 + }, + { + "epoch": 7.485017978425889, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18735 + }, + { + "epoch": 7.485417499001199, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18736 + }, + { + "epoch": 7.485817019576508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18737 + }, + { + "epoch": 7.486216540151818, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18738 + }, + { + "epoch": 7.486616060727127, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18739 + }, + { + "epoch": 7.487015581302437, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18740 + }, + { + "epoch": 7.487415101877747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18741 + }, + { + "epoch": 7.487814622453056, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18742 + }, + { + "epoch": 7.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18743 + }, + { + "epoch": 7.488613663603676, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18744 + }, + { + "epoch": 7.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18745 + }, + { + "epoch": 7.489412704754295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18746 + }, + { + "epoch": 7.4898122253296044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18747 + }, + { + "epoch": 7.490211745904914, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18748 + }, + { + "epoch": 7.490611266480224, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18749 + }, + { + "epoch": 7.491010787055533, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18750 + }, + { + "epoch": 7.491410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18751 + }, + { + "epoch": 7.4918098282061525, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18752 + }, + { + "epoch": 7.492209348781462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18753 + }, + { + "epoch": 7.492608869356772, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18754 + }, + { + "epoch": 7.493008389932082, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18755 + }, + { + "epoch": 7.493407910507391, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18756 + }, + { + "epoch": 7.4938074310827005, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18757 + }, + { + "epoch": 7.49420695165801, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18758 + }, + { + "epoch": 7.49460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18759 + }, + { + "epoch": 7.49500599280863, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18760 + }, + { + "epoch": 7.49540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18761 + }, + { + "epoch": 7.495805033959249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18762 + }, + { + "epoch": 7.496204554534558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18763 + }, + { + "epoch": 7.496604075109868, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18764 + }, + { + "epoch": 7.497003595685178, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18765 + }, + { + "epoch": 7.497403116260488, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18766 + }, + { + "epoch": 7.497802636835797, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18767 + }, + { + "epoch": 7.498202157411106, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18768 + }, + { + "epoch": 7.498601677986416, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18769 + }, + { + "epoch": 7.499001198561726, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18770 + }, + { + "epoch": 7.499400719137036, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18771 + }, + { + "epoch": 7.4998002397123456, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18772 + }, + { + "epoch": 7.5001997602876544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18773 + }, + { + "epoch": 7.500599280862964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18774 + }, + { + "epoch": 7.500998801438274, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18775 + }, + { + "epoch": 7.501398322013584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18776 + }, + { + "epoch": 7.501797842588894, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18777 + }, + { + "epoch": 7.502197363164203, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18778 + }, + { + "epoch": 7.502596883739512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18779 + }, + { + "epoch": 7.502996404314822, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18780 + }, + { + "epoch": 7.503395924890132, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18781 + }, + { + "epoch": 7.503795445465442, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18782 + }, + { + "epoch": 7.504194966040751, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18783 + }, + { + "epoch": 7.50459448661606, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18784 + }, + { + "epoch": 7.50499400719137, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18785 + }, + { + "epoch": 7.50539352776668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18786 + }, + { + "epoch": 7.50579304834199, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18787 + }, + { + "epoch": 7.5061925689172995, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18788 + }, + { + "epoch": 7.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18789 + }, + { + "epoch": 7.506991610067918, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18790 + }, + { + "epoch": 7.507391130643228, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18791 + }, + { + "epoch": 7.507790651218538, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18792 + }, + { + "epoch": 7.5081901717938475, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18793 + }, + { + "epoch": 7.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18794 + }, + { + "epoch": 7.508989212944467, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18795 + }, + { + "epoch": 7.509388733519776, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18796 + }, + { + "epoch": 7.509788254095086, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18797 + }, + { + "epoch": 7.5101877746703956, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18798 + }, + { + "epoch": 7.510587295245705, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18799 + }, + { + "epoch": 7.510986815821015, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18800 + }, + { + "epoch": 7.511386336396324, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18801 + }, + { + "epoch": 7.511785856971634, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18802 + }, + { + "epoch": 7.512185377546944, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18803 + }, + { + "epoch": 7.512584898122253, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18804 + }, + { + "epoch": 7.512984418697563, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 18805 + }, + { + "epoch": 7.513383939272872, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18806 + }, + { + "epoch": 7.513783459848182, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18807 + }, + { + "epoch": 7.514182980423492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18808 + }, + { + "epoch": 7.514582500998801, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18809 + }, + { + "epoch": 7.514982021574111, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18810 + }, + { + "epoch": 7.515381542149421, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18811 + }, + { + "epoch": 7.515781062724731, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18812 + }, + { + "epoch": 7.51618058330004, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18813 + }, + { + "epoch": 7.5165801038753495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18814 + }, + { + "epoch": 7.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18815 + }, + { + "epoch": 7.517379145025969, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18816 + }, + { + "epoch": 7.517778665601279, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18817 + }, + { + "epoch": 7.518178186176588, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18818 + }, + { + "epoch": 7.5185777067518975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18819 + }, + { + "epoch": 7.518977227327207, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18820 + }, + { + "epoch": 7.519376747902517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18821 + }, + { + "epoch": 7.519776268477827, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18822 + }, + { + "epoch": 7.520175789053136, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18823 + }, + { + "epoch": 7.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18824 + }, + { + "epoch": 7.520974830203755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18825 + }, + { + "epoch": 7.521374350779065, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18826 + }, + { + "epoch": 7.521773871354375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18827 + }, + { + "epoch": 7.522173391929685, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18828 + }, + { + "epoch": 7.5225729125049945, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18829 + }, + { + "epoch": 7.522972433080303, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18830 + }, + { + "epoch": 7.523371953655613, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18831 + }, + { + "epoch": 7.523771474230923, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18832 + }, + { + "epoch": 7.524170994806233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18833 + }, + { + "epoch": 7.5245705153815425, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18834 + }, + { + "epoch": 7.524970035956851, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18835 + }, + { + "epoch": 7.525369556532161, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18836 + }, + { + "epoch": 7.525769077107471, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18837 + }, + { + "epoch": 7.526168597682781, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18838 + }, + { + "epoch": 7.526568118258091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18839 + }, + { + "epoch": 7.5269676388333995, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18840 + }, + { + "epoch": 7.527367159408709, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18841 + }, + { + "epoch": 7.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18842 + }, + { + "epoch": 7.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18843 + }, + { + "epoch": 7.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18844 + }, + { + "epoch": 7.528965241709948, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18845 + }, + { + "epoch": 7.529364762285257, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18846 + }, + { + "epoch": 7.529764282860567, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 18847 + }, + { + "epoch": 7.530163803435877, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18848 + }, + { + "epoch": 7.530563324011187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18849 + }, + { + "epoch": 7.5309628445864965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18850 + }, + { + "epoch": 7.531362365161806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18851 + }, + { + "epoch": 7.531761885737115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18852 + }, + { + "epoch": 7.532161406312425, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18853 + }, + { + "epoch": 7.532560926887735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18854 + }, + { + "epoch": 7.5329604474630445, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18855 + }, + { + "epoch": 7.533359968038354, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18856 + }, + { + "epoch": 7.533759488613663, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18857 + }, + { + "epoch": 7.534159009188973, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18858 + }, + { + "epoch": 7.534558529764283, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18859 + }, + { + "epoch": 7.5349580503395925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18860 + }, + { + "epoch": 7.535357570914902, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18861 + }, + { + "epoch": 7.535757091490212, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18862 + }, + { + "epoch": 7.536156612065521, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18863 + }, + { + "epoch": 7.536556132640831, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18864 + }, + { + "epoch": 7.536955653216141, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18865 + }, + { + "epoch": 7.53735517379145, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18866 + }, + { + "epoch": 7.53775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18867 + }, + { + "epoch": 7.53815421494207, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18868 + }, + { + "epoch": 7.538553735517379, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18869 + }, + { + "epoch": 7.538953256092689, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18870 + }, + { + "epoch": 7.539352776667998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18871 + }, + { + "epoch": 7.539752297243308, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 18872 + }, + { + "epoch": 7.540151817818618, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18873 + }, + { + "epoch": 7.540551338393927, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18874 + }, + { + "epoch": 7.540950858969237, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18875 + }, + { + "epoch": 7.5413503795445465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18876 + }, + { + "epoch": 7.541749900119856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18877 + }, + { + "epoch": 7.542149420695166, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18878 + }, + { + "epoch": 7.542548941270476, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18879 + }, + { + "epoch": 7.542948461845785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18880 + }, + { + "epoch": 7.5433479824210945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18881 + }, + { + "epoch": 7.543747502996404, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18882 + }, + { + "epoch": 7.544147023571714, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18883 + }, + { + "epoch": 7.544546544147024, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18884 + }, + { + "epoch": 7.544946064722334, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18885 + }, + { + "epoch": 7.5453455852976425, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18886 + }, + { + "epoch": 7.545745105872952, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18887 + }, + { + "epoch": 7.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18888 + }, + { + "epoch": 7.546544147023572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18889 + }, + { + "epoch": 7.546943667598882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18890 + }, + { + "epoch": 7.547343188174191, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18891 + }, + { + "epoch": 7.5477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18892 + }, + { + "epoch": 7.54814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18893 + }, + { + "epoch": 7.54854174990012, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18894 + }, + { + "epoch": 7.54894127047543, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18895 + }, + { + "epoch": 7.549340791050739, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18896 + }, + { + "epoch": 7.549740311626048, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18897 + }, + { + "epoch": 7.550139832201358, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18898 + }, + { + "epoch": 7.550539352776668, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18899 + }, + { + "epoch": 7.550938873351978, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18900 + }, + { + "epoch": 7.551338393927288, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18901 + }, + { + "epoch": 7.551737914502597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18902 + }, + { + "epoch": 7.552137435077906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18903 + }, + { + "epoch": 7.552536955653216, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18904 + }, + { + "epoch": 7.552936476228526, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18905 + }, + { + "epoch": 7.553335996803836, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18906 + }, + { + "epoch": 7.553735517379145, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18907 + }, + { + "epoch": 7.554135037954454, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18908 + }, + { + "epoch": 7.554534558529764, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18909 + }, + { + "epoch": 7.554934079105074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18910 + }, + { + "epoch": 7.555333599680384, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18911 + }, + { + "epoch": 7.555733120255693, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18912 + }, + { + "epoch": 7.556132640831002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18913 + }, + { + "epoch": 7.556532161406312, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18914 + }, + { + "epoch": 7.556931681981622, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18915 + }, + { + "epoch": 7.557331202556932, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18916 + }, + { + "epoch": 7.5577307231322415, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18917 + }, + { + "epoch": 7.558130243707551, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18918 + }, + { + "epoch": 7.558529764282861, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18919 + }, + { + "epoch": 7.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18920 + }, + { + "epoch": 7.55932880543348, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18921 + }, + { + "epoch": 7.5597283260087895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18922 + }, + { + "epoch": 7.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18923 + }, + { + "epoch": 7.560527367159409, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18924 + }, + { + "epoch": 7.560926887734718, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18925 + }, + { + "epoch": 7.561326408310028, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18926 + }, + { + "epoch": 7.561725928885338, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18927 + }, + { + "epoch": 7.562125449460647, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18928 + }, + { + "epoch": 7.562524970035957, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18929 + }, + { + "epoch": 7.562924490611266, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 18930 + }, + { + "epoch": 7.563324011186576, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18931 + }, + { + "epoch": 7.563723531761886, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18932 + }, + { + "epoch": 7.564123052337195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18933 + }, + { + "epoch": 7.564522572912505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18934 + }, + { + "epoch": 7.564922093487815, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18935 + }, + { + "epoch": 7.565321614063124, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18936 + }, + { + "epoch": 7.565721134638434, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18937 + }, + { + "epoch": 7.566120655213743, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18938 + }, + { + "epoch": 7.566520175789053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18939 + }, + { + "epoch": 7.566919696364363, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18940 + }, + { + "epoch": 7.567319216939673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18941 + }, + { + "epoch": 7.567718737514982, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 18942 + }, + { + "epoch": 7.5681182580902915, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18943 + }, + { + "epoch": 7.568517778665601, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18944 + }, + { + "epoch": 7.568917299240911, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18945 + }, + { + "epoch": 7.569316819816221, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18946 + }, + { + "epoch": 7.56971634039153, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18947 + }, + { + "epoch": 7.5701158609668395, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18948 + }, + { + "epoch": 7.570515381542149, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18949 + }, + { + "epoch": 7.570914902117459, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 18950 + }, + { + "epoch": 7.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18951 + }, + { + "epoch": 7.571713943268079, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18952 + }, + { + "epoch": 7.572113463843388, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18953 + }, + { + "epoch": 7.572512984418697, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18954 + }, + { + "epoch": 7.572912504994007, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18955 + }, + { + "epoch": 7.573312025569317, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18956 + }, + { + "epoch": 7.573711546144627, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18957 + }, + { + "epoch": 7.5741110667199365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18958 + }, + { + "epoch": 7.574510587295245, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18959 + }, + { + "epoch": 7.574910107870555, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 18960 + }, + { + "epoch": 7.575309628445865, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18961 + }, + { + "epoch": 7.575709149021175, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18962 + }, + { + "epoch": 7.5761086695964845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18963 + }, + { + "epoch": 7.5765081901717934, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18964 + }, + { + "epoch": 7.576907710747103, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18965 + }, + { + "epoch": 7.577307231322413, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18966 + }, + { + "epoch": 7.577706751897723, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 18967 + }, + { + "epoch": 7.578106272473033, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18968 + }, + { + "epoch": 7.578505793048342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18969 + }, + { + "epoch": 7.578905313623651, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18970 + }, + { + "epoch": 7.579304834198961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18971 + }, + { + "epoch": 7.579704354774271, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 18972 + }, + { + "epoch": 7.580103875349581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18973 + }, + { + "epoch": 7.58050339592489, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18974 + }, + { + "epoch": 7.5809029165002, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18975 + }, + { + "epoch": 7.581302437075509, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18976 + }, + { + "epoch": 7.581701957650819, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18977 + }, + { + "epoch": 7.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18978 + }, + { + "epoch": 7.5825009988014385, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18979 + }, + { + "epoch": 7.582900519376748, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18980 + }, + { + "epoch": 7.583300039952057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 18981 + }, + { + "epoch": 7.583699560527367, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18982 + }, + { + "epoch": 7.584099081102677, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18983 + }, + { + "epoch": 7.5844986016779865, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18984 + }, + { + "epoch": 7.584898122253296, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18985 + }, + { + "epoch": 7.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18986 + }, + { + "epoch": 7.585697163403915, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18987 + }, + { + "epoch": 7.586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18988 + }, + { + "epoch": 7.5864962045545346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18989 + }, + { + "epoch": 7.586895725129844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18990 + }, + { + "epoch": 7.587295245705154, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18991 + }, + { + "epoch": 7.587694766280464, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18992 + }, + { + "epoch": 7.588094286855773, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18993 + }, + { + "epoch": 7.588493807431083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18994 + }, + { + "epoch": 7.588893328006392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18995 + }, + { + "epoch": 7.589292848581702, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18996 + }, + { + "epoch": 7.589692369157012, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18997 + }, + { + "epoch": 7.590091889732321, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18998 + }, + { + "epoch": 7.590491410307631, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18999 + }, + { + "epoch": 7.59089093088294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19000 + }, + { + "epoch": 7.59129045145825, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19001 + }, + { + "epoch": 7.59168997203356, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19002 + }, + { + "epoch": 7.592089492608869, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 19003 + }, + { + "epoch": 7.592489013184179, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19004 + }, + { + "epoch": 7.5928885337594885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19005 + }, + { + "epoch": 7.593288054334798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19006 + }, + { + "epoch": 7.593687574910108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19007 + }, + { + "epoch": 7.594087095485418, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19008 + }, + { + "epoch": 7.594486616060728, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19009 + }, + { + "epoch": 7.5948861366360365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19010 + }, + { + "epoch": 7.595285657211346, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19011 + }, + { + "epoch": 7.595685177786656, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19012 + }, + { + "epoch": 7.596084698361966, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19013 + }, + { + "epoch": 7.596484218937276, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 19014 + }, + { + "epoch": 7.5968837395125846, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19015 + }, + { + "epoch": 7.597283260087894, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19016 + }, + { + "epoch": 7.597682780663204, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19017 + }, + { + "epoch": 7.598082301238514, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19018 + }, + { + "epoch": 7.598481821813824, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19019 + }, + { + "epoch": 7.598881342389133, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19020 + }, + { + "epoch": 7.599280862964442, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19021 + }, + { + "epoch": 7.599680383539752, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19022 + }, + { + "epoch": 7.600079904115062, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19023 + }, + { + "epoch": 7.600479424690372, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19024 + }, + { + "epoch": 7.6008789452656815, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19025 + }, + { + "epoch": 7.60127846584099, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19026 + }, + { + "epoch": 7.6016779864163, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19027 + }, + { + "epoch": 7.60207750699161, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19028 + }, + { + "epoch": 7.60247702756692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19029 + }, + { + "epoch": 7.60287654814223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19030 + }, + { + "epoch": 7.603276068717539, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19031 + }, + { + "epoch": 7.603675589292848, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19032 + }, + { + "epoch": 7.604075109868158, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19033 + }, + { + "epoch": 7.604474630443468, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19034 + }, + { + "epoch": 7.604874151018778, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19035 + }, + { + "epoch": 7.605273671594087, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19036 + }, + { + "epoch": 7.605673192169396, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19037 + }, + { + "epoch": 7.606072712744706, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19038 + }, + { + "epoch": 7.606472233320016, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19039 + }, + { + "epoch": 7.606871753895326, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19040 + }, + { + "epoch": 7.6072712744706354, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19041 + }, + { + "epoch": 7.607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19042 + }, + { + "epoch": 7.608070315621254, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19043 + }, + { + "epoch": 7.608469836196564, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19044 + }, + { + "epoch": 7.608869356771874, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19045 + }, + { + "epoch": 7.6092688773471835, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19046 + }, + { + "epoch": 7.609668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19047 + }, + { + "epoch": 7.610067918497803, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19048 + }, + { + "epoch": 7.610467439073112, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19049 + }, + { + "epoch": 7.610866959648422, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19050 + }, + { + "epoch": 7.6112664802237315, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19051 + }, + { + "epoch": 7.611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 19052 + }, + { + "epoch": 7.612065521374351, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19053 + }, + { + "epoch": 7.61246504194966, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19054 + }, + { + "epoch": 7.61286456252497, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19055 + }, + { + "epoch": 7.61326408310028, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19056 + }, + { + "epoch": 7.613663603675589, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19057 + }, + { + "epoch": 7.614063124250899, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19058 + }, + { + "epoch": 7.614462644826209, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19059 + }, + { + "epoch": 7.614862165401518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19060 + }, + { + "epoch": 7.615261685976828, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19061 + }, + { + "epoch": 7.615661206552137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19062 + }, + { + "epoch": 7.616060727127447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19063 + }, + { + "epoch": 7.616460247702757, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19064 + }, + { + "epoch": 7.616859768278067, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19065 + }, + { + "epoch": 7.617259288853376, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19066 + }, + { + "epoch": 7.6176588094286855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19067 + }, + { + "epoch": 7.618058330003995, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19068 + }, + { + "epoch": 7.618457850579305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19069 + }, + { + "epoch": 7.618857371154615, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19070 + }, + { + "epoch": 7.619256891729924, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19071 + }, + { + "epoch": 7.6196564123052335, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19072 + }, + { + "epoch": 7.620055932880543, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19073 + }, + { + "epoch": 7.620455453455853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19074 + }, + { + "epoch": 7.620854974031163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19075 + }, + { + "epoch": 7.621254494606472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19076 + }, + { + "epoch": 7.6216540151817815, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19077 + }, + { + "epoch": 7.622053535757091, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19078 + }, + { + "epoch": 7.622453056332401, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19079 + }, + { + "epoch": 7.622852576907711, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19080 + }, + { + "epoch": 7.623252097483021, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19081 + }, + { + "epoch": 7.6236516180583305, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19082 + }, + { + "epoch": 7.624051138633639, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19083 + }, + { + "epoch": 7.624450659208949, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19084 + }, + { + "epoch": 7.624850179784259, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19085 + }, + { + "epoch": 7.625249700359569, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19086 + }, + { + "epoch": 7.6256492209348785, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19087 + }, + { + "epoch": 7.626048741510187, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19088 + }, + { + "epoch": 7.626448262085497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19089 + }, + { + "epoch": 7.626847782660807, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19090 + }, + { + "epoch": 7.627247303236117, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19091 + }, + { + "epoch": 7.627646823811427, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19092 + }, + { + "epoch": 7.6280463443867355, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19093 + }, + { + "epoch": 7.628445864962045, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19094 + }, + { + "epoch": 7.628845385537355, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19095 + }, + { + "epoch": 7.629244906112665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19096 + }, + { + "epoch": 7.629644426687975, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19097 + }, + { + "epoch": 7.630043947263284, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19098 + }, + { + "epoch": 7.630443467838594, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19099 + }, + { + "epoch": 7.630842988413903, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19100 + }, + { + "epoch": 7.631242508989213, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19101 + }, + { + "epoch": 7.631642029564523, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19102 + }, + { + "epoch": 7.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19103 + }, + { + "epoch": 7.632441070715142, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19104 + }, + { + "epoch": 7.632840591290451, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19105 + }, + { + "epoch": 7.633240111865761, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19106 + }, + { + "epoch": 7.633639632441071, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19107 + }, + { + "epoch": 7.6340391530163805, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19108 + }, + { + "epoch": 7.63443867359169, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19109 + }, + { + "epoch": 7.634838194166999, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19110 + }, + { + "epoch": 7.635237714742309, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19111 + }, + { + "epoch": 7.635637235317619, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19112 + }, + { + "epoch": 7.6360367558929285, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19113 + }, + { + "epoch": 7.636436276468238, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19114 + }, + { + "epoch": 7.636835797043548, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19115 + }, + { + "epoch": 7.637235317618857, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 19116 + }, + { + "epoch": 7.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19117 + }, + { + "epoch": 7.638034358769477, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19118 + }, + { + "epoch": 7.638433879344786, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19119 + }, + { + "epoch": 7.638833399920096, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19120 + }, + { + "epoch": 7.639232920495406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19121 + }, + { + "epoch": 7.639632441070715, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19122 + }, + { + "epoch": 7.640031961646025, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19123 + }, + { + "epoch": 7.640431482221334, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19124 + }, + { + "epoch": 7.640831002796644, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19125 + }, + { + "epoch": 7.641230523371954, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19126 + }, + { + "epoch": 7.641630043947263, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 19127 + }, + { + "epoch": 7.642029564522573, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19128 + }, + { + "epoch": 7.642429085097882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19129 + }, + { + "epoch": 7.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19130 + }, + { + "epoch": 7.643228126248502, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19131 + }, + { + "epoch": 7.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19132 + }, + { + "epoch": 7.644027167399121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 19133 + }, + { + "epoch": 7.6444266879744305, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19134 + }, + { + "epoch": 7.64482620854974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19135 + }, + { + "epoch": 7.64522572912505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19136 + }, + { + "epoch": 7.64562524970036, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19137 + }, + { + "epoch": 7.64602477027567, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19138 + }, + { + "epoch": 7.6464242908509785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19139 + }, + { + "epoch": 7.646823811426288, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 19140 + }, + { + "epoch": 7.647223332001598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19141 + }, + { + "epoch": 7.647622852576908, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19142 + }, + { + "epoch": 7.648022373152218, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 19143 + }, + { + "epoch": 7.648421893727527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19144 + }, + { + "epoch": 7.648821414302836, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19145 + }, + { + "epoch": 7.649220934878146, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19146 + }, + { + "epoch": 7.649620455453456, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19147 + }, + { + "epoch": 7.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19148 + }, + { + "epoch": 7.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19149 + }, + { + "epoch": 7.650819017179384, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 19150 + }, + { + "epoch": 7.651218537754694, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19151 + }, + { + "epoch": 7.651618058330004, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19152 + }, + { + "epoch": 7.652017578905314, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19153 + }, + { + "epoch": 7.6524170994806235, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19154 + }, + { + "epoch": 7.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19155 + }, + { + "epoch": 7.653216140631242, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19156 + }, + { + "epoch": 7.653615661206552, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19157 + }, + { + "epoch": 7.654015181781862, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19158 + }, + { + "epoch": 7.654414702357172, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19159 + }, + { + "epoch": 7.654814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19160 + }, + { + "epoch": 7.65521374350779, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19161 + }, + { + "epoch": 7.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19162 + }, + { + "epoch": 7.65601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 19163 + }, + { + "epoch": 7.65641230523372, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19164 + }, + { + "epoch": 7.656811825809029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19165 + }, + { + "epoch": 7.657211346384338, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19166 + }, + { + "epoch": 7.657610866959648, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 19167 + }, + { + "epoch": 7.658010387534958, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19168 + }, + { + "epoch": 7.658409908110268, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19169 + }, + { + "epoch": 7.6588094286855775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19170 + }, + { + "epoch": 7.659208949260887, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19171 + }, + { + "epoch": 7.659608469836197, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19172 + }, + { + "epoch": 7.660007990411506, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19173 + }, + { + "epoch": 7.660407510986816, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19174 + }, + { + "epoch": 7.6608070315621255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19175 + }, + { + "epoch": 7.661206552137435, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19176 + }, + { + "epoch": 7.661606072712745, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19177 + }, + { + "epoch": 7.662005593288054, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19178 + }, + { + "epoch": 7.662405113863364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19179 + }, + { + "epoch": 7.6628046344386735, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19180 + }, + { + "epoch": 7.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19181 + }, + { + "epoch": 7.663603675589293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19182 + }, + { + "epoch": 7.664003196164602, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19183 + }, + { + "epoch": 7.664402716739912, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19184 + }, + { + "epoch": 7.664802237315222, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 19185 + }, + { + "epoch": 7.665201757890531, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19186 + }, + { + "epoch": 7.665601278465841, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 19187 + }, + { + "epoch": 7.666000799041151, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19188 + }, + { + "epoch": 7.666400319616461, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19189 + }, + { + "epoch": 7.66679984019177, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19190 + }, + { + "epoch": 7.667199360767079, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19191 + }, + { + "epoch": 7.667598881342389, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19192 + }, + { + "epoch": 7.667998401917699, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19193 + }, + { + "epoch": 7.668397922493009, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19194 + }, + { + "epoch": 7.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19195 + }, + { + "epoch": 7.6691969636436275, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19196 + }, + { + "epoch": 7.669596484218937, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19197 + }, + { + "epoch": 7.669996004794247, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19198 + }, + { + "epoch": 7.670395525369557, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19199 + }, + { + "epoch": 7.670795045944866, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19200 + }, + { + "epoch": 7.6711945665201755, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19201 + }, + { + "epoch": 7.671594087095485, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19202 + }, + { + "epoch": 7.671993607670795, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19203 + }, + { + "epoch": 7.672393128246105, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 19204 + }, + { + "epoch": 7.672792648821415, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 19205 + }, + { + "epoch": 7.6731921693967235, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19206 + }, + { + "epoch": 7.673591689972033, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19207 + }, + { + "epoch": 7.673991210547343, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19208 + }, + { + "epoch": 7.674390731122653, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19209 + }, + { + "epoch": 7.674790251697963, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19210 + }, + { + "epoch": 7.6751897722732725, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19211 + }, + { + "epoch": 7.675589292848581, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19212 + }, + { + "epoch": 7.675988813423891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19213 + }, + { + "epoch": 7.676388333999201, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19214 + }, + { + "epoch": 7.676787854574511, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19215 + }, + { + "epoch": 7.6771873751498205, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19216 + }, + { + "epoch": 7.677586895725129, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 19217 + }, + { + "epoch": 7.677986416300439, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 19218 + }, + { + "epoch": 7.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19219 + }, + { + "epoch": 7.678785457451059, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19220 + }, + { + "epoch": 7.679184978026369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19221 + }, + { + "epoch": 7.679584498601678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19222 + }, + { + "epoch": 7.679984019176987, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19223 + }, + { + "epoch": 7.680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19224 + }, + { + "epoch": 7.680783060327607, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19225 + }, + { + "epoch": 7.681182580902917, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19226 + }, + { + "epoch": 7.681582101478226, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19227 + }, + { + "epoch": 7.681981622053536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19228 + }, + { + "epoch": 7.682381142628845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19229 + }, + { + "epoch": 7.682780663204155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19230 + }, + { + "epoch": 7.683180183779465, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19231 + }, + { + "epoch": 7.683579704354774, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19232 + }, + { + "epoch": 7.683979224930084, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19233 + }, + { + "epoch": 7.684378745505393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19234 + }, + { + "epoch": 7.684778266080703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19235 + }, + { + "epoch": 7.685177786656013, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19236 + }, + { + "epoch": 7.6855773072313225, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19237 + }, + { + "epoch": 7.685976827806632, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19238 + }, + { + "epoch": 7.686376348381942, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19239 + }, + { + "epoch": 7.686775868957251, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19240 + }, + { + "epoch": 7.687175389532561, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19241 + }, + { + "epoch": 7.6875749101078705, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19242 + }, + { + "epoch": 7.68797443068318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19243 + }, + { + "epoch": 7.68837395125849, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19244 + }, + { + "epoch": 7.6887734718338, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19245 + }, + { + "epoch": 7.689172992409109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19246 + }, + { + "epoch": 7.689572512984419, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19247 + }, + { + "epoch": 7.689972033559728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 19248 + }, + { + "epoch": 7.690371554135038, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19249 + }, + { + "epoch": 7.690771074710348, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19250 + } + ], + "logging_steps": 1.0, + "max_steps": 250300, + "num_input_tokens_seen": 0, + "num_train_epochs": 100, + "save_steps": 250, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.8695666066820956e+18, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-19250/training_args.bin b/checkpoint-19250/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2bc6e8c6a7b45b79e23a7e4b282dcb7f0e746bc4 --- /dev/null +++ b/checkpoint-19250/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9307aa53030dab795f2b0f5373487f7b4270f37adff792b3db36f71bcb8dae52 +size 5816 diff --git a/checkpoint-19500/config.json b/checkpoint-19500/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fe7dbf7d28e62328ecc2df097527bd51ec161f9d --- /dev/null +++ b/checkpoint-19500/config.json @@ -0,0 +1,73 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 6, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 512, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 64, + "intermediate_size": 8960, + "layerwise_start_idx": 21, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 7, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f128c512-mix-1.0-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-19500/generation_config.json b/checkpoint-19500/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-19500/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-19500/model-00001-of-00002.safetensors b/checkpoint-19500/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d4dfa0e271ad6b1218b0b6dda4436f6cc4136fca --- /dev/null +++ b/checkpoint-19500/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d27d5d92f7450f116e4ffe3c18c20b36cb3179a3dea6ecdb427983cd651d0659 +size 4754815448 diff --git a/checkpoint-19500/model-00002-of-00002.safetensors b/checkpoint-19500/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c76cf00e7a9c7ce6c6d8a7a347afcbf18b661e2d --- /dev/null +++ b/checkpoint-19500/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d096d5260908edb86a73ee04889ac23d52b4054c4adabbc4a0e6ec76854d6f +size 840447856 diff --git a/checkpoint-19500/model.safetensors.index.json b/checkpoint-19500/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea3b7f98260439a7c863a0290604b077271f337 --- /dev/null +++ b/checkpoint-19500/model.safetensors.index.json @@ -0,0 +1,1076 @@ +{ + "metadata": { + "total_size": 5595140920 + }, + "weight_map": { + "embed_tokens.weight": "model-00001-of-00002.safetensors", + "img2text.bias": "model-00001-of-00002.safetensors", + "img2text.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.weight": "model-00001-of-00002.safetensors", + "layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.gate": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.22.gate": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.23.gate": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.24.gate": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.25.gate": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.26.gate": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.27.gate": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "norm.weight": "model-00001-of-00002.safetensors", + "patch_embedder.proj.bias": "model-00001-of-00002.safetensors", + "patch_embedder.proj.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.bias": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_out.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.conv_out.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.encoder.conv_in.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_in.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-19500/optimizer.pt b/checkpoint-19500/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..ba7e2fae99c776709526a034688ddfebae90e87f --- /dev/null +++ b/checkpoint-19500/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c82f60fbfe305de0acf10555f18b272eeee35810f070687ab9ef05ffd1ea3de0 +size 5620161007 diff --git a/checkpoint-19500/rng_state_0.pth b/checkpoint-19500/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..238c015acc73bbb43326968c5e5cd05d5b2587ac --- /dev/null +++ b/checkpoint-19500/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:11a19bf2a3df25c0089a8ac351e368bdd3e41ad230b3e45d8e795f77ea49b106 +size 14960 diff --git a/checkpoint-19500/rng_state_1.pth b/checkpoint-19500/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..47ce6293ba17d8789b64af91c45b978203399634 --- /dev/null +++ b/checkpoint-19500/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ba1e3f4ffecd06a9331d84faa16ff499d9df136237a9b2b1d330766b5d2227b2 +size 14960 diff --git a/checkpoint-19500/rng_state_2.pth b/checkpoint-19500/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..a168e18b97965a4d6514aa26701dc4507b7cee72 --- /dev/null +++ b/checkpoint-19500/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1d2bc7e47aa897e2249313f6bf28ae4cc541d22886eb79959f25339c08787656 +size 14960 diff --git a/checkpoint-19500/rng_state_3.pth b/checkpoint-19500/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..c476e57e14cf1bfa656f9374db6756982438ad1d --- /dev/null +++ b/checkpoint-19500/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:48fb596880b020a14093ba097b6bcf8cb73f8dede0e186418c9ae2c8bb996e1b +size 14960 diff --git a/checkpoint-19500/scheduler.pt b/checkpoint-19500/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..283564df3b3be05333af71a77a163fdd351906dc --- /dev/null +++ b/checkpoint-19500/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4afa35f4b32b417eec751a4bb43398efb94f15390672a3160295126c10272c6b +size 1064 diff --git a/checkpoint-19500/trainer_state.json b/checkpoint-19500/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..ab0cedf0492576c16e6ac5ba800f378779610f3c --- /dev/null +++ b/checkpoint-19500/trainer_state.json @@ -0,0 +1,136533 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.7906512185377546, + "eval_steps": 500, + "global_step": 19500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00039952057530962844, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 3.1824, + "step": 1 + }, + { + "epoch": 0.0007990411506192569, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.9778, + "step": 2 + }, + { + "epoch": 0.0011985617259288853, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 2.7824, + "step": 3 + }, + { + "epoch": 0.0015980823012385138, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 2.6074, + "step": 4 + }, + { + "epoch": 0.001997602876548142, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.4209, + "step": 5 + }, + { + "epoch": 0.0023971234518577705, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 2.2482, + "step": 6 + }, + { + "epoch": 0.002796644027167399, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 2.1017, + "step": 7 + }, + { + "epoch": 0.0031961646024770275, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 2.0034, + "step": 8 + }, + { + "epoch": 0.003595685177786656, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.9567, + "step": 9 + }, + { + "epoch": 0.003995205753096284, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.9417, + "step": 10 + }, + { + "epoch": 0.0043947263284059125, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.9526, + "step": 11 + }, + { + "epoch": 0.004794246903715541, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.9613, + "step": 12 + }, + { + "epoch": 0.0051937674790251695, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.9467, + "step": 13 + }, + { + "epoch": 0.005593288054334798, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.9215, + "step": 14 + }, + { + "epoch": 0.0059928086296444265, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.9154, + "step": 15 + }, + { + "epoch": 0.006392329204954055, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.9064, + "step": 16 + }, + { + "epoch": 0.0067918497802636835, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.888, + "step": 17 + }, + { + "epoch": 0.007191370355573312, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.8729, + "step": 18 + }, + { + "epoch": 0.0075908909308829405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8425, + "step": 19 + }, + { + "epoch": 0.007990411506192568, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.8468, + "step": 20 + }, + { + "epoch": 0.008389932081502197, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.8418, + "step": 21 + }, + { + "epoch": 0.008789452656811825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.825, + "step": 22 + }, + { + "epoch": 0.009188973232121454, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8176, + "step": 23 + }, + { + "epoch": 0.009588493807431082, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.7827, + "step": 24 + }, + { + "epoch": 0.00998801438274071, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.7695, + "step": 25 + }, + { + "epoch": 0.010387534958050339, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.78, + "step": 26 + }, + { + "epoch": 0.010787055533359968, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.7588, + "step": 27 + }, + { + "epoch": 0.011186576108669596, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.743, + "step": 28 + }, + { + "epoch": 0.011586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.7357, + "step": 29 + }, + { + "epoch": 0.011985617259288853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.7271, + "step": 30 + }, + { + "epoch": 0.012385137834598482, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.7053, + "step": 31 + }, + { + "epoch": 0.01278465840990811, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 32 + }, + { + "epoch": 0.013184178985217739, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.6831, + "step": 33 + }, + { + "epoch": 0.013583699560527367, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 34 + }, + { + "epoch": 0.013983220135836995, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6682, + "step": 35 + }, + { + "epoch": 0.014382740711146624, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6573, + "step": 36 + }, + { + "epoch": 0.014782261286456252, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.6507, + "step": 37 + }, + { + "epoch": 0.015181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.6424, + "step": 38 + }, + { + "epoch": 0.01558130243707551, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.6251, + "step": 39 + }, + { + "epoch": 0.015980823012385136, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.6215, + "step": 40 + }, + { + "epoch": 0.016380343587694766, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.6226, + "step": 41 + }, + { + "epoch": 0.016779864163004393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.6005, + "step": 42 + }, + { + "epoch": 0.017179384738314023, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.6224, + "step": 43 + }, + { + "epoch": 0.01757890531362365, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5901, + "step": 44 + }, + { + "epoch": 0.01797842588893328, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5962, + "step": 45 + }, + { + "epoch": 0.018377946464242907, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 46 + }, + { + "epoch": 0.018777467039552537, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5697, + "step": 47 + }, + { + "epoch": 0.019176987614862164, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 48 + }, + { + "epoch": 0.019576508190171794, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.5531, + "step": 49 + }, + { + "epoch": 0.01997602876548142, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.554, + "step": 50 + }, + { + "epoch": 0.02037554934079105, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.5539, + "step": 51 + }, + { + "epoch": 0.020775069916100678, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.5591, + "step": 52 + }, + { + "epoch": 0.02117459049141031, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.545, + "step": 53 + }, + { + "epoch": 0.021574111066719935, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5221, + "step": 54 + }, + { + "epoch": 0.021973631642029565, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.5234, + "step": 55 + }, + { + "epoch": 0.022373152217339192, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.5186, + "step": 56 + }, + { + "epoch": 0.022772672792648822, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.5273, + "step": 57 + }, + { + "epoch": 0.02317219336795845, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5201, + "step": 58 + }, + { + "epoch": 0.02357171394326808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5108, + "step": 59 + }, + { + "epoch": 0.023971234518577706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.5078, + "step": 60 + }, + { + "epoch": 0.024370755093887336, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.5026, + "step": 61 + }, + { + "epoch": 0.024770275669196963, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.5076, + "step": 62 + }, + { + "epoch": 0.025169796244506593, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4956, + "step": 63 + }, + { + "epoch": 0.02556931681981622, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4925, + "step": 64 + }, + { + "epoch": 0.02596883739512585, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.5045, + "step": 65 + }, + { + "epoch": 0.026368357970435477, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.4843, + "step": 66 + }, + { + "epoch": 0.026767878545745107, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4807, + "step": 67 + }, + { + "epoch": 0.027167399121054734, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 68 + }, + { + "epoch": 0.027566919696364364, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4784, + "step": 69 + }, + { + "epoch": 0.02796644027167399, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4758, + "step": 70 + }, + { + "epoch": 0.02836596084698362, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4871, + "step": 71 + }, + { + "epoch": 0.028765481422293248, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4753, + "step": 72 + }, + { + "epoch": 0.029165001997602878, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4922, + "step": 73 + }, + { + "epoch": 0.029564522572912505, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4759, + "step": 74 + }, + { + "epoch": 0.02996404314822213, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.4703, + "step": 75 + }, + { + "epoch": 0.030363563723531762, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4666, + "step": 76 + }, + { + "epoch": 0.03076308429884139, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4628, + "step": 77 + }, + { + "epoch": 0.03116260487415102, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.4656, + "step": 78 + }, + { + "epoch": 0.031562125449460646, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4695, + "step": 79 + }, + { + "epoch": 0.03196164602477027, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4517, + "step": 80 + }, + { + "epoch": 0.032361166600079906, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4577, + "step": 81 + }, + { + "epoch": 0.03276068717538953, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4526, + "step": 82 + }, + { + "epoch": 0.03316020775069916, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4476, + "step": 83 + }, + { + "epoch": 0.033559728326008786, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.4499, + "step": 84 + }, + { + "epoch": 0.03395924890131842, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.4549, + "step": 85 + }, + { + "epoch": 0.03435876947662805, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4548, + "step": 86 + }, + { + "epoch": 0.034758290051937674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4525, + "step": 87 + }, + { + "epoch": 0.0351578106272473, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 88 + }, + { + "epoch": 0.035557331202556934, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4343, + "step": 89 + }, + { + "epoch": 0.03595685177786656, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4398, + "step": 90 + }, + { + "epoch": 0.03635637235317619, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4431, + "step": 91 + }, + { + "epoch": 0.036755892928485814, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.438, + "step": 92 + }, + { + "epoch": 0.03715541350379545, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4291, + "step": 93 + }, + { + "epoch": 0.037554934079105075, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.4373, + "step": 94 + }, + { + "epoch": 0.0379544546544147, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4409, + "step": 95 + }, + { + "epoch": 0.03835397522972433, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4305, + "step": 96 + }, + { + "epoch": 0.03875349580503396, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.4337, + "step": 97 + }, + { + "epoch": 0.03915301638034359, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4309, + "step": 98 + }, + { + "epoch": 0.039552536955653216, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 99 + }, + { + "epoch": 0.03995205753096284, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 100 + }, + { + "epoch": 0.040351578106272476, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.42, + "step": 101 + }, + { + "epoch": 0.0407510986815821, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4211, + "step": 102 + }, + { + "epoch": 0.04115061925689173, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4342, + "step": 103 + }, + { + "epoch": 0.041550139832201356, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4191, + "step": 104 + }, + { + "epoch": 0.04194966040751099, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.4234, + "step": 105 + }, + { + "epoch": 0.04234918098282062, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 106 + }, + { + "epoch": 0.042748701558130243, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4308, + "step": 107 + }, + { + "epoch": 0.04314822213343987, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.4189, + "step": 108 + }, + { + "epoch": 0.043547742708749504, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 109 + }, + { + "epoch": 0.04394726328405913, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4278, + "step": 110 + }, + { + "epoch": 0.04434678385936876, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 111 + }, + { + "epoch": 0.044746304434678384, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.4276, + "step": 112 + }, + { + "epoch": 0.04514582500998801, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.4203, + "step": 113 + }, + { + "epoch": 0.045545345585297645, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4313, + "step": 114 + }, + { + "epoch": 0.04594486616060727, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4186, + "step": 115 + }, + { + "epoch": 0.0463443867359169, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4215, + "step": 116 + }, + { + "epoch": 0.046743907311226525, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.4272, + "step": 117 + }, + { + "epoch": 0.04714342788653616, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4262, + "step": 118 + }, + { + "epoch": 0.047542948461845785, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 119 + }, + { + "epoch": 0.04794246903715541, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4071, + "step": 120 + }, + { + "epoch": 0.04834198961246504, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4241, + "step": 121 + }, + { + "epoch": 0.04874151018777467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 122 + }, + { + "epoch": 0.0491410307630843, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4252, + "step": 123 + }, + { + "epoch": 0.049540551338393926, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.424, + "step": 124 + }, + { + "epoch": 0.04994007191370355, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 125 + }, + { + "epoch": 0.05033959248901319, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.412, + "step": 126 + }, + { + "epoch": 0.05073911306432281, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4199, + "step": 127 + }, + { + "epoch": 0.05113863363963244, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 128 + }, + { + "epoch": 0.05153815421494207, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 129 + }, + { + "epoch": 0.0519376747902517, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 130 + }, + { + "epoch": 0.05233719536556133, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4049, + "step": 131 + }, + { + "epoch": 0.052736715940870954, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4134, + "step": 132 + }, + { + "epoch": 0.05313623651618058, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 133 + }, + { + "epoch": 0.053535757091490214, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4156, + "step": 134 + }, + { + "epoch": 0.05393527766679984, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 135 + }, + { + "epoch": 0.05433479824210947, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 136 + }, + { + "epoch": 0.054734318817419095, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 137 + }, + { + "epoch": 0.05513383939272873, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 138 + }, + { + "epoch": 0.055533359968038355, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.428, + "step": 139 + }, + { + "epoch": 0.05593288054334798, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 140 + }, + { + "epoch": 0.05633240111865761, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.4083, + "step": 141 + }, + { + "epoch": 0.05673192169396724, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 142 + }, + { + "epoch": 0.05713144226927687, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 143 + }, + { + "epoch": 0.057530962844586496, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 144 + }, + { + "epoch": 0.05793048341989612, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 145 + }, + { + "epoch": 0.058330003995205756, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 146 + }, + { + "epoch": 0.05872952457051538, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.4168, + "step": 147 + }, + { + "epoch": 0.05912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 148 + }, + { + "epoch": 0.05952856572113464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 149 + }, + { + "epoch": 0.05992808629644426, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 150 + }, + { + "epoch": 0.0603276068717539, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4171, + "step": 151 + }, + { + "epoch": 0.060727127447063524, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 152 + }, + { + "epoch": 0.06112664802237315, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 153 + }, + { + "epoch": 0.06152616859768278, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3966, + "step": 154 + }, + { + "epoch": 0.06192568917299241, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.4121, + "step": 155 + }, + { + "epoch": 0.06232520974830204, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 156 + }, + { + "epoch": 0.06272473032361167, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 157 + }, + { + "epoch": 0.06312425089892129, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 158 + }, + { + "epoch": 0.06352377147423093, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.413, + "step": 159 + }, + { + "epoch": 0.06392329204954054, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4174, + "step": 160 + }, + { + "epoch": 0.06432281262485018, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4133, + "step": 161 + }, + { + "epoch": 0.06472233320015981, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 162 + }, + { + "epoch": 0.06512185377546943, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4128, + "step": 163 + }, + { + "epoch": 0.06552137435077907, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3945, + "step": 164 + }, + { + "epoch": 0.0659208949260887, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 165 + }, + { + "epoch": 0.06632041550139832, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 166 + }, + { + "epoch": 0.06671993607670795, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4104, + "step": 167 + }, + { + "epoch": 0.06711945665201757, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.4144, + "step": 168 + }, + { + "epoch": 0.0675189772273272, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4059, + "step": 169 + }, + { + "epoch": 0.06791849780263684, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 170 + }, + { + "epoch": 0.06831801837794646, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 171 + }, + { + "epoch": 0.0687175389532561, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 172 + }, + { + "epoch": 0.06911705952856573, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 173 + }, + { + "epoch": 0.06951658010387535, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 174 + }, + { + "epoch": 0.06991610067918498, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4097, + "step": 175 + }, + { + "epoch": 0.0703156212544946, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 176 + }, + { + "epoch": 0.07071514182980423, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4052, + "step": 177 + }, + { + "epoch": 0.07111466240511387, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.417, + "step": 178 + }, + { + "epoch": 0.07151418298042349, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 179 + }, + { + "epoch": 0.07191370355573312, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 180 + }, + { + "epoch": 0.07231322413104276, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 181 + }, + { + "epoch": 0.07271274470635238, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 182 + }, + { + "epoch": 0.07311226528166201, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 183 + }, + { + "epoch": 0.07351178585697163, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 184 + }, + { + "epoch": 0.07391130643228126, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 185 + }, + { + "epoch": 0.0743108270075909, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 186 + }, + { + "epoch": 0.07471034758290052, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 187 + }, + { + "epoch": 0.07510986815821015, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 188 + }, + { + "epoch": 0.07550938873351977, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.4132, + "step": 189 + }, + { + "epoch": 0.0759089093088294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.4067, + "step": 190 + }, + { + "epoch": 0.07630842988413904, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 191 + }, + { + "epoch": 0.07670795045944866, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 192 + }, + { + "epoch": 0.07710747103475829, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3964, + "step": 193 + }, + { + "epoch": 0.07750699161006792, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 194 + }, + { + "epoch": 0.07790651218537754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3979, + "step": 195 + }, + { + "epoch": 0.07830603276068718, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 196 + }, + { + "epoch": 0.0787055533359968, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4111, + "step": 197 + }, + { + "epoch": 0.07910507391130643, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 198 + }, + { + "epoch": 0.07950459448661606, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 199 + }, + { + "epoch": 0.07990411506192568, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 200 + }, + { + "epoch": 0.08030363563723532, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 201 + }, + { + "epoch": 0.08070315621254495, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 202 + }, + { + "epoch": 0.08110267678785457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4021, + "step": 203 + }, + { + "epoch": 0.0815021973631642, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 204 + }, + { + "epoch": 0.08190171793847383, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 205 + }, + { + "epoch": 0.08230123851378346, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 206 + }, + { + "epoch": 0.08270075908909309, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.4057, + "step": 207 + }, + { + "epoch": 0.08310027966440271, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4158, + "step": 208 + }, + { + "epoch": 0.08349980023971235, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3898, + "step": 209 + }, + { + "epoch": 0.08389932081502198, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 210 + }, + { + "epoch": 0.0842988413903316, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.4193, + "step": 211 + }, + { + "epoch": 0.08469836196564123, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 212 + }, + { + "epoch": 0.08509788254095085, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 213 + }, + { + "epoch": 0.08549740311626049, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 214 + }, + { + "epoch": 0.08589692369157012, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4098, + "step": 215 + }, + { + "epoch": 0.08629644426687974, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.4015, + "step": 216 + }, + { + "epoch": 0.08669596484218937, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.4113, + "step": 217 + }, + { + "epoch": 0.08709548541749901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 218 + }, + { + "epoch": 0.08749500599280863, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 219 + }, + { + "epoch": 0.08789452656811826, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 220 + }, + { + "epoch": 0.08829404714342788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 221 + }, + { + "epoch": 0.08869356771873751, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.409, + "step": 222 + }, + { + "epoch": 0.08909308829404715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 223 + }, + { + "epoch": 0.08949260886935677, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 224 + }, + { + "epoch": 0.0898921294446664, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 225 + }, + { + "epoch": 0.09029165001997602, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 226 + }, + { + "epoch": 0.09069117059528566, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3904, + "step": 227 + }, + { + "epoch": 0.09109069117059529, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 228 + }, + { + "epoch": 0.09149021174590491, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 229 + }, + { + "epoch": 0.09188973232121454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 230 + }, + { + "epoch": 0.09228925289652418, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 231 + }, + { + "epoch": 0.0926887734718338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 232 + }, + { + "epoch": 0.09308829404714343, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 233 + }, + { + "epoch": 0.09348781462245305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 234 + }, + { + "epoch": 0.09388733519776268, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 235 + }, + { + "epoch": 0.09428685577307232, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 236 + }, + { + "epoch": 0.09468637634838194, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 237 + }, + { + "epoch": 0.09508589692369157, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 238 + }, + { + "epoch": 0.0954854174990012, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 239 + }, + { + "epoch": 0.09588493807431082, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4064, + "step": 240 + }, + { + "epoch": 0.09628445864962046, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 241 + }, + { + "epoch": 0.09668397922493008, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 242 + }, + { + "epoch": 0.09708349980023971, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 243 + }, + { + "epoch": 0.09748302037554935, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 244 + }, + { + "epoch": 0.09788254095085897, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3961, + "step": 245 + }, + { + "epoch": 0.0982820615261686, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4061, + "step": 246 + }, + { + "epoch": 0.09868158210147823, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 247 + }, + { + "epoch": 0.09908110267678785, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 248 + }, + { + "epoch": 0.09948062325209749, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 249 + }, + { + "epoch": 0.0998801438274071, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 250 + }, + { + "epoch": 0.10027966440271674, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 251 + }, + { + "epoch": 0.10067918497802637, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 252 + }, + { + "epoch": 0.10107870555333599, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.4024, + "step": 253 + }, + { + "epoch": 0.10147822612864563, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 254 + }, + { + "epoch": 0.10187774670395526, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 255 + }, + { + "epoch": 0.10227726727926488, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 256 + }, + { + "epoch": 0.10267678785457451, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 257 + }, + { + "epoch": 0.10307630842988413, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 258 + }, + { + "epoch": 0.10347582900519377, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 259 + }, + { + "epoch": 0.1038753495805034, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 260 + }, + { + "epoch": 0.10427487015581302, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 261 + }, + { + "epoch": 0.10467439073112265, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 262 + }, + { + "epoch": 0.10507391130643227, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 263 + }, + { + "epoch": 0.10547343188174191, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 264 + }, + { + "epoch": 0.10587295245705154, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3988, + "step": 265 + }, + { + "epoch": 0.10627247303236116, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 266 + }, + { + "epoch": 0.1066719936076708, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.4079, + "step": 267 + }, + { + "epoch": 0.10707151418298043, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 268 + }, + { + "epoch": 0.10747103475829005, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3954, + "step": 269 + }, + { + "epoch": 0.10787055533359968, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 270 + }, + { + "epoch": 0.1082700759089093, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 271 + }, + { + "epoch": 0.10866959648421894, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 272 + }, + { + "epoch": 0.10906911705952857, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 273 + }, + { + "epoch": 0.10946863763483819, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3899, + "step": 274 + }, + { + "epoch": 0.10986815821014782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 275 + }, + { + "epoch": 0.11026767878545746, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4012, + "step": 276 + }, + { + "epoch": 0.11066719936076708, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 277 + }, + { + "epoch": 0.11106671993607671, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3948, + "step": 278 + }, + { + "epoch": 0.11146624051138633, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4175, + "step": 279 + }, + { + "epoch": 0.11186576108669596, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 280 + }, + { + "epoch": 0.1122652816620056, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 281 + }, + { + "epoch": 0.11266480223731522, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3994, + "step": 282 + }, + { + "epoch": 0.11306432281262485, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 283 + }, + { + "epoch": 0.11346384338793448, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 284 + }, + { + "epoch": 0.1138633639632441, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 285 + }, + { + "epoch": 0.11426288453855374, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 286 + }, + { + "epoch": 0.11466240511386336, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4045, + "step": 287 + }, + { + "epoch": 0.11506192568917299, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 288 + }, + { + "epoch": 0.11546144626448263, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 289 + }, + { + "epoch": 0.11586096683979225, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 290 + }, + { + "epoch": 0.11626048741510188, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 291 + }, + { + "epoch": 0.11666000799041151, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 292 + }, + { + "epoch": 0.11705952856572113, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 293 + }, + { + "epoch": 0.11745904914103077, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3991, + "step": 294 + }, + { + "epoch": 0.11785856971634039, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 295 + }, + { + "epoch": 0.11825809029165002, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 296 + }, + { + "epoch": 0.11865761086695965, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 297 + }, + { + "epoch": 0.11905713144226927, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 298 + }, + { + "epoch": 0.11945665201757891, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 299 + }, + { + "epoch": 0.11985617259288853, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 300 + }, + { + "epoch": 0.12025569316819816, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3998, + "step": 301 + }, + { + "epoch": 0.1206552137435078, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 302 + }, + { + "epoch": 0.12105473431881741, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 303 + }, + { + "epoch": 0.12145425489412705, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 304 + }, + { + "epoch": 0.12185377546943668, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 305 + }, + { + "epoch": 0.1222532960447463, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 306 + }, + { + "epoch": 0.12265281662005593, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 307 + }, + { + "epoch": 0.12305233719536555, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 308 + }, + { + "epoch": 0.12345185777067519, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 309 + }, + { + "epoch": 0.12385137834598482, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 310 + }, + { + "epoch": 0.12425089892129444, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 311 + }, + { + "epoch": 0.12465041949660408, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 312 + }, + { + "epoch": 0.1250499400719137, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 313 + }, + { + "epoch": 0.12544946064722334, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 314 + }, + { + "epoch": 0.12584898122253296, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 315 + }, + { + "epoch": 0.12624850179784258, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 316 + }, + { + "epoch": 0.12664802237315223, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 317 + }, + { + "epoch": 0.12704754294846185, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 318 + }, + { + "epoch": 0.12744706352377147, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 319 + }, + { + "epoch": 0.1278465840990811, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 320 + }, + { + "epoch": 0.12824610467439074, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 321 + }, + { + "epoch": 0.12864562524970036, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3987, + "step": 322 + }, + { + "epoch": 0.12904514582500998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 323 + }, + { + "epoch": 0.12944466640031962, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 324 + }, + { + "epoch": 0.12984418697562924, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 325 + }, + { + "epoch": 0.13024370755093886, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 326 + }, + { + "epoch": 0.1306432281262485, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 327 + }, + { + "epoch": 0.13104274870155813, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 328 + }, + { + "epoch": 0.13144226927686775, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 329 + }, + { + "epoch": 0.1318417898521774, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 330 + }, + { + "epoch": 0.13224131042748702, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 331 + }, + { + "epoch": 0.13264083100279664, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 332 + }, + { + "epoch": 0.1330403515781063, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 333 + }, + { + "epoch": 0.1334398721534159, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 334 + }, + { + "epoch": 0.13383939272872553, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 335 + }, + { + "epoch": 0.13423891330403515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 336 + }, + { + "epoch": 0.1346384338793448, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 337 + }, + { + "epoch": 0.1350379544546544, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 338 + }, + { + "epoch": 0.13543747502996403, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 339 + }, + { + "epoch": 0.13583699560527368, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 340 + }, + { + "epoch": 0.1362365161805833, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 341 + }, + { + "epoch": 0.13663603675589292, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 342 + }, + { + "epoch": 0.13703555733120257, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3889, + "step": 343 + }, + { + "epoch": 0.1374350779065122, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3995, + "step": 344 + }, + { + "epoch": 0.1378345984818218, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 345 + }, + { + "epoch": 0.13823411905713145, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 346 + }, + { + "epoch": 0.13863363963244107, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3983, + "step": 347 + }, + { + "epoch": 0.1390331602077507, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 348 + }, + { + "epoch": 0.13943268078306031, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 349 + }, + { + "epoch": 0.13983220135836996, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 350 + }, + { + "epoch": 0.14023172193367958, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 351 + }, + { + "epoch": 0.1406312425089892, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 352 + }, + { + "epoch": 0.14103076308429885, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 353 + }, + { + "epoch": 0.14143028365960847, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.391, + "step": 354 + }, + { + "epoch": 0.1418298042349181, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3937, + "step": 355 + }, + { + "epoch": 0.14222932481022774, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3858, + "step": 356 + }, + { + "epoch": 0.14262884538553736, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 357 + }, + { + "epoch": 0.14302836596084698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 358 + }, + { + "epoch": 0.14342788653615662, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 359 + }, + { + "epoch": 0.14382740711146624, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 360 + }, + { + "epoch": 0.14422692768677586, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 361 + }, + { + "epoch": 0.1446264482620855, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 362 + }, + { + "epoch": 0.14502596883739513, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 363 + }, + { + "epoch": 0.14542548941270475, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 364 + }, + { + "epoch": 0.14582500998801437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 365 + }, + { + "epoch": 0.14622453056332402, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 366 + }, + { + "epoch": 0.14662405113863364, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 367 + }, + { + "epoch": 0.14702357171394326, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 368 + }, + { + "epoch": 0.1474230922892529, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 369 + }, + { + "epoch": 0.14782261286456252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 370 + }, + { + "epoch": 0.14822213343987214, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 371 + }, + { + "epoch": 0.1486216540151818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 372 + }, + { + "epoch": 0.1490211745904914, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 373 + }, + { + "epoch": 0.14942069516580103, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3942, + "step": 374 + }, + { + "epoch": 0.14982021574111068, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 375 + }, + { + "epoch": 0.1502197363164203, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 376 + }, + { + "epoch": 0.15061925689172992, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3903, + "step": 377 + }, + { + "epoch": 0.15101877746703954, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 378 + }, + { + "epoch": 0.1514182980423492, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 379 + }, + { + "epoch": 0.1518178186176588, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 380 + }, + { + "epoch": 0.15221733919296843, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 381 + }, + { + "epoch": 0.15261685976827807, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 382 + }, + { + "epoch": 0.1530163803435877, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 383 + }, + { + "epoch": 0.1534159009188973, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 384 + }, + { + "epoch": 0.15381542149420696, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 385 + }, + { + "epoch": 0.15421494206951658, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 386 + }, + { + "epoch": 0.1546144626448262, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 387 + }, + { + "epoch": 0.15501398322013585, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 388 + }, + { + "epoch": 0.15541350379544547, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 389 + }, + { + "epoch": 0.1558130243707551, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 390 + }, + { + "epoch": 0.15621254494606474, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 391 + }, + { + "epoch": 0.15661206552137436, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 392 + }, + { + "epoch": 0.15701158609668397, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 393 + }, + { + "epoch": 0.1574111066719936, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 394 + }, + { + "epoch": 0.15781062724730324, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 395 + }, + { + "epoch": 0.15821014782261286, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 396 + }, + { + "epoch": 0.15860966839792248, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 397 + }, + { + "epoch": 0.15900918897323213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 398 + }, + { + "epoch": 0.15940870954854175, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 399 + }, + { + "epoch": 0.15980823012385137, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3929, + "step": 400 + }, + { + "epoch": 0.16020775069916102, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3835, + "step": 401 + }, + { + "epoch": 0.16060727127447064, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 402 + }, + { + "epoch": 0.16100679184978026, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 403 + }, + { + "epoch": 0.1614063124250899, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.4032, + "step": 404 + }, + { + "epoch": 0.16180583300039952, + "grad_norm": 3.984375, + "learning_rate": 0.0002, + "loss": 1.48, + "step": 405 + }, + { + "epoch": 0.16220535357570914, + "grad_norm": 2.234375, + "learning_rate": 0.0002, + "loss": 1.4185, + "step": 406 + }, + { + "epoch": 0.1626048741510188, + "grad_norm": 3.921875, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 407 + }, + { + "epoch": 0.1630043947263284, + "grad_norm": 3.578125, + "learning_rate": 0.0002, + "loss": 1.4966, + "step": 408 + }, + { + "epoch": 0.16340391530163803, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 409 + }, + { + "epoch": 0.16380343587694765, + "grad_norm": 2.671875, + "learning_rate": 0.0002, + "loss": 1.4878, + "step": 410 + }, + { + "epoch": 0.1642029564522573, + "grad_norm": 2.34375, + "learning_rate": 0.0002, + "loss": 1.491, + "step": 411 + }, + { + "epoch": 0.16460247702756692, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 412 + }, + { + "epoch": 0.16500199760287654, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4232, + "step": 413 + }, + { + "epoch": 0.16540151817818619, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.4109, + "step": 414 + }, + { + "epoch": 0.1658010387534958, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.4187, + "step": 415 + }, + { + "epoch": 0.16620055932880543, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 416 + }, + { + "epoch": 0.16660007990411507, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 417 + }, + { + "epoch": 0.1669996004794247, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 418 + }, + { + "epoch": 0.1673991210547343, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3982, + "step": 419 + }, + { + "epoch": 0.16779864163004396, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 420 + }, + { + "epoch": 0.16819816220535358, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 421 + }, + { + "epoch": 0.1685976827806632, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 422 + }, + { + "epoch": 0.16899720335597282, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3933, + "step": 423 + }, + { + "epoch": 0.16939672393128247, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 424 + }, + { + "epoch": 0.1697962445065921, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3909, + "step": 425 + }, + { + "epoch": 0.1701957650819017, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3978, + "step": 426 + }, + { + "epoch": 0.17059528565721135, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 427 + }, + { + "epoch": 0.17099480623252097, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3955, + "step": 428 + }, + { + "epoch": 0.1713943268078306, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 429 + }, + { + "epoch": 0.17179384738314024, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 430 + }, + { + "epoch": 0.17219336795844986, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3965, + "step": 431 + }, + { + "epoch": 0.17259288853375948, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 432 + }, + { + "epoch": 0.17299240910906913, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 433 + }, + { + "epoch": 0.17339192968437875, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 434 + }, + { + "epoch": 0.17379145025968837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 435 + }, + { + "epoch": 0.17419097083499802, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 436 + }, + { + "epoch": 0.17459049141030764, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 437 + }, + { + "epoch": 0.17499001198561726, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3928, + "step": 438 + }, + { + "epoch": 0.17538953256092688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 439 + }, + { + "epoch": 0.17578905313623652, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 440 + }, + { + "epoch": 0.17618857371154614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 441 + }, + { + "epoch": 0.17658809428685576, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3912, + "step": 442 + }, + { + "epoch": 0.1769876148621654, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 443 + }, + { + "epoch": 0.17738713543747503, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 444 + }, + { + "epoch": 0.17778665601278465, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 445 + }, + { + "epoch": 0.1781861765880943, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 446 + }, + { + "epoch": 0.17858569716340392, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 447 + }, + { + "epoch": 0.17898521773871354, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 448 + }, + { + "epoch": 0.17938473831402318, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 449 + }, + { + "epoch": 0.1797842588893328, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 450 + }, + { + "epoch": 0.18018377946464242, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 451 + }, + { + "epoch": 0.18058330003995204, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 452 + }, + { + "epoch": 0.1809828206152617, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 453 + }, + { + "epoch": 0.1813823411905713, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 454 + }, + { + "epoch": 0.18178186176588093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.4001, + "step": 455 + }, + { + "epoch": 0.18218138234119058, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 456 + }, + { + "epoch": 0.1825809029165002, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 457 + }, + { + "epoch": 0.18298042349180982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 458 + }, + { + "epoch": 0.18337994406711947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 459 + }, + { + "epoch": 0.18377946464242909, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 460 + }, + { + "epoch": 0.1841789852177387, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 461 + }, + { + "epoch": 0.18457850579304835, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 462 + }, + { + "epoch": 0.18497802636835797, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 463 + }, + { + "epoch": 0.1853775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 464 + }, + { + "epoch": 0.18577706751897724, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 465 + }, + { + "epoch": 0.18617658809428686, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3796, + "step": 466 + }, + { + "epoch": 0.18657610866959648, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 467 + }, + { + "epoch": 0.1869756292449061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 468 + }, + { + "epoch": 0.18737514982021575, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 469 + }, + { + "epoch": 0.18777467039552537, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 470 + }, + { + "epoch": 0.188174190970835, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 471 + }, + { + "epoch": 0.18857371154614463, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 472 + }, + { + "epoch": 0.18897323212145425, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 473 + }, + { + "epoch": 0.18937275269676387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 474 + }, + { + "epoch": 0.18977227327207352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 475 + }, + { + "epoch": 0.19017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 476 + }, + { + "epoch": 0.19057131442269276, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 477 + }, + { + "epoch": 0.1909708349980024, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 478 + }, + { + "epoch": 0.19137035557331203, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 479 + }, + { + "epoch": 0.19176987614862165, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3807, + "step": 480 + }, + { + "epoch": 0.1921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 481 + }, + { + "epoch": 0.19256891729924092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 482 + }, + { + "epoch": 0.19296843787455054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 483 + }, + { + "epoch": 0.19336795844986016, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 484 + }, + { + "epoch": 0.1937674790251698, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 485 + }, + { + "epoch": 0.19416699960047942, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 486 + }, + { + "epoch": 0.19456652017578904, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 487 + }, + { + "epoch": 0.1949660407510987, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 488 + }, + { + "epoch": 0.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 489 + }, + { + "epoch": 0.19576508190171793, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 490 + }, + { + "epoch": 0.19616460247702758, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3816, + "step": 491 + }, + { + "epoch": 0.1965641230523372, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 492 + }, + { + "epoch": 0.19696364362764682, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3824, + "step": 493 + }, + { + "epoch": 0.19736316420295646, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.387, + "step": 494 + }, + { + "epoch": 0.19776268477826608, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 495 + }, + { + "epoch": 0.1981622053535757, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 496 + }, + { + "epoch": 0.19856172592888532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 497 + }, + { + "epoch": 0.19896124650419497, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 498 + }, + { + "epoch": 0.1993607670795046, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 499 + }, + { + "epoch": 0.1997602876548142, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 500 + }, + { + "epoch": 0.20015980823012386, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 501 + }, + { + "epoch": 0.20055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 502 + }, + { + "epoch": 0.2009588493807431, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 503 + }, + { + "epoch": 0.20135836995605275, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 504 + }, + { + "epoch": 0.20175789053136237, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 505 + }, + { + "epoch": 0.20215741110667199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 506 + }, + { + "epoch": 0.20255693168198163, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 507 + }, + { + "epoch": 0.20295645225729125, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 508 + }, + { + "epoch": 0.20335597283260087, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 509 + }, + { + "epoch": 0.20375549340791052, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 510 + }, + { + "epoch": 0.20415501398322014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.384, + "step": 511 + }, + { + "epoch": 0.20455453455852976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 512 + }, + { + "epoch": 0.20495405513383938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 513 + }, + { + "epoch": 0.20535357570914903, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 514 + }, + { + "epoch": 0.20575309628445865, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 515 + }, + { + "epoch": 0.20615261685976827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 516 + }, + { + "epoch": 0.20655213743507791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.38, + "step": 517 + }, + { + "epoch": 0.20695165801038753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 518 + }, + { + "epoch": 0.20735117858569715, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 519 + }, + { + "epoch": 0.2077506991610068, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 520 + }, + { + "epoch": 0.20815021973631642, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 521 + }, + { + "epoch": 0.20854974031162604, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 522 + }, + { + "epoch": 0.2089492608869357, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 523 + }, + { + "epoch": 0.2093487814622453, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 524 + }, + { + "epoch": 0.20974830203755493, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3845, + "step": 525 + }, + { + "epoch": 0.21014782261286455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 526 + }, + { + "epoch": 0.2105473431881742, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 527 + }, + { + "epoch": 0.21094686376348382, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3958, + "step": 528 + }, + { + "epoch": 0.21134638433879344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 529 + }, + { + "epoch": 0.21174590491410308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 530 + }, + { + "epoch": 0.2121454254894127, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 531 + }, + { + "epoch": 0.21254494606472232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 532 + }, + { + "epoch": 0.21294446664003197, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 533 + }, + { + "epoch": 0.2133439872153416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 534 + }, + { + "epoch": 0.2137435077906512, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 535 + }, + { + "epoch": 0.21414302836596086, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 536 + }, + { + "epoch": 0.21454254894127048, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 537 + }, + { + "epoch": 0.2149420695165801, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 538 + }, + { + "epoch": 0.21534159009188975, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 539 + }, + { + "epoch": 0.21574111066719937, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 540 + }, + { + "epoch": 0.21614063124250898, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 541 + }, + { + "epoch": 0.2165401518178186, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 542 + }, + { + "epoch": 0.21693967239312825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3754, + "step": 543 + }, + { + "epoch": 0.21733919296843787, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 544 + }, + { + "epoch": 0.2177387135437475, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 545 + }, + { + "epoch": 0.21813823411905714, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 546 + }, + { + "epoch": 0.21853775469436676, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3887, + "step": 547 + }, + { + "epoch": 0.21893727526967638, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 548 + }, + { + "epoch": 0.21933679584498603, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 549 + }, + { + "epoch": 0.21973631642029565, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 550 + }, + { + "epoch": 0.22013583699560527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 551 + }, + { + "epoch": 0.2205353575709149, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 552 + }, + { + "epoch": 0.22093487814622453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 553 + }, + { + "epoch": 0.22133439872153415, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 554 + }, + { + "epoch": 0.2217339192968438, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3992, + "step": 555 + }, + { + "epoch": 0.22213343987215342, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 556 + }, + { + "epoch": 0.22253296044746304, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 557 + }, + { + "epoch": 0.22293248102277266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 558 + }, + { + "epoch": 0.2233320015980823, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 559 + }, + { + "epoch": 0.22373152217339193, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 560 + }, + { + "epoch": 0.22413104274870155, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 561 + }, + { + "epoch": 0.2245305633240112, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 562 + }, + { + "epoch": 0.22493008389932082, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 563 + }, + { + "epoch": 0.22532960447463043, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 564 + }, + { + "epoch": 0.22572912504994008, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 565 + }, + { + "epoch": 0.2261286456252497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 566 + }, + { + "epoch": 0.22652816620055932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 567 + }, + { + "epoch": 0.22692768677586897, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 568 + }, + { + "epoch": 0.2273272073511786, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 569 + }, + { + "epoch": 0.2277267279264882, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 570 + }, + { + "epoch": 0.22812624850179783, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 571 + }, + { + "epoch": 0.22852576907710748, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 572 + }, + { + "epoch": 0.2289252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 573 + }, + { + "epoch": 0.22932481022772672, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 574 + }, + { + "epoch": 0.22972433080303636, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 575 + }, + { + "epoch": 0.23012385137834598, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 576 + }, + { + "epoch": 0.2305233719536556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3786, + "step": 577 + }, + { + "epoch": 0.23092289252896525, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 578 + }, + { + "epoch": 0.23132241310427487, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 579 + }, + { + "epoch": 0.2317219336795845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 580 + }, + { + "epoch": 0.23212145425489414, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 581 + }, + { + "epoch": 0.23252097483020376, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3781, + "step": 582 + }, + { + "epoch": 0.23292049540551338, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 583 + }, + { + "epoch": 0.23332001598082303, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 584 + }, + { + "epoch": 0.23371953655613265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 585 + }, + { + "epoch": 0.23411905713144227, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 586 + }, + { + "epoch": 0.23451857770675189, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 587 + }, + { + "epoch": 0.23491809828206153, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 588 + }, + { + "epoch": 0.23531761885737115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 589 + }, + { + "epoch": 0.23571713943268077, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 590 + }, + { + "epoch": 0.23611666000799042, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 591 + }, + { + "epoch": 0.23651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 592 + }, + { + "epoch": 0.23691570115860966, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 593 + }, + { + "epoch": 0.2373152217339193, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 594 + }, + { + "epoch": 0.23771474230922893, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 595 + }, + { + "epoch": 0.23811426288453855, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 596 + }, + { + "epoch": 0.2385137834598482, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3914, + "step": 597 + }, + { + "epoch": 0.23891330403515781, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 598 + }, + { + "epoch": 0.23931282461046743, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 599 + }, + { + "epoch": 0.23971234518577705, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 600 + }, + { + "epoch": 0.2401118657610867, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 601 + }, + { + "epoch": 0.24051138633639632, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 602 + }, + { + "epoch": 0.24091090691170594, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 603 + }, + { + "epoch": 0.2413104274870156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 604 + }, + { + "epoch": 0.2417099480623252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 605 + }, + { + "epoch": 0.24210946863763483, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 606 + }, + { + "epoch": 0.24250898921294448, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 607 + }, + { + "epoch": 0.2429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 608 + }, + { + "epoch": 0.24330803036356372, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 609 + }, + { + "epoch": 0.24370755093887336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 610 + }, + { + "epoch": 0.24410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 611 + }, + { + "epoch": 0.2445065920894926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 612 + }, + { + "epoch": 0.24490611266480225, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 613 + }, + { + "epoch": 0.24530563324011187, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 614 + }, + { + "epoch": 0.2457051538154215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 615 + }, + { + "epoch": 0.2461046743907311, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3783, + "step": 616 + }, + { + "epoch": 0.24650419496604076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 617 + }, + { + "epoch": 0.24690371554135038, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 618 + }, + { + "epoch": 0.24730323611666, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 619 + }, + { + "epoch": 0.24770275669196964, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 620 + }, + { + "epoch": 0.24810227726727926, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 621 + }, + { + "epoch": 0.24850179784258888, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 622 + }, + { + "epoch": 0.24890131841789853, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 623 + }, + { + "epoch": 0.24930083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 624 + }, + { + "epoch": 0.24970035956851777, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 625 + }, + { + "epoch": 0.2500998801438274, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 626 + }, + { + "epoch": 0.250499400719137, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3748, + "step": 627 + }, + { + "epoch": 0.2508989212944467, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 628 + }, + { + "epoch": 0.2512984418697563, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.377, + "step": 629 + }, + { + "epoch": 0.2516979624450659, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 630 + }, + { + "epoch": 0.25209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 631 + }, + { + "epoch": 0.25249700359568517, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 632 + }, + { + "epoch": 0.2528965241709948, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 633 + }, + { + "epoch": 0.25329604474630446, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 634 + }, + { + "epoch": 0.2536955653216141, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 635 + }, + { + "epoch": 0.2540950858969237, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 636 + }, + { + "epoch": 0.2544946064722333, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 637 + }, + { + "epoch": 0.25489412704754294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 638 + }, + { + "epoch": 0.25529364762285256, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 639 + }, + { + "epoch": 0.2556931681981622, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 640 + }, + { + "epoch": 0.25609268877347185, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 641 + }, + { + "epoch": 0.2564922093487815, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 642 + }, + { + "epoch": 0.2568917299240911, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 643 + }, + { + "epoch": 0.2572912504994007, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 644 + }, + { + "epoch": 0.25769077107471033, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 645 + }, + { + "epoch": 0.25809029165001995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 646 + }, + { + "epoch": 0.25848981222532963, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 647 + }, + { + "epoch": 0.25888933280063925, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 648 + }, + { + "epoch": 0.25928885337594887, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 649 + }, + { + "epoch": 0.2596883739512585, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 650 + }, + { + "epoch": 0.2600878945265681, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 651 + }, + { + "epoch": 0.26048741510187773, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 652 + }, + { + "epoch": 0.26088693567718735, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3866, + "step": 653 + }, + { + "epoch": 0.261286456252497, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 654 + }, + { + "epoch": 0.26168597682780664, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 655 + }, + { + "epoch": 0.26208549740311626, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 656 + }, + { + "epoch": 0.2624850179784259, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 657 + }, + { + "epoch": 0.2628845385537355, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3802, + "step": 658 + }, + { + "epoch": 0.2632840591290451, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 659 + }, + { + "epoch": 0.2636835797043548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 660 + }, + { + "epoch": 0.2640831002796644, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 661 + }, + { + "epoch": 0.26448262085497404, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 662 + }, + { + "epoch": 0.26488214143028366, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 663 + }, + { + "epoch": 0.2652816620055933, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 664 + }, + { + "epoch": 0.2656811825809029, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 665 + }, + { + "epoch": 0.2660807031562126, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 666 + }, + { + "epoch": 0.2664802237315222, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 667 + }, + { + "epoch": 0.2668797443068318, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 668 + }, + { + "epoch": 0.26727926488214143, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 669 + }, + { + "epoch": 0.26767878545745105, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 670 + }, + { + "epoch": 0.26807830603276067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 671 + }, + { + "epoch": 0.2684778266080703, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 672 + }, + { + "epoch": 0.26887734718337997, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 673 + }, + { + "epoch": 0.2692768677586896, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 674 + }, + { + "epoch": 0.2696763883339992, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 675 + }, + { + "epoch": 0.2700759089093088, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 676 + }, + { + "epoch": 0.27047542948461845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 677 + }, + { + "epoch": 0.27087495005992807, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 678 + }, + { + "epoch": 0.27127447063523774, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 679 + }, + { + "epoch": 0.27167399121054736, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 680 + }, + { + "epoch": 0.272073511785857, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 681 + }, + { + "epoch": 0.2724730323611666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 682 + }, + { + "epoch": 0.2728725529364762, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 683 + }, + { + "epoch": 0.27327207351178584, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 684 + }, + { + "epoch": 0.27367159408709546, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 685 + }, + { + "epoch": 0.27407111466240514, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3747, + "step": 686 + }, + { + "epoch": 0.27447063523771476, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 687 + }, + { + "epoch": 0.2748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 688 + }, + { + "epoch": 0.275269676388334, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 689 + }, + { + "epoch": 0.2756691969636436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 690 + }, + { + "epoch": 0.27606871753895323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 691 + }, + { + "epoch": 0.2764682381142629, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 692 + }, + { + "epoch": 0.27686775868957253, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 693 + }, + { + "epoch": 0.27726727926488215, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 694 + }, + { + "epoch": 0.27766679984019177, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3842, + "step": 695 + }, + { + "epoch": 0.2780663204155014, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 696 + }, + { + "epoch": 0.278465840990811, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 697 + }, + { + "epoch": 0.27886536156612063, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 698 + }, + { + "epoch": 0.2792648821414303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 699 + }, + { + "epoch": 0.2796644027167399, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 700 + }, + { + "epoch": 0.28006392329204954, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3742, + "step": 701 + }, + { + "epoch": 0.28046344386735916, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 702 + }, + { + "epoch": 0.2808629644426688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 703 + }, + { + "epoch": 0.2812624850179784, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 704 + }, + { + "epoch": 0.2816620055932881, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 705 + }, + { + "epoch": 0.2820615261685977, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 706 + }, + { + "epoch": 0.2824610467439073, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 707 + }, + { + "epoch": 0.28286056731921694, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 708 + }, + { + "epoch": 0.28326008789452656, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 709 + }, + { + "epoch": 0.2836596084698362, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 710 + }, + { + "epoch": 0.2840591290451458, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 711 + }, + { + "epoch": 0.2844586496204555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 712 + }, + { + "epoch": 0.2848581701957651, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 713 + }, + { + "epoch": 0.2852576907710747, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 714 + }, + { + "epoch": 0.28565721134638433, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 715 + }, + { + "epoch": 0.28605673192169395, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 716 + }, + { + "epoch": 0.28645625249700357, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 717 + }, + { + "epoch": 0.28685577307231325, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 718 + }, + { + "epoch": 0.28725529364762287, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 719 + }, + { + "epoch": 0.2876548142229325, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 720 + }, + { + "epoch": 0.2880543347982421, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3826, + "step": 721 + }, + { + "epoch": 0.2884538553735517, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 722 + }, + { + "epoch": 0.28885337594886135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 723 + }, + { + "epoch": 0.289252896524171, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 724 + }, + { + "epoch": 0.28965241709948064, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 725 + }, + { + "epoch": 0.29005193767479026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 726 + }, + { + "epoch": 0.2904514582500999, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 727 + }, + { + "epoch": 0.2908509788254095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 728 + }, + { + "epoch": 0.2912504994007191, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3854, + "step": 729 + }, + { + "epoch": 0.29165001997602874, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 730 + }, + { + "epoch": 0.2920495405513384, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 731 + }, + { + "epoch": 0.29244906112664804, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 732 + }, + { + "epoch": 0.29284858170195766, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 733 + }, + { + "epoch": 0.2932481022772673, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 734 + }, + { + "epoch": 0.2936476228525769, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 735 + }, + { + "epoch": 0.2940471434278865, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 736 + }, + { + "epoch": 0.2944466640031962, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 737 + }, + { + "epoch": 0.2948461845785058, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 738 + }, + { + "epoch": 0.29524570515381543, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 739 + }, + { + "epoch": 0.29564522572912505, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 740 + }, + { + "epoch": 0.29604474630443467, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 741 + }, + { + "epoch": 0.2964442668797443, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 742 + }, + { + "epoch": 0.2968437874550539, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 743 + }, + { + "epoch": 0.2972433080303636, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 744 + }, + { + "epoch": 0.2976428286056732, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 745 + }, + { + "epoch": 0.2980423491809828, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 746 + }, + { + "epoch": 0.29844186975629244, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 747 + }, + { + "epoch": 0.29884139033160206, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 748 + }, + { + "epoch": 0.2992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 749 + }, + { + "epoch": 0.29964043148222136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.382, + "step": 750 + }, + { + "epoch": 0.300039952057531, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 751 + }, + { + "epoch": 0.3004394726328406, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 752 + }, + { + "epoch": 0.3008389932081502, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 753 + }, + { + "epoch": 0.30123851378345984, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 754 + }, + { + "epoch": 0.30163803435876946, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 755 + }, + { + "epoch": 0.3020375549340791, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 756 + }, + { + "epoch": 0.30243707550938875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 757 + }, + { + "epoch": 0.3028365960846984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 758 + }, + { + "epoch": 0.303236116660008, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 759 + }, + { + "epoch": 0.3036356372353176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3698, + "step": 760 + }, + { + "epoch": 0.30403515781062723, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 761 + }, + { + "epoch": 0.30443467838593685, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3761, + "step": 762 + }, + { + "epoch": 0.3048341989612465, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 763 + }, + { + "epoch": 0.30523371953655615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 764 + }, + { + "epoch": 0.30563324011186577, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 765 + }, + { + "epoch": 0.3060327606871754, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 766 + }, + { + "epoch": 0.306432281262485, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 767 + }, + { + "epoch": 0.3068318018377946, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 768 + }, + { + "epoch": 0.3072313224131043, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3694, + "step": 769 + }, + { + "epoch": 0.3076308429884139, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 770 + }, + { + "epoch": 0.30803036356372354, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 771 + }, + { + "epoch": 0.30842988413903316, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 772 + }, + { + "epoch": 0.3088294047143428, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 773 + }, + { + "epoch": 0.3092289252896524, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 774 + }, + { + "epoch": 0.309628445864962, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3775, + "step": 775 + }, + { + "epoch": 0.3100279664402717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 776 + }, + { + "epoch": 0.3104274870155813, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 777 + }, + { + "epoch": 0.31082700759089094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 778 + }, + { + "epoch": 0.31122652816620056, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 779 + }, + { + "epoch": 0.3116260487415102, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 780 + }, + { + "epoch": 0.3120255693168198, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 781 + }, + { + "epoch": 0.31242508989212947, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 782 + }, + { + "epoch": 0.3128246104674391, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 783 + }, + { + "epoch": 0.3132241310427487, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 784 + }, + { + "epoch": 0.31362365161805833, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 785 + }, + { + "epoch": 0.31402317219336795, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 786 + }, + { + "epoch": 0.31442269276867757, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 787 + }, + { + "epoch": 0.3148222133439872, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 788 + }, + { + "epoch": 0.31522173391929686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 789 + }, + { + "epoch": 0.3156212544946065, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 790 + }, + { + "epoch": 0.3160207750699161, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 791 + }, + { + "epoch": 0.3164202956452257, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3728, + "step": 792 + }, + { + "epoch": 0.31681981622053534, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 793 + }, + { + "epoch": 0.31721933679584496, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 794 + }, + { + "epoch": 0.31761885737115464, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 795 + }, + { + "epoch": 0.31801837794646426, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 796 + }, + { + "epoch": 0.3184178985217739, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 797 + }, + { + "epoch": 0.3188174190970835, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 798 + }, + { + "epoch": 0.3192169396723931, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 799 + }, + { + "epoch": 0.31961646024770274, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 800 + }, + { + "epoch": 0.32001598082301236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 801 + }, + { + "epoch": 0.32041550139832203, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 802 + }, + { + "epoch": 0.32081502197363165, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 803 + }, + { + "epoch": 0.3212145425489413, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 804 + }, + { + "epoch": 0.3216140631242509, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 805 + }, + { + "epoch": 0.3220135836995605, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 806 + }, + { + "epoch": 0.32241310427487013, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 807 + }, + { + "epoch": 0.3228126248501798, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 808 + }, + { + "epoch": 0.3232121454254894, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 809 + }, + { + "epoch": 0.32361166600079905, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 810 + }, + { + "epoch": 0.32401118657610867, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 811 + }, + { + "epoch": 0.3244107071514183, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 812 + }, + { + "epoch": 0.3248102277267279, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 813 + }, + { + "epoch": 0.3252097483020376, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.373, + "step": 814 + }, + { + "epoch": 0.3256092688773472, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 815 + }, + { + "epoch": 0.3260087894526568, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 816 + }, + { + "epoch": 0.32640831002796644, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 817 + }, + { + "epoch": 0.32680783060327606, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 818 + }, + { + "epoch": 0.3272073511785857, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 819 + }, + { + "epoch": 0.3276068717538953, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 820 + }, + { + "epoch": 0.328006392329205, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 821 + }, + { + "epoch": 0.3284059129045146, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 822 + }, + { + "epoch": 0.3288054334798242, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 823 + }, + { + "epoch": 0.32920495405513384, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 824 + }, + { + "epoch": 0.32960447463044346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 825 + }, + { + "epoch": 0.3300039952057531, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 826 + }, + { + "epoch": 0.33040351578106275, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 827 + }, + { + "epoch": 0.33080303635637237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 828 + }, + { + "epoch": 0.331202556931682, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 829 + }, + { + "epoch": 0.3316020775069916, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 830 + }, + { + "epoch": 0.33200159808230123, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 831 + }, + { + "epoch": 0.33240111865761085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 832 + }, + { + "epoch": 0.33280063923292047, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 833 + }, + { + "epoch": 0.33320015980823015, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 834 + }, + { + "epoch": 0.33359968038353977, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 835 + }, + { + "epoch": 0.3339992009588494, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 836 + }, + { + "epoch": 0.334398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 837 + }, + { + "epoch": 0.3347982421094686, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 838 + }, + { + "epoch": 0.33519776268477824, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 839 + }, + { + "epoch": 0.3355972832600879, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 840 + }, + { + "epoch": 0.33599680383539754, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 841 + }, + { + "epoch": 0.33639632441070716, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 842 + }, + { + "epoch": 0.3367958449860168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 843 + }, + { + "epoch": 0.3371953655613264, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 844 + }, + { + "epoch": 0.337594886136636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 845 + }, + { + "epoch": 0.33799440671194564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 846 + }, + { + "epoch": 0.3383939272872553, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 847 + }, + { + "epoch": 0.33879344786256493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 848 + }, + { + "epoch": 0.33919296843787455, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 849 + }, + { + "epoch": 0.3395924890131842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 850 + }, + { + "epoch": 0.3399920095884938, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 851 + }, + { + "epoch": 0.3403915301638034, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 852 + }, + { + "epoch": 0.3407910507391131, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 853 + }, + { + "epoch": 0.3411905713144227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 854 + }, + { + "epoch": 0.34159009188973233, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 855 + }, + { + "epoch": 0.34198961246504195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 856 + }, + { + "epoch": 0.34238913304035157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 857 + }, + { + "epoch": 0.3427886536156612, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 858 + }, + { + "epoch": 0.34318817419097086, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 859 + }, + { + "epoch": 0.3435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 860 + }, + { + "epoch": 0.3439872153415901, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 861 + }, + { + "epoch": 0.3443867359168997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 862 + }, + { + "epoch": 0.34478625649220934, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 863 + }, + { + "epoch": 0.34518577706751896, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 864 + }, + { + "epoch": 0.3455852976428286, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 865 + }, + { + "epoch": 0.34598481821813826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 866 + }, + { + "epoch": 0.3463843387934479, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3776, + "step": 867 + }, + { + "epoch": 0.3467838593687575, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 868 + }, + { + "epoch": 0.3471833799440671, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 869 + }, + { + "epoch": 0.34758290051937674, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 870 + }, + { + "epoch": 0.34798242109468636, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 871 + }, + { + "epoch": 0.34838194166999603, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 872 + }, + { + "epoch": 0.34878146224530565, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 873 + }, + { + "epoch": 0.34918098282061527, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 874 + }, + { + "epoch": 0.3495805033959249, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 875 + }, + { + "epoch": 0.3499800239712345, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 876 + }, + { + "epoch": 0.35037954454654413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 877 + }, + { + "epoch": 0.35077906512185375, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 878 + }, + { + "epoch": 0.3511785856971634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 879 + }, + { + "epoch": 0.35157810627247305, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 880 + }, + { + "epoch": 0.35197762684778267, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 881 + }, + { + "epoch": 0.3523771474230923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 882 + }, + { + "epoch": 0.3527766679984019, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 883 + }, + { + "epoch": 0.3531761885737115, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 884 + }, + { + "epoch": 0.3535757091490212, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 885 + }, + { + "epoch": 0.3539752297243308, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 886 + }, + { + "epoch": 0.35437475029964044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 887 + }, + { + "epoch": 0.35477427087495006, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3751, + "step": 888 + }, + { + "epoch": 0.3551737914502597, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 889 + }, + { + "epoch": 0.3555733120255693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 890 + }, + { + "epoch": 0.3559728326008789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 891 + }, + { + "epoch": 0.3563723531761886, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 892 + }, + { + "epoch": 0.3567718737514982, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 893 + }, + { + "epoch": 0.35717139432680783, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 894 + }, + { + "epoch": 0.35757091490211745, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 895 + }, + { + "epoch": 0.3579704354774271, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3758, + "step": 896 + }, + { + "epoch": 0.3583699560527367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 897 + }, + { + "epoch": 0.35876947662804637, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 898 + }, + { + "epoch": 0.359168997203356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 899 + }, + { + "epoch": 0.3595685177786656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 900 + }, + { + "epoch": 0.35996803835397523, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 901 + }, + { + "epoch": 0.36036755892928485, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3774, + "step": 902 + }, + { + "epoch": 0.36076707950459447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 903 + }, + { + "epoch": 0.3611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 904 + }, + { + "epoch": 0.36156612065521376, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 905 + }, + { + "epoch": 0.3619656412305234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 906 + }, + { + "epoch": 0.362365161805833, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 907 + }, + { + "epoch": 0.3627646823811426, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 908 + }, + { + "epoch": 0.36316420295645224, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 909 + }, + { + "epoch": 0.36356372353176186, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 910 + }, + { + "epoch": 0.36396324410707154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 911 + }, + { + "epoch": 0.36436276468238116, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 912 + }, + { + "epoch": 0.3647622852576908, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 913 + }, + { + "epoch": 0.3651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 914 + }, + { + "epoch": 0.36556132640831, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 915 + }, + { + "epoch": 0.36596084698361964, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 916 + }, + { + "epoch": 0.3663603675589293, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 917 + }, + { + "epoch": 0.36675988813423893, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 918 + }, + { + "epoch": 0.36715940870954855, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 919 + }, + { + "epoch": 0.36755892928485817, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 920 + }, + { + "epoch": 0.3679584498601678, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 921 + }, + { + "epoch": 0.3683579704354774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 922 + }, + { + "epoch": 0.36875749101078703, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 923 + }, + { + "epoch": 0.3691570115860967, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 924 + }, + { + "epoch": 0.3695565321614063, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 925 + }, + { + "epoch": 0.36995605273671595, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 926 + }, + { + "epoch": 0.37035557331202557, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 927 + }, + { + "epoch": 0.3707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 928 + }, + { + "epoch": 0.3711546144626448, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 929 + }, + { + "epoch": 0.3715541350379545, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 930 + }, + { + "epoch": 0.3719536556132641, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 931 + }, + { + "epoch": 0.3723531761885737, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 932 + }, + { + "epoch": 0.37275269676388334, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 933 + }, + { + "epoch": 0.37315221733919296, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 934 + }, + { + "epoch": 0.3735517379145026, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 935 + }, + { + "epoch": 0.3739512584898122, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 936 + }, + { + "epoch": 0.3743507790651219, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 937 + }, + { + "epoch": 0.3747502996404315, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 938 + }, + { + "epoch": 0.3751498202157411, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 939 + }, + { + "epoch": 0.37554934079105073, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 940 + }, + { + "epoch": 0.37594886136636035, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 941 + }, + { + "epoch": 0.37634838194167, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 942 + }, + { + "epoch": 0.37674790251697965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 943 + }, + { + "epoch": 0.37714742309228927, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 944 + }, + { + "epoch": 0.3775469436675989, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 945 + }, + { + "epoch": 0.3779464642429085, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 946 + }, + { + "epoch": 0.37834598481821813, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 947 + }, + { + "epoch": 0.37874550539352775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 948 + }, + { + "epoch": 0.37914502596883737, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 949 + }, + { + "epoch": 0.37954454654414704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 950 + }, + { + "epoch": 0.37994406711945666, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 951 + }, + { + "epoch": 0.3803435876947663, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 952 + }, + { + "epoch": 0.3807431082700759, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 953 + }, + { + "epoch": 0.3811426288453855, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 954 + }, + { + "epoch": 0.38154214942069514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 955 + }, + { + "epoch": 0.3819416699960048, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 956 + }, + { + "epoch": 0.38234119057131444, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 957 + }, + { + "epoch": 0.38274071114662406, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 958 + }, + { + "epoch": 0.3831402317219337, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 959 + }, + { + "epoch": 0.3835397522972433, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 960 + }, + { + "epoch": 0.3839392728725529, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 961 + }, + { + "epoch": 0.3843387934478626, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3661, + "step": 962 + }, + { + "epoch": 0.3847383140231722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 963 + }, + { + "epoch": 0.38513783459848183, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 964 + }, + { + "epoch": 0.38553735517379145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 965 + }, + { + "epoch": 0.38593687574910107, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 966 + }, + { + "epoch": 0.3863363963244107, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 967 + }, + { + "epoch": 0.3867359168997203, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 968 + }, + { + "epoch": 0.38713543747503, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 969 + }, + { + "epoch": 0.3875349580503396, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 970 + }, + { + "epoch": 0.3879344786256492, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 971 + }, + { + "epoch": 0.38833399920095885, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 972 + }, + { + "epoch": 0.38873351977626847, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 973 + }, + { + "epoch": 0.3891330403515781, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 974 + }, + { + "epoch": 0.38953256092688776, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 975 + }, + { + "epoch": 0.3899320815021974, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 976 + }, + { + "epoch": 0.390331602077507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 977 + }, + { + "epoch": 0.3907311226528166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 978 + }, + { + "epoch": 0.39113064322812624, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 979 + }, + { + "epoch": 0.39153016380343586, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 980 + }, + { + "epoch": 0.3919296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 981 + }, + { + "epoch": 0.39232920495405516, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 982 + }, + { + "epoch": 0.3927287255293648, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 983 + }, + { + "epoch": 0.3931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 984 + }, + { + "epoch": 0.393527766679984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 985 + }, + { + "epoch": 0.39392728725529363, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 986 + }, + { + "epoch": 0.39432680783060325, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 987 + }, + { + "epoch": 0.39472632840591293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 988 + }, + { + "epoch": 0.39512584898122255, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 989 + }, + { + "epoch": 0.39552536955653217, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 990 + }, + { + "epoch": 0.3959248901318418, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 991 + }, + { + "epoch": 0.3963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 992 + }, + { + "epoch": 0.39672393128246103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 993 + }, + { + "epoch": 0.39712345185777065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 994 + }, + { + "epoch": 0.3975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 995 + }, + { + "epoch": 0.39792249300838994, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 996 + }, + { + "epoch": 0.39832201358369956, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 997 + }, + { + "epoch": 0.3987215341590092, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 998 + }, + { + "epoch": 0.3991210547343188, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 999 + }, + { + "epoch": 0.3995205753096284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1000 + }, + { + "epoch": 0.3999200958849381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1001 + }, + { + "epoch": 0.4003196164602477, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1002 + }, + { + "epoch": 0.40071913703555734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1003 + }, + { + "epoch": 0.40111865761086696, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1004 + }, + { + "epoch": 0.4015181781861766, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 1005 + }, + { + "epoch": 0.4019176987614862, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1006 + }, + { + "epoch": 0.4023172193367959, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1007 + }, + { + "epoch": 0.4027167399121055, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1008 + }, + { + "epoch": 0.4031162604874151, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1009 + }, + { + "epoch": 0.40351578106272473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1010 + }, + { + "epoch": 0.40391530163803435, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1011 + }, + { + "epoch": 0.40431482221334397, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1012 + }, + { + "epoch": 0.4047143427886536, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 1013 + }, + { + "epoch": 0.40511386336396327, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1014 + }, + { + "epoch": 0.4055133839392729, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1015 + }, + { + "epoch": 0.4059129045145825, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1016 + }, + { + "epoch": 0.4063124250898921, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1017 + }, + { + "epoch": 0.40671194566520175, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 1018 + }, + { + "epoch": 0.40711146624051137, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1019 + }, + { + "epoch": 0.40751098681582104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1020 + }, + { + "epoch": 0.40791050739113066, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1021 + }, + { + "epoch": 0.4083100279664403, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1022 + }, + { + "epoch": 0.4087095485417499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1023 + }, + { + "epoch": 0.4091090691170595, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 1024 + }, + { + "epoch": 0.40950858969236914, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1025 + }, + { + "epoch": 0.40990811026767876, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1026 + }, + { + "epoch": 0.41030763084298844, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1027 + }, + { + "epoch": 0.41070715141829806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1028 + }, + { + "epoch": 0.4111066719936077, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1029 + }, + { + "epoch": 0.4115061925689173, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1030 + }, + { + "epoch": 0.4119057131442269, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1031 + }, + { + "epoch": 0.41230523371953653, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1032 + }, + { + "epoch": 0.4127047542948462, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1033 + }, + { + "epoch": 0.41310427487015583, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1034 + }, + { + "epoch": 0.41350379544546545, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1035 + }, + { + "epoch": 0.41390331602077507, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1036 + }, + { + "epoch": 0.4143028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1037 + }, + { + "epoch": 0.4147023571713943, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1038 + }, + { + "epoch": 0.41510187774670393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1039 + }, + { + "epoch": 0.4155013983220136, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1040 + }, + { + "epoch": 0.4159009188973232, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1041 + }, + { + "epoch": 0.41630043947263284, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1042 + }, + { + "epoch": 0.41669996004794246, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1043 + }, + { + "epoch": 0.4170994806232521, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1044 + }, + { + "epoch": 0.4174990011985617, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1045 + }, + { + "epoch": 0.4178985217738714, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 1046 + }, + { + "epoch": 0.418298042349181, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1047 + }, + { + "epoch": 0.4186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1048 + }, + { + "epoch": 0.41909708349980024, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3711, + "step": 1049 + }, + { + "epoch": 0.41949660407510986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1050 + }, + { + "epoch": 0.4198961246504195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1051 + }, + { + "epoch": 0.4202956452257291, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1052 + }, + { + "epoch": 0.4206951658010388, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1053 + }, + { + "epoch": 0.4210946863763484, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1054 + }, + { + "epoch": 0.421494206951658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1055 + }, + { + "epoch": 0.42189372752696763, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1056 + }, + { + "epoch": 0.42229324810227725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 1057 + }, + { + "epoch": 0.42269276867758687, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1058 + }, + { + "epoch": 0.42309228925289655, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 1059 + }, + { + "epoch": 0.42349180982820617, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1060 + }, + { + "epoch": 0.4238913304035158, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1061 + }, + { + "epoch": 0.4242908509788254, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1062 + }, + { + "epoch": 0.424690371554135, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1063 + }, + { + "epoch": 0.42508989212944465, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1064 + }, + { + "epoch": 0.4254894127047543, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 1065 + }, + { + "epoch": 0.42588893328006394, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1066 + }, + { + "epoch": 0.42628845385537356, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1067 + }, + { + "epoch": 0.4266879744306832, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 1068 + }, + { + "epoch": 0.4270874950059928, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1069 + }, + { + "epoch": 0.4274870155813024, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1070 + }, + { + "epoch": 0.42788653615661204, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1071 + }, + { + "epoch": 0.4282860567319217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1072 + }, + { + "epoch": 0.42868557730723134, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1073 + }, + { + "epoch": 0.42908509788254096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1074 + }, + { + "epoch": 0.4294846184578506, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1075 + }, + { + "epoch": 0.4298841390331602, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1076 + }, + { + "epoch": 0.4302836596084698, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1077 + }, + { + "epoch": 0.4306831801837795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1078 + }, + { + "epoch": 0.4310827007590891, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1079 + }, + { + "epoch": 0.43148222133439873, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1080 + }, + { + "epoch": 0.43188174190970835, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1081 + }, + { + "epoch": 0.43228126248501797, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1082 + }, + { + "epoch": 0.4326807830603276, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 1083 + }, + { + "epoch": 0.4330803036356372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1084 + }, + { + "epoch": 0.4334798242109469, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1085 + }, + { + "epoch": 0.4338793447862565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1086 + }, + { + "epoch": 0.4342788653615661, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 1087 + }, + { + "epoch": 0.43467838593687574, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1088 + }, + { + "epoch": 0.43507790651218536, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1089 + }, + { + "epoch": 0.435477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1090 + }, + { + "epoch": 0.43587694766280466, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1091 + }, + { + "epoch": 0.4362764682381143, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1092 + }, + { + "epoch": 0.4366759888134239, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1093 + }, + { + "epoch": 0.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1094 + }, + { + "epoch": 0.43747502996404314, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1095 + }, + { + "epoch": 0.43787455053935276, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1096 + }, + { + "epoch": 0.4382740711146624, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1097 + }, + { + "epoch": 0.43867359168997205, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1098 + }, + { + "epoch": 0.4390731122652817, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1099 + }, + { + "epoch": 0.4394726328405913, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1100 + }, + { + "epoch": 0.4398721534159009, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1101 + }, + { + "epoch": 0.44027167399121053, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1102 + }, + { + "epoch": 0.44067119456652015, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 1103 + }, + { + "epoch": 0.4410707151418298, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1104 + }, + { + "epoch": 0.44147023571713945, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1105 + }, + { + "epoch": 0.44186975629244907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1106 + }, + { + "epoch": 0.4422692768677587, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1107 + }, + { + "epoch": 0.4426687974430683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1108 + }, + { + "epoch": 0.4430683180183779, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1109 + }, + { + "epoch": 0.4434678385936876, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 1110 + }, + { + "epoch": 0.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1111 + }, + { + "epoch": 0.44426687974430684, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1112 + }, + { + "epoch": 0.44466640031961646, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1113 + }, + { + "epoch": 0.4450659208949261, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1114 + }, + { + "epoch": 0.4454654414702357, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1115 + }, + { + "epoch": 0.4458649620455453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1116 + }, + { + "epoch": 0.446264482620855, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 1117 + }, + { + "epoch": 0.4466640031961646, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1118 + }, + { + "epoch": 0.44706352377147424, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1119 + }, + { + "epoch": 0.44746304434678386, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1120 + }, + { + "epoch": 0.4478625649220935, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1121 + }, + { + "epoch": 0.4482620854974031, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1122 + }, + { + "epoch": 0.44866160607271277, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1123 + }, + { + "epoch": 0.4490611266480224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1124 + }, + { + "epoch": 0.449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 1125 + }, + { + "epoch": 0.44986016779864163, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1126 + }, + { + "epoch": 0.45025968837395125, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1127 + }, + { + "epoch": 0.45065920894926087, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1128 + }, + { + "epoch": 0.4510587295245705, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1129 + }, + { + "epoch": 0.45145825009988017, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1130 + }, + { + "epoch": 0.4518577706751898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1131 + }, + { + "epoch": 0.4522572912504994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1132 + }, + { + "epoch": 0.452656811825809, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 1133 + }, + { + "epoch": 0.45305633240111864, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1134 + }, + { + "epoch": 0.45345585297642826, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1135 + }, + { + "epoch": 0.45385537355173794, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1136 + }, + { + "epoch": 0.45425489412704756, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1137 + }, + { + "epoch": 0.4546544147023572, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1138 + }, + { + "epoch": 0.4550539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 1139 + }, + { + "epoch": 0.4554534558529764, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1140 + }, + { + "epoch": 0.45585297642828604, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 1141 + }, + { + "epoch": 0.45625249700359566, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1142 + }, + { + "epoch": 0.45665201757890533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1143 + }, + { + "epoch": 0.45705153815421495, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1144 + }, + { + "epoch": 0.4574510587295246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1145 + }, + { + "epoch": 0.4578505793048342, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1146 + }, + { + "epoch": 0.4582500998801438, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1147 + }, + { + "epoch": 0.45864962045545343, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1148 + }, + { + "epoch": 0.4590491410307631, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1149 + }, + { + "epoch": 0.45944866160607273, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1150 + }, + { + "epoch": 0.45984818218138235, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3724, + "step": 1151 + }, + { + "epoch": 0.46024770275669197, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1152 + }, + { + "epoch": 0.4606472233320016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1153 + }, + { + "epoch": 0.4610467439073112, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1154 + }, + { + "epoch": 0.4614462644826209, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1155 + }, + { + "epoch": 0.4618457850579305, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1156 + }, + { + "epoch": 0.4622453056332401, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1157 + }, + { + "epoch": 0.46264482620854974, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1158 + }, + { + "epoch": 0.46304434678385936, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1159 + }, + { + "epoch": 0.463443867359169, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 1160 + }, + { + "epoch": 0.4638433879344786, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1161 + }, + { + "epoch": 0.4642429085097883, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1162 + }, + { + "epoch": 0.4646424290850979, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1163 + }, + { + "epoch": 0.4650419496604075, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 1164 + }, + { + "epoch": 0.46544147023571714, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1165 + }, + { + "epoch": 0.46584099081102676, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1166 + }, + { + "epoch": 0.4662405113863364, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1167 + }, + { + "epoch": 0.46664003196164605, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1168 + }, + { + "epoch": 0.46703955253695567, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1169 + }, + { + "epoch": 0.4674390731122653, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1170 + }, + { + "epoch": 0.4678385936875749, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1171 + }, + { + "epoch": 0.46823811426288453, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1172 + }, + { + "epoch": 0.46863763483819415, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1173 + }, + { + "epoch": 0.46903715541350377, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1174 + }, + { + "epoch": 0.46943667598881345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1175 + }, + { + "epoch": 0.46983619656412307, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1176 + }, + { + "epoch": 0.4702357171394327, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1177 + }, + { + "epoch": 0.4706352377147423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 1178 + }, + { + "epoch": 0.4710347582900519, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1179 + }, + { + "epoch": 0.47143427886536154, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1180 + }, + { + "epoch": 0.4718337994406712, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 1181 + }, + { + "epoch": 0.47223332001598084, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1182 + }, + { + "epoch": 0.47263284059129046, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1183 + }, + { + "epoch": 0.4730323611666001, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1184 + }, + { + "epoch": 0.4734318817419097, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1185 + }, + { + "epoch": 0.4738314023172193, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1186 + }, + { + "epoch": 0.47423092289252894, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1187 + }, + { + "epoch": 0.4746304434678386, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1188 + }, + { + "epoch": 0.47502996404314823, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1189 + }, + { + "epoch": 0.47542948461845785, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 1190 + }, + { + "epoch": 0.4758290051937675, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1191 + }, + { + "epoch": 0.4762285257690771, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1192 + }, + { + "epoch": 0.4766280463443867, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1193 + }, + { + "epoch": 0.4770275669196964, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1194 + }, + { + "epoch": 0.477427087495006, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1195 + }, + { + "epoch": 0.47782660807031563, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1196 + }, + { + "epoch": 0.47822612864562525, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1197 + }, + { + "epoch": 0.47862564922093487, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1198 + }, + { + "epoch": 0.4790251697962445, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1199 + }, + { + "epoch": 0.4794246903715541, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1200 + }, + { + "epoch": 0.4798242109468638, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1201 + }, + { + "epoch": 0.4802237315221734, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1202 + }, + { + "epoch": 0.480623252097483, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1203 + }, + { + "epoch": 0.48102277267279264, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1204 + }, + { + "epoch": 0.48142229324810226, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1205 + }, + { + "epoch": 0.4818218138234119, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1206 + }, + { + "epoch": 0.48222133439872156, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1207 + }, + { + "epoch": 0.4826208549740312, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1208 + }, + { + "epoch": 0.4830203755493408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1209 + }, + { + "epoch": 0.4834198961246504, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1210 + }, + { + "epoch": 0.48381941669996004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1211 + }, + { + "epoch": 0.48421893727526966, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1212 + }, + { + "epoch": 0.48461845785057933, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1213 + }, + { + "epoch": 0.48501797842588895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1214 + }, + { + "epoch": 0.48541749900119857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1215 + }, + { + "epoch": 0.4858170195765082, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1216 + }, + { + "epoch": 0.4862165401518178, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1217 + }, + { + "epoch": 0.48661606072712743, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 1218 + }, + { + "epoch": 0.48701558130243705, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1219 + }, + { + "epoch": 0.4874151018777467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1220 + }, + { + "epoch": 0.48781462245305635, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1221 + }, + { + "epoch": 0.48821414302836597, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1222 + }, + { + "epoch": 0.4886136636036756, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1223 + }, + { + "epoch": 0.4890131841789852, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1224 + }, + { + "epoch": 0.4894127047542948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1225 + }, + { + "epoch": 0.4898122253296045, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1226 + }, + { + "epoch": 0.4902117459049141, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1227 + }, + { + "epoch": 0.49061126648022374, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1228 + }, + { + "epoch": 0.49101078705553336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1229 + }, + { + "epoch": 0.491410307630843, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 1230 + }, + { + "epoch": 0.4918098282061526, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1231 + }, + { + "epoch": 0.4922093487814622, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1232 + }, + { + "epoch": 0.4926088693567719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1233 + }, + { + "epoch": 0.4930083899320815, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1234 + }, + { + "epoch": 0.49340791050739113, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1235 + }, + { + "epoch": 0.49380743108270075, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1236 + }, + { + "epoch": 0.4942069516580104, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 1237 + }, + { + "epoch": 0.49460647223332, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1238 + }, + { + "epoch": 0.49500599280862967, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1239 + }, + { + "epoch": 0.4954055133839393, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1240 + }, + { + "epoch": 0.4958050339592489, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1241 + }, + { + "epoch": 0.49620455453455853, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1242 + }, + { + "epoch": 0.49660407510986815, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1243 + }, + { + "epoch": 0.49700359568517777, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1244 + }, + { + "epoch": 0.4974031162604874, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1245 + }, + { + "epoch": 0.49780263683579706, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1246 + }, + { + "epoch": 0.4982021574111067, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1247 + }, + { + "epoch": 0.4986016779864163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1248 + }, + { + "epoch": 0.4990011985617259, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1249 + }, + { + "epoch": 0.49940071913703554, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1250 + }, + { + "epoch": 0.49980023971234516, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1251 + }, + { + "epoch": 0.5001997602876548, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1252 + }, + { + "epoch": 0.5005992808629645, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1253 + }, + { + "epoch": 0.500998801438274, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1254 + }, + { + "epoch": 0.5013983220135837, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1255 + }, + { + "epoch": 0.5017978425888934, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1256 + }, + { + "epoch": 0.5021973631642029, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1257 + }, + { + "epoch": 0.5025968837395126, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1258 + }, + { + "epoch": 0.5029964043148222, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1259 + }, + { + "epoch": 0.5033959248901319, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1260 + }, + { + "epoch": 0.5037954454654414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1261 + }, + { + "epoch": 0.5041949660407511, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1262 + }, + { + "epoch": 0.5045944866160608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 1263 + }, + { + "epoch": 0.5049940071913703, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1264 + }, + { + "epoch": 0.50539352776668, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1265 + }, + { + "epoch": 0.5057930483419896, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1266 + }, + { + "epoch": 0.5061925689172992, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1267 + }, + { + "epoch": 0.5065920894926089, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3678, + "step": 1268 + }, + { + "epoch": 0.5069916100679185, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.365, + "step": 1269 + }, + { + "epoch": 0.5073911306432282, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1270 + }, + { + "epoch": 0.5077906512185377, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1271 + }, + { + "epoch": 0.5081901717938474, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1272 + }, + { + "epoch": 0.508589692369157, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1273 + }, + { + "epoch": 0.5089892129444666, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1274 + }, + { + "epoch": 0.5093887335197763, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1275 + }, + { + "epoch": 0.5097882540950859, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1276 + }, + { + "epoch": 0.5101877746703956, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1277 + }, + { + "epoch": 0.5105872952457051, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 1278 + }, + { + "epoch": 0.5109868158210148, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1279 + }, + { + "epoch": 0.5113863363963244, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1280 + }, + { + "epoch": 0.511785856971634, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1281 + }, + { + "epoch": 0.5121853775469437, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1282 + }, + { + "epoch": 0.5125848981222533, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1283 + }, + { + "epoch": 0.512984418697563, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1284 + }, + { + "epoch": 0.5133839392728725, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1285 + }, + { + "epoch": 0.5137834598481822, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.368, + "step": 1286 + }, + { + "epoch": 0.5141829804234919, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1287 + }, + { + "epoch": 0.5145825009988014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1288 + }, + { + "epoch": 0.5149820215741111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1289 + }, + { + "epoch": 0.5153815421494207, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1290 + }, + { + "epoch": 0.5157810627247303, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1291 + }, + { + "epoch": 0.5161805833000399, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 1292 + }, + { + "epoch": 0.5165801038753496, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1293 + }, + { + "epoch": 0.5169796244506593, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1294 + }, + { + "epoch": 0.5173791450259688, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1295 + }, + { + "epoch": 0.5177786656012785, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 1296 + }, + { + "epoch": 0.5181781861765881, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1297 + }, + { + "epoch": 0.5185777067518977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1298 + }, + { + "epoch": 0.5189772273272073, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1299 + }, + { + "epoch": 0.519376747902517, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1300 + }, + { + "epoch": 0.5197762684778267, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1301 + }, + { + "epoch": 0.5201757890531362, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1302 + }, + { + "epoch": 0.5205753096284459, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1303 + }, + { + "epoch": 0.5209748302037555, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1304 + }, + { + "epoch": 0.5213743507790651, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1305 + }, + { + "epoch": 0.5217738713543747, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1306 + }, + { + "epoch": 0.5221733919296844, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1307 + }, + { + "epoch": 0.522572912504994, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1308 + }, + { + "epoch": 0.5229724330803036, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1309 + }, + { + "epoch": 0.5233719536556133, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1310 + }, + { + "epoch": 0.5237714742309229, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1311 + }, + { + "epoch": 0.5241709948062325, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1312 + }, + { + "epoch": 0.5245705153815422, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1313 + }, + { + "epoch": 0.5249700359568518, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1314 + }, + { + "epoch": 0.5253695565321614, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1315 + }, + { + "epoch": 0.525769077107471, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1316 + }, + { + "epoch": 0.5261685976827807, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1317 + }, + { + "epoch": 0.5265681182580902, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1318 + }, + { + "epoch": 0.5269676388333999, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 1319 + }, + { + "epoch": 0.5273671594087096, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1320 + }, + { + "epoch": 0.5277666799840192, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1321 + }, + { + "epoch": 0.5281662005593288, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1322 + }, + { + "epoch": 0.5285657211346384, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1323 + }, + { + "epoch": 0.5289652417099481, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1324 + }, + { + "epoch": 0.5293647622852576, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1325 + }, + { + "epoch": 0.5297642828605673, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1326 + }, + { + "epoch": 0.530163803435877, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1327 + }, + { + "epoch": 0.5305633240111866, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 1328 + }, + { + "epoch": 0.5309628445864962, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1329 + }, + { + "epoch": 0.5313623651618058, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1330 + }, + { + "epoch": 0.5317618857371155, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1331 + }, + { + "epoch": 0.5321614063124251, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1332 + }, + { + "epoch": 0.5325609268877347, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1333 + }, + { + "epoch": 0.5329604474630444, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1334 + }, + { + "epoch": 0.533359968038354, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1335 + }, + { + "epoch": 0.5337594886136636, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1336 + }, + { + "epoch": 0.5341590091889732, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1337 + }, + { + "epoch": 0.5345585297642829, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1338 + }, + { + "epoch": 0.5349580503395925, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1339 + }, + { + "epoch": 0.5353575709149021, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1340 + }, + { + "epoch": 0.5357570914902118, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1341 + }, + { + "epoch": 0.5361566120655213, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1342 + }, + { + "epoch": 0.536556132640831, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1343 + }, + { + "epoch": 0.5369556532161406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1344 + }, + { + "epoch": 0.5373551737914503, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1345 + }, + { + "epoch": 0.5377546943667599, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1346 + }, + { + "epoch": 0.5381542149420695, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1347 + }, + { + "epoch": 0.5385537355173792, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1348 + }, + { + "epoch": 0.5389532560926887, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 1349 + }, + { + "epoch": 0.5393527766679984, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1350 + }, + { + "epoch": 0.539752297243308, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1351 + }, + { + "epoch": 0.5401518178186177, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1352 + }, + { + "epoch": 0.5405513383939273, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1353 + }, + { + "epoch": 0.5409508589692369, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 1354 + }, + { + "epoch": 0.5413503795445466, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1355 + }, + { + "epoch": 0.5417499001198561, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1356 + }, + { + "epoch": 0.5421494206951658, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1357 + }, + { + "epoch": 0.5425489412704755, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1358 + }, + { + "epoch": 0.542948461845785, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1359 + }, + { + "epoch": 0.5433479824210947, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1360 + }, + { + "epoch": 0.5437475029964043, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1361 + }, + { + "epoch": 0.544147023571714, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1362 + }, + { + "epoch": 0.5445465441470235, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1363 + }, + { + "epoch": 0.5449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1364 + }, + { + "epoch": 0.5453455852976429, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1365 + }, + { + "epoch": 0.5457451058729524, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1366 + }, + { + "epoch": 0.5461446264482621, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1367 + }, + { + "epoch": 0.5465441470235717, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 1368 + }, + { + "epoch": 0.5469436675988814, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1369 + }, + { + "epoch": 0.5473431881741909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1370 + }, + { + "epoch": 0.5477427087495006, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1371 + }, + { + "epoch": 0.5481422293248103, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 1372 + }, + { + "epoch": 0.5485417499001198, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1373 + }, + { + "epoch": 0.5489412704754295, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1374 + }, + { + "epoch": 0.5493407910507391, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1375 + }, + { + "epoch": 0.5497403116260487, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1376 + }, + { + "epoch": 0.5501398322013584, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1377 + }, + { + "epoch": 0.550539352776668, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1378 + }, + { + "epoch": 0.5509388733519777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1379 + }, + { + "epoch": 0.5513383939272872, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1380 + }, + { + "epoch": 0.5517379145025969, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1381 + }, + { + "epoch": 0.5521374350779065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1382 + }, + { + "epoch": 0.5525369556532161, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1383 + }, + { + "epoch": 0.5529364762285258, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1384 + }, + { + "epoch": 0.5533359968038354, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1385 + }, + { + "epoch": 0.5537355173791451, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1386 + }, + { + "epoch": 0.5541350379544546, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1387 + }, + { + "epoch": 0.5545345585297643, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1388 + }, + { + "epoch": 0.5549340791050739, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1389 + }, + { + "epoch": 0.5553335996803835, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1390 + }, + { + "epoch": 0.5557331202556932, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1391 + }, + { + "epoch": 0.5561326408310028, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1392 + }, + { + "epoch": 0.5565321614063125, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1393 + }, + { + "epoch": 0.556931681981622, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 1394 + }, + { + "epoch": 0.5573312025569317, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1395 + }, + { + "epoch": 0.5577307231322413, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1396 + }, + { + "epoch": 0.5581302437075509, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1397 + }, + { + "epoch": 0.5585297642828606, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1398 + }, + { + "epoch": 0.5589292848581702, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1399 + }, + { + "epoch": 0.5593288054334798, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1400 + }, + { + "epoch": 0.5597283260087894, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1401 + }, + { + "epoch": 0.5601278465840991, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1402 + }, + { + "epoch": 0.5605273671594088, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1403 + }, + { + "epoch": 0.5609268877347183, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1404 + }, + { + "epoch": 0.561326408310028, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1405 + }, + { + "epoch": 0.5617259288853376, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1406 + }, + { + "epoch": 0.5621254494606472, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 1407 + }, + { + "epoch": 0.5625249700359568, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1408 + }, + { + "epoch": 0.5629244906112665, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1409 + }, + { + "epoch": 0.5633240111865762, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1410 + }, + { + "epoch": 0.5637235317618857, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1411 + }, + { + "epoch": 0.5641230523371954, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1412 + }, + { + "epoch": 0.564522572912505, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1413 + }, + { + "epoch": 0.5649220934878146, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1414 + }, + { + "epoch": 0.5653216140631242, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1415 + }, + { + "epoch": 0.5657211346384339, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1416 + }, + { + "epoch": 0.5661206552137436, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1417 + }, + { + "epoch": 0.5665201757890531, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1418 + }, + { + "epoch": 0.5669196963643628, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1419 + }, + { + "epoch": 0.5673192169396724, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1420 + }, + { + "epoch": 0.567718737514982, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1421 + }, + { + "epoch": 0.5681182580902916, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1422 + }, + { + "epoch": 0.5685177786656013, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1423 + }, + { + "epoch": 0.568917299240911, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1424 + }, + { + "epoch": 0.5693168198162205, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1425 + }, + { + "epoch": 0.5697163403915302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1426 + }, + { + "epoch": 0.5701158609668397, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1427 + }, + { + "epoch": 0.5705153815421494, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1428 + }, + { + "epoch": 0.5709149021174591, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1429 + }, + { + "epoch": 0.5713144226927687, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1430 + }, + { + "epoch": 0.5717139432680783, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1431 + }, + { + "epoch": 0.5721134638433879, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1432 + }, + { + "epoch": 0.5725129844186976, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1433 + }, + { + "epoch": 0.5729125049940071, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1434 + }, + { + "epoch": 0.5733120255693168, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1435 + }, + { + "epoch": 0.5737115461446265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1436 + }, + { + "epoch": 0.5741110667199361, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1437 + }, + { + "epoch": 0.5745105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1438 + }, + { + "epoch": 0.5749101078705553, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1439 + }, + { + "epoch": 0.575309628445865, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1440 + }, + { + "epoch": 0.5757091490211745, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1441 + }, + { + "epoch": 0.5761086695964842, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1442 + }, + { + "epoch": 0.5765081901717939, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1443 + }, + { + "epoch": 0.5769077107471035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1444 + }, + { + "epoch": 0.5773072313224131, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1445 + }, + { + "epoch": 0.5777067518977227, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1446 + }, + { + "epoch": 0.5781062724730324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1447 + }, + { + "epoch": 0.578505793048342, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1448 + }, + { + "epoch": 0.5789053136236516, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1449 + }, + { + "epoch": 0.5793048341989613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 1450 + }, + { + "epoch": 0.5797043547742708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1451 + }, + { + "epoch": 0.5801038753495805, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1452 + }, + { + "epoch": 0.5805033959248901, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1453 + }, + { + "epoch": 0.5809029165001998, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1454 + }, + { + "epoch": 0.5813024370755094, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1455 + }, + { + "epoch": 0.581701957650819, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1456 + }, + { + "epoch": 0.5821014782261287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1457 + }, + { + "epoch": 0.5825009988014382, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1458 + }, + { + "epoch": 0.5829005193767479, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1459 + }, + { + "epoch": 0.5833000399520575, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1460 + }, + { + "epoch": 0.5836995605273672, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1461 + }, + { + "epoch": 0.5840990811026768, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1462 + }, + { + "epoch": 0.5844986016779864, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1463 + }, + { + "epoch": 0.5848981222532961, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1464 + }, + { + "epoch": 0.5852976428286056, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1465 + }, + { + "epoch": 0.5856971634039153, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1466 + }, + { + "epoch": 0.5860966839792249, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1467 + }, + { + "epoch": 0.5864962045545346, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1468 + }, + { + "epoch": 0.5868957251298442, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1469 + }, + { + "epoch": 0.5872952457051538, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1470 + }, + { + "epoch": 0.5876947662804635, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1471 + }, + { + "epoch": 0.588094286855773, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1472 + }, + { + "epoch": 0.5884938074310827, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1473 + }, + { + "epoch": 0.5888933280063924, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1474 + }, + { + "epoch": 0.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1475 + }, + { + "epoch": 0.5896923691570116, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1476 + }, + { + "epoch": 0.5900918897323212, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1477 + }, + { + "epoch": 0.5904914103076309, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1478 + }, + { + "epoch": 0.5908909308829404, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1479 + }, + { + "epoch": 0.5912904514582501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1480 + }, + { + "epoch": 0.5916899720335598, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1481 + }, + { + "epoch": 0.5920894926088693, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1482 + }, + { + "epoch": 0.592489013184179, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1483 + }, + { + "epoch": 0.5928885337594886, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1484 + }, + { + "epoch": 0.5932880543347983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1485 + }, + { + "epoch": 0.5936875749101078, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1486 + }, + { + "epoch": 0.5940870954854175, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1487 + }, + { + "epoch": 0.5944866160607272, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1488 + }, + { + "epoch": 0.5948861366360367, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1489 + }, + { + "epoch": 0.5952856572113464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1490 + }, + { + "epoch": 0.595685177786656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1491 + }, + { + "epoch": 0.5960846983619656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1492 + }, + { + "epoch": 0.5964842189372753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1493 + }, + { + "epoch": 0.5968837395125849, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1494 + }, + { + "epoch": 0.5972832600878946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1495 + }, + { + "epoch": 0.5976827806632041, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1496 + }, + { + "epoch": 0.5980823012385138, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1497 + }, + { + "epoch": 0.5984818218138234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 1498 + }, + { + "epoch": 0.598881342389133, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1499 + }, + { + "epoch": 0.5992808629644427, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1500 + }, + { + "epoch": 0.5996803835397523, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1501 + }, + { + "epoch": 0.600079904115062, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1502 + }, + { + "epoch": 0.6004794246903715, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1503 + }, + { + "epoch": 0.6008789452656812, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1504 + }, + { + "epoch": 0.6012784658409908, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1505 + }, + { + "epoch": 0.6016779864163004, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1506 + }, + { + "epoch": 0.6020775069916101, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1507 + }, + { + "epoch": 0.6024770275669197, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1508 + }, + { + "epoch": 0.6028765481422294, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1509 + }, + { + "epoch": 0.6032760687175389, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1510 + }, + { + "epoch": 0.6036755892928486, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1511 + }, + { + "epoch": 0.6040751098681582, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1512 + }, + { + "epoch": 0.6044746304434678, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3669, + "step": 1513 + }, + { + "epoch": 0.6048741510187775, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1514 + }, + { + "epoch": 0.6052736715940871, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1515 + }, + { + "epoch": 0.6056731921693967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1516 + }, + { + "epoch": 0.6060727127447063, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1517 + }, + { + "epoch": 0.606472233320016, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1518 + }, + { + "epoch": 0.6068717538953257, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1519 + }, + { + "epoch": 0.6072712744706352, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1520 + }, + { + "epoch": 0.6076707950459449, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1521 + }, + { + "epoch": 0.6080703156212545, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1522 + }, + { + "epoch": 0.6084698361965641, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1523 + }, + { + "epoch": 0.6088693567718737, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1524 + }, + { + "epoch": 0.6092688773471834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1525 + }, + { + "epoch": 0.609668397922493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1526 + }, + { + "epoch": 0.6100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1527 + }, + { + "epoch": 0.6104674390731123, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1528 + }, + { + "epoch": 0.6108669596484219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1529 + }, + { + "epoch": 0.6112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1530 + }, + { + "epoch": 0.6116660007990411, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1531 + }, + { + "epoch": 0.6120655213743508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1532 + }, + { + "epoch": 0.6124650419496604, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1533 + }, + { + "epoch": 0.61286456252497, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1534 + }, + { + "epoch": 0.6132640831002797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1535 + }, + { + "epoch": 0.6136636036755893, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1536 + }, + { + "epoch": 0.6140631242508989, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1537 + }, + { + "epoch": 0.6144626448262086, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1538 + }, + { + "epoch": 0.6148621654015182, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1539 + }, + { + "epoch": 0.6152616859768278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1540 + }, + { + "epoch": 0.6156612065521374, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1541 + }, + { + "epoch": 0.6160607271274471, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1542 + }, + { + "epoch": 0.6164602477027566, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1543 + }, + { + "epoch": 0.6168597682780663, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1544 + }, + { + "epoch": 0.617259288853376, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1545 + }, + { + "epoch": 0.6176588094286856, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1546 + }, + { + "epoch": 0.6180583300039952, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1547 + }, + { + "epoch": 0.6184578505793048, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1548 + }, + { + "epoch": 0.6188573711546145, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1549 + }, + { + "epoch": 0.619256891729924, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1550 + }, + { + "epoch": 0.6196564123052337, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1551 + }, + { + "epoch": 0.6200559328805434, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1552 + }, + { + "epoch": 0.620455453455853, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1553 + }, + { + "epoch": 0.6208549740311626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 1554 + }, + { + "epoch": 0.6212544946064722, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1555 + }, + { + "epoch": 0.6216540151817819, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1556 + }, + { + "epoch": 0.6220535357570914, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1557 + }, + { + "epoch": 0.6224530563324011, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1558 + }, + { + "epoch": 0.6228525769077108, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1559 + }, + { + "epoch": 0.6232520974830204, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1560 + }, + { + "epoch": 0.62365161805833, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1561 + }, + { + "epoch": 0.6240511386336396, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1562 + }, + { + "epoch": 0.6244506592089493, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1563 + }, + { + "epoch": 0.6248501797842589, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1564 + }, + { + "epoch": 0.6252497003595685, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1565 + }, + { + "epoch": 0.6256492209348782, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1566 + }, + { + "epoch": 0.6260487415101877, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1567 + }, + { + "epoch": 0.6264482620854974, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1568 + }, + { + "epoch": 0.626847782660807, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1569 + }, + { + "epoch": 0.6272473032361167, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1570 + }, + { + "epoch": 0.6276468238114263, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1571 + }, + { + "epoch": 0.6280463443867359, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1572 + }, + { + "epoch": 0.6284458649620456, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1573 + }, + { + "epoch": 0.6288453855373551, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1574 + }, + { + "epoch": 0.6292449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1575 + }, + { + "epoch": 0.6296444266879744, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1576 + }, + { + "epoch": 0.630043947263284, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1577 + }, + { + "epoch": 0.6304434678385937, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1578 + }, + { + "epoch": 0.6308429884139033, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1579 + }, + { + "epoch": 0.631242508989213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1580 + }, + { + "epoch": 0.6316420295645225, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1581 + }, + { + "epoch": 0.6320415501398322, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1582 + }, + { + "epoch": 0.6324410707151419, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1583 + }, + { + "epoch": 0.6328405912904514, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1584 + }, + { + "epoch": 0.6332401118657611, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1585 + }, + { + "epoch": 0.6336396324410707, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1586 + }, + { + "epoch": 0.6340391530163804, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1587 + }, + { + "epoch": 0.6344386735916899, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1588 + }, + { + "epoch": 0.6348381941669996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1589 + }, + { + "epoch": 0.6352377147423093, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1590 + }, + { + "epoch": 0.6356372353176188, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1591 + }, + { + "epoch": 0.6360367558929285, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1592 + }, + { + "epoch": 0.6364362764682381, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1593 + }, + { + "epoch": 0.6368357970435478, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1594 + }, + { + "epoch": 0.6372353176188573, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1595 + }, + { + "epoch": 0.637634838194167, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1596 + }, + { + "epoch": 0.6380343587694767, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1597 + }, + { + "epoch": 0.6384338793447862, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1598 + }, + { + "epoch": 0.6388333999200959, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1599 + }, + { + "epoch": 0.6392329204954055, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1600 + }, + { + "epoch": 0.6396324410707152, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1601 + }, + { + "epoch": 0.6400319616460247, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1602 + }, + { + "epoch": 0.6404314822213344, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1603 + }, + { + "epoch": 0.6408310027966441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1604 + }, + { + "epoch": 0.6412305233719536, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1605 + }, + { + "epoch": 0.6416300439472633, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1606 + }, + { + "epoch": 0.6420295645225729, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1607 + }, + { + "epoch": 0.6424290850978825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1608 + }, + { + "epoch": 0.6428286056731922, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1609 + }, + { + "epoch": 0.6432281262485018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1610 + }, + { + "epoch": 0.6436276468238115, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1611 + }, + { + "epoch": 0.644027167399121, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1612 + }, + { + "epoch": 0.6444266879744307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1613 + }, + { + "epoch": 0.6448262085497403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1614 + }, + { + "epoch": 0.6452257291250499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1615 + }, + { + "epoch": 0.6456252497003596, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 1616 + }, + { + "epoch": 0.6460247702756692, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1617 + }, + { + "epoch": 0.6464242908509789, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 1618 + }, + { + "epoch": 0.6468238114262884, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1619 + }, + { + "epoch": 0.6472233320015981, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1620 + }, + { + "epoch": 0.6476228525769077, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1621 + }, + { + "epoch": 0.6480223731522173, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1622 + }, + { + "epoch": 0.648421893727527, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1623 + }, + { + "epoch": 0.6488214143028366, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1624 + }, + { + "epoch": 0.6492209348781462, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1625 + }, + { + "epoch": 0.6496204554534558, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1626 + }, + { + "epoch": 0.6500199760287655, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1627 + }, + { + "epoch": 0.6504194966040752, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1628 + }, + { + "epoch": 0.6508190171793847, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1629 + }, + { + "epoch": 0.6512185377546944, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1630 + }, + { + "epoch": 0.651618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1631 + }, + { + "epoch": 0.6520175789053136, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1632 + }, + { + "epoch": 0.6524170994806232, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1633 + }, + { + "epoch": 0.6528166200559329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1634 + }, + { + "epoch": 0.6532161406312426, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1635 + }, + { + "epoch": 0.6536156612065521, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 1636 + }, + { + "epoch": 0.6540151817818618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1637 + }, + { + "epoch": 0.6544147023571714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1638 + }, + { + "epoch": 0.654814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1639 + }, + { + "epoch": 0.6552137435077906, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1640 + }, + { + "epoch": 0.6556132640831003, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1641 + }, + { + "epoch": 0.65601278465841, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1642 + }, + { + "epoch": 0.6564123052337195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1643 + }, + { + "epoch": 0.6568118258090292, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1644 + }, + { + "epoch": 0.6572113463843388, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1645 + }, + { + "epoch": 0.6576108669596484, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1646 + }, + { + "epoch": 0.658010387534958, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1647 + }, + { + "epoch": 0.6584099081102677, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1648 + }, + { + "epoch": 0.6588094286855773, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1649 + }, + { + "epoch": 0.6592089492608869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 1650 + }, + { + "epoch": 0.6596084698361966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 1651 + }, + { + "epoch": 0.6600079904115062, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1652 + }, + { + "epoch": 0.6604075109868158, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1653 + }, + { + "epoch": 0.6608070315621255, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1654 + }, + { + "epoch": 0.6612065521374351, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1655 + }, + { + "epoch": 0.6616060727127447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1656 + }, + { + "epoch": 0.6620055932880543, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1657 + }, + { + "epoch": 0.662405113863364, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1658 + }, + { + "epoch": 0.6628046344386735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1659 + }, + { + "epoch": 0.6632041550139832, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1660 + }, + { + "epoch": 0.6636036755892929, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1661 + }, + { + "epoch": 0.6640031961646025, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1662 + }, + { + "epoch": 0.6644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1663 + }, + { + "epoch": 0.6648022373152217, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1664 + }, + { + "epoch": 0.6652017578905314, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1665 + }, + { + "epoch": 0.6656012784658409, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1666 + }, + { + "epoch": 0.6660007990411506, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1667 + }, + { + "epoch": 0.6664003196164603, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1668 + }, + { + "epoch": 0.6667998401917699, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1669 + }, + { + "epoch": 0.6671993607670795, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1670 + }, + { + "epoch": 0.6675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1671 + }, + { + "epoch": 0.6679984019176988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1672 + }, + { + "epoch": 0.6683979224930084, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1673 + }, + { + "epoch": 0.668797443068318, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1674 + }, + { + "epoch": 0.6691969636436277, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1675 + }, + { + "epoch": 0.6695964842189372, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1676 + }, + { + "epoch": 0.6699960047942469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 1677 + }, + { + "epoch": 0.6703955253695565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1678 + }, + { + "epoch": 0.6707950459448662, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1679 + }, + { + "epoch": 0.6711945665201758, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1680 + }, + { + "epoch": 0.6715940870954854, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1681 + }, + { + "epoch": 0.6719936076707951, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1682 + }, + { + "epoch": 0.6723931282461046, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1683 + }, + { + "epoch": 0.6727926488214143, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1684 + }, + { + "epoch": 0.6731921693967239, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1685 + }, + { + "epoch": 0.6735916899720336, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1686 + }, + { + "epoch": 0.6739912105473432, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1687 + }, + { + "epoch": 0.6743907311226528, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1688 + }, + { + "epoch": 0.6747902516979625, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1689 + }, + { + "epoch": 0.675189772273272, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1690 + }, + { + "epoch": 0.6755892928485817, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1691 + }, + { + "epoch": 0.6759888134238913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1692 + }, + { + "epoch": 0.676388333999201, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1693 + }, + { + "epoch": 0.6767878545745106, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1694 + }, + { + "epoch": 0.6771873751498202, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1695 + }, + { + "epoch": 0.6775868957251299, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1696 + }, + { + "epoch": 0.6779864163004394, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1697 + }, + { + "epoch": 0.6783859368757491, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1698 + }, + { + "epoch": 0.6787854574510588, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1699 + }, + { + "epoch": 0.6791849780263683, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 1700 + }, + { + "epoch": 0.679584498601678, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1701 + }, + { + "epoch": 0.6799840191769876, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1702 + }, + { + "epoch": 0.6803835397522973, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1703 + }, + { + "epoch": 0.6807830603276068, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1704 + }, + { + "epoch": 0.6811825809029165, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1705 + }, + { + "epoch": 0.6815821014782262, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1706 + }, + { + "epoch": 0.6819816220535357, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1707 + }, + { + "epoch": 0.6823811426288454, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1708 + }, + { + "epoch": 0.682780663204155, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1709 + }, + { + "epoch": 0.6831801837794647, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1710 + }, + { + "epoch": 0.6835797043547742, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1711 + }, + { + "epoch": 0.6839792249300839, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1712 + }, + { + "epoch": 0.6843787455053936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1713 + }, + { + "epoch": 0.6847782660807031, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1714 + }, + { + "epoch": 0.6851777866560128, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1715 + }, + { + "epoch": 0.6855773072313224, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1716 + }, + { + "epoch": 0.685976827806632, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1717 + }, + { + "epoch": 0.6863763483819417, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1718 + }, + { + "epoch": 0.6867758689572513, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1719 + }, + { + "epoch": 0.687175389532561, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1720 + }, + { + "epoch": 0.6875749101078705, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1721 + }, + { + "epoch": 0.6879744306831802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1722 + }, + { + "epoch": 0.6883739512584898, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1723 + }, + { + "epoch": 0.6887734718337994, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1724 + }, + { + "epoch": 0.6891729924091091, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1725 + }, + { + "epoch": 0.6895725129844187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1726 + }, + { + "epoch": 0.6899720335597284, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1727 + }, + { + "epoch": 0.6903715541350379, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1728 + }, + { + "epoch": 0.6907710747103476, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1729 + }, + { + "epoch": 0.6911705952856572, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1730 + }, + { + "epoch": 0.6915701158609668, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1731 + }, + { + "epoch": 0.6919696364362765, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1732 + }, + { + "epoch": 0.6923691570115861, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1733 + }, + { + "epoch": 0.6927686775868958, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1734 + }, + { + "epoch": 0.6931681981622053, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1735 + }, + { + "epoch": 0.693567718737515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1736 + }, + { + "epoch": 0.6939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1737 + }, + { + "epoch": 0.6943667598881342, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1738 + }, + { + "epoch": 0.6947662804634439, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 1739 + }, + { + "epoch": 0.6951658010387535, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1740 + }, + { + "epoch": 0.6955653216140631, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1741 + }, + { + "epoch": 0.6959648421893727, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 1742 + }, + { + "epoch": 0.6963643627646824, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1743 + }, + { + "epoch": 0.6967638833399921, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1744 + }, + { + "epoch": 0.6971634039153016, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1745 + }, + { + "epoch": 0.6975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1746 + }, + { + "epoch": 0.6979624450659209, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1747 + }, + { + "epoch": 0.6983619656412305, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1748 + }, + { + "epoch": 0.6987614862165401, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1749 + }, + { + "epoch": 0.6991610067918498, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1750 + }, + { + "epoch": 0.6995605273671595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 1751 + }, + { + "epoch": 0.699960047942469, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1752 + }, + { + "epoch": 0.7003595685177787, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1753 + }, + { + "epoch": 0.7007590890930883, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1754 + }, + { + "epoch": 0.7011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1755 + }, + { + "epoch": 0.7015581302437075, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1756 + }, + { + "epoch": 0.7019576508190172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1757 + }, + { + "epoch": 0.7023571713943269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1758 + }, + { + "epoch": 0.7027566919696364, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1759 + }, + { + "epoch": 0.7031562125449461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1760 + }, + { + "epoch": 0.7035557331202557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1761 + }, + { + "epoch": 0.7039552536955653, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1762 + }, + { + "epoch": 0.7043547742708749, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1763 + }, + { + "epoch": 0.7047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1764 + }, + { + "epoch": 0.7051538154214942, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1765 + }, + { + "epoch": 0.7055533359968038, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1766 + }, + { + "epoch": 0.7059528565721135, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1767 + }, + { + "epoch": 0.706352377147423, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1768 + }, + { + "epoch": 0.7067518977227327, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1769 + }, + { + "epoch": 0.7071514182980424, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 1770 + }, + { + "epoch": 0.707550938873352, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1771 + }, + { + "epoch": 0.7079504594486616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1772 + }, + { + "epoch": 0.7083499800239712, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1773 + }, + { + "epoch": 0.7087495005992809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1774 + }, + { + "epoch": 0.7091490211745904, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1775 + }, + { + "epoch": 0.7095485417499001, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 1776 + }, + { + "epoch": 0.7099480623252098, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1777 + }, + { + "epoch": 0.7103475829005194, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1778 + }, + { + "epoch": 0.710747103475829, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1779 + }, + { + "epoch": 0.7111466240511386, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1780 + }, + { + "epoch": 0.7115461446264483, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1781 + }, + { + "epoch": 0.7119456652017578, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1782 + }, + { + "epoch": 0.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1783 + }, + { + "epoch": 0.7127447063523772, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1784 + }, + { + "epoch": 0.7131442269276868, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1785 + }, + { + "epoch": 0.7135437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1786 + }, + { + "epoch": 0.713943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1787 + }, + { + "epoch": 0.7143427886536157, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1788 + }, + { + "epoch": 0.7147423092289253, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1789 + }, + { + "epoch": 0.7151418298042349, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1790 + }, + { + "epoch": 0.7155413503795446, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1791 + }, + { + "epoch": 0.7159408709548541, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1792 + }, + { + "epoch": 0.7163403915301638, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1793 + }, + { + "epoch": 0.7167399121054734, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1794 + }, + { + "epoch": 0.7171394326807831, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1795 + }, + { + "epoch": 0.7175389532560927, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1796 + }, + { + "epoch": 0.7179384738314023, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1797 + }, + { + "epoch": 0.718337994406712, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1798 + }, + { + "epoch": 0.7187375149820215, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1799 + }, + { + "epoch": 0.7191370355573312, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 1800 + }, + { + "epoch": 0.7195365561326408, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1801 + }, + { + "epoch": 0.7199360767079505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1802 + }, + { + "epoch": 0.7203355972832601, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 1803 + }, + { + "epoch": 0.7207351178585697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1804 + }, + { + "epoch": 0.7211346384338794, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1805 + }, + { + "epoch": 0.7215341590091889, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1806 + }, + { + "epoch": 0.7219336795844986, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1807 + }, + { + "epoch": 0.7223332001598082, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1808 + }, + { + "epoch": 0.7227327207351179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1809 + }, + { + "epoch": 0.7231322413104275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1810 + }, + { + "epoch": 0.7235317618857371, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1811 + }, + { + "epoch": 0.7239312824610468, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1812 + }, + { + "epoch": 0.7243308030363563, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1813 + }, + { + "epoch": 0.724730323611666, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1814 + }, + { + "epoch": 0.7251298441869757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1815 + }, + { + "epoch": 0.7255293647622852, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1816 + }, + { + "epoch": 0.7259288853375949, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1817 + }, + { + "epoch": 0.7263284059129045, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1818 + }, + { + "epoch": 0.7267279264882142, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1819 + }, + { + "epoch": 0.7271274470635237, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1820 + }, + { + "epoch": 0.7275269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1821 + }, + { + "epoch": 0.7279264882141431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1822 + }, + { + "epoch": 0.7283260087894526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1823 + }, + { + "epoch": 0.7287255293647623, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1824 + }, + { + "epoch": 0.7291250499400719, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 1825 + }, + { + "epoch": 0.7295245705153816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1826 + }, + { + "epoch": 0.7299240910906911, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1827 + }, + { + "epoch": 0.7303236116660008, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1828 + }, + { + "epoch": 0.7307231322413105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 1829 + }, + { + "epoch": 0.73112265281662, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1830 + }, + { + "epoch": 0.7315221733919297, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1831 + }, + { + "epoch": 0.7319216939672393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1832 + }, + { + "epoch": 0.732321214542549, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1833 + }, + { + "epoch": 0.7327207351178586, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1834 + }, + { + "epoch": 0.7331202556931682, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1835 + }, + { + "epoch": 0.7335197762684779, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1836 + }, + { + "epoch": 0.7339192968437874, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1837 + }, + { + "epoch": 0.7343188174190971, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1838 + }, + { + "epoch": 0.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1839 + }, + { + "epoch": 0.7351178585697163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1840 + }, + { + "epoch": 0.735517379145026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1841 + }, + { + "epoch": 0.7359168997203356, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1842 + }, + { + "epoch": 0.7363164202956453, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 1843 + }, + { + "epoch": 0.7367159408709548, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1844 + }, + { + "epoch": 0.7371154614462645, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1845 + }, + { + "epoch": 0.7375149820215741, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1846 + }, + { + "epoch": 0.7379145025968837, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1847 + }, + { + "epoch": 0.7383140231721934, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1848 + }, + { + "epoch": 0.738713543747503, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 1849 + }, + { + "epoch": 0.7391130643228127, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1850 + }, + { + "epoch": 0.7395125848981222, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1851 + }, + { + "epoch": 0.7399121054734319, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1852 + }, + { + "epoch": 0.7403116260487415, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1853 + }, + { + "epoch": 0.7407111466240511, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 1854 + }, + { + "epoch": 0.7411106671993608, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1855 + }, + { + "epoch": 0.7415101877746704, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1856 + }, + { + "epoch": 0.74190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1857 + }, + { + "epoch": 0.7423092289252896, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1858 + }, + { + "epoch": 0.7427087495005993, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1859 + }, + { + "epoch": 0.743108270075909, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1860 + }, + { + "epoch": 0.7435077906512185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1861 + }, + { + "epoch": 0.7439073112265282, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1862 + }, + { + "epoch": 0.7443068318018378, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 1863 + }, + { + "epoch": 0.7447063523771474, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1864 + }, + { + "epoch": 0.745105872952457, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1865 + }, + { + "epoch": 0.7455053935277667, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1866 + }, + { + "epoch": 0.7459049141030764, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1867 + }, + { + "epoch": 0.7463044346783859, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1868 + }, + { + "epoch": 0.7467039552536956, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1869 + }, + { + "epoch": 0.7471034758290052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 1870 + }, + { + "epoch": 0.7475029964043148, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1871 + }, + { + "epoch": 0.7479025169796244, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1872 + }, + { + "epoch": 0.7483020375549341, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1873 + }, + { + "epoch": 0.7487015581302437, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1874 + }, + { + "epoch": 0.7491010787055533, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1875 + }, + { + "epoch": 0.749500599280863, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1876 + }, + { + "epoch": 0.7499001198561726, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1877 + }, + { + "epoch": 0.7502996404314822, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1878 + }, + { + "epoch": 0.7506991610067919, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1879 + }, + { + "epoch": 0.7510986815821015, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1880 + }, + { + "epoch": 0.7514982021574111, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 1881 + }, + { + "epoch": 0.7518977227327207, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1882 + }, + { + "epoch": 0.7522972433080304, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1883 + }, + { + "epoch": 0.75269676388334, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1884 + }, + { + "epoch": 0.7530962844586496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1885 + }, + { + "epoch": 0.7534958050339593, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1886 + }, + { + "epoch": 0.7538953256092689, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1887 + }, + { + "epoch": 0.7542948461845785, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1888 + }, + { + "epoch": 0.7546943667598881, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1889 + }, + { + "epoch": 0.7550938873351978, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 1890 + }, + { + "epoch": 0.7554934079105073, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1891 + }, + { + "epoch": 0.755892928485817, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1892 + }, + { + "epoch": 0.7562924490611267, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1893 + }, + { + "epoch": 0.7566919696364363, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1894 + }, + { + "epoch": 0.7570914902117459, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1895 + }, + { + "epoch": 0.7574910107870555, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1896 + }, + { + "epoch": 0.7578905313623652, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1897 + }, + { + "epoch": 0.7582900519376747, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 1898 + }, + { + "epoch": 0.7586895725129844, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1899 + }, + { + "epoch": 0.7590890930882941, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1900 + }, + { + "epoch": 0.7594886136636037, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1901 + }, + { + "epoch": 0.7598881342389133, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1902 + }, + { + "epoch": 0.7602876548142229, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1903 + }, + { + "epoch": 0.7606871753895326, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1904 + }, + { + "epoch": 0.7610866959648422, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 1905 + }, + { + "epoch": 0.7614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1906 + }, + { + "epoch": 0.7618857371154615, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1907 + }, + { + "epoch": 0.762285257690771, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1908 + }, + { + "epoch": 0.7626847782660807, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1909 + }, + { + "epoch": 0.7630842988413903, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1910 + }, + { + "epoch": 0.7634838194167, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1911 + }, + { + "epoch": 0.7638833399920096, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1912 + }, + { + "epoch": 0.7642828605673192, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1913 + }, + { + "epoch": 0.7646823811426289, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1914 + }, + { + "epoch": 0.7650819017179384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1915 + }, + { + "epoch": 0.7654814222932481, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1916 + }, + { + "epoch": 0.7658809428685577, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1917 + }, + { + "epoch": 0.7662804634438674, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1918 + }, + { + "epoch": 0.766679984019177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1919 + }, + { + "epoch": 0.7670795045944866, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1920 + }, + { + "epoch": 0.7674790251697963, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1921 + }, + { + "epoch": 0.7678785457451058, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1922 + }, + { + "epoch": 0.7682780663204155, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1923 + }, + { + "epoch": 0.7686775868957252, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1924 + }, + { + "epoch": 0.7690771074710347, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1925 + }, + { + "epoch": 0.7694766280463444, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 1926 + }, + { + "epoch": 0.769876148621654, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1927 + }, + { + "epoch": 0.7702756691969637, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1928 + }, + { + "epoch": 0.7706751897722732, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1929 + }, + { + "epoch": 0.7710747103475829, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1930 + }, + { + "epoch": 0.7714742309228926, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1931 + }, + { + "epoch": 0.7718737514982021, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1932 + }, + { + "epoch": 0.7722732720735118, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1933 + }, + { + "epoch": 0.7726727926488214, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1934 + }, + { + "epoch": 0.7730723132241311, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1935 + }, + { + "epoch": 0.7734718337994406, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1936 + }, + { + "epoch": 0.7738713543747503, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1937 + }, + { + "epoch": 0.77427087495006, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1938 + }, + { + "epoch": 0.7746703955253695, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 1939 + }, + { + "epoch": 0.7750699161006792, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1940 + }, + { + "epoch": 0.7754694366759888, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1941 + }, + { + "epoch": 0.7758689572512985, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 1942 + }, + { + "epoch": 0.776268477826608, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 1943 + }, + { + "epoch": 0.7766679984019177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1944 + }, + { + "epoch": 0.7770675189772274, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 1945 + }, + { + "epoch": 0.7774670395525369, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1946 + }, + { + "epoch": 0.7778665601278466, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1947 + }, + { + "epoch": 0.7782660807031562, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1948 + }, + { + "epoch": 0.7786656012784658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1949 + }, + { + "epoch": 0.7790651218537755, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 1950 + }, + { + "epoch": 0.7794646424290851, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1951 + }, + { + "epoch": 0.7798641630043948, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1952 + }, + { + "epoch": 0.7802636835797043, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 1953 + }, + { + "epoch": 0.780663204155014, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1954 + }, + { + "epoch": 0.7810627247303236, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1955 + }, + { + "epoch": 0.7814622453056332, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1956 + }, + { + "epoch": 0.7818617658809429, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1957 + }, + { + "epoch": 0.7822612864562525, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1958 + }, + { + "epoch": 0.7826608070315622, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1959 + }, + { + "epoch": 0.7830603276068717, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1960 + }, + { + "epoch": 0.7834598481821814, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1961 + }, + { + "epoch": 0.783859368757491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1962 + }, + { + "epoch": 0.7842588893328006, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1963 + }, + { + "epoch": 0.7846584099081103, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1964 + }, + { + "epoch": 0.7850579304834199, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1965 + }, + { + "epoch": 0.7854574510587295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1966 + }, + { + "epoch": 0.7858569716340391, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 1967 + }, + { + "epoch": 0.7862564922093488, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1968 + }, + { + "epoch": 0.7866560127846585, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 1969 + }, + { + "epoch": 0.787055533359968, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1970 + }, + { + "epoch": 0.7874550539352777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 1971 + }, + { + "epoch": 0.7878545745105873, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1972 + }, + { + "epoch": 0.7882540950858969, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1973 + }, + { + "epoch": 0.7886536156612065, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1974 + }, + { + "epoch": 0.7890531362365162, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1975 + }, + { + "epoch": 0.7894526568118259, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1976 + }, + { + "epoch": 0.7898521773871354, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1977 + }, + { + "epoch": 0.7902516979624451, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1978 + }, + { + "epoch": 0.7906512185377547, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 1979 + }, + { + "epoch": 0.7910507391130643, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1980 + }, + { + "epoch": 0.7914502596883739, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1981 + }, + { + "epoch": 0.7918497802636836, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 1982 + }, + { + "epoch": 0.7922493008389933, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 1983 + }, + { + "epoch": 0.7926488214143028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1984 + }, + { + "epoch": 0.7930483419896125, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1985 + }, + { + "epoch": 0.7934478625649221, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1986 + }, + { + "epoch": 0.7938473831402317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1987 + }, + { + "epoch": 0.7942469037155413, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1988 + }, + { + "epoch": 0.794646424290851, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 1989 + }, + { + "epoch": 0.7950459448661606, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1990 + }, + { + "epoch": 0.7954454654414702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 1991 + }, + { + "epoch": 0.7958449860167799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1992 + }, + { + "epoch": 0.7962445065920895, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1993 + }, + { + "epoch": 0.7966440271673991, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1994 + }, + { + "epoch": 0.7970435477427088, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1995 + }, + { + "epoch": 0.7974430683180184, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 1996 + }, + { + "epoch": 0.797842588893328, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1997 + }, + { + "epoch": 0.7982421094686376, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1998 + }, + { + "epoch": 0.7986416300439473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1999 + }, + { + "epoch": 0.7990411506192568, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2000 + }, + { + "epoch": 0.7994406711945665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2001 + }, + { + "epoch": 0.7998401917698762, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2002 + }, + { + "epoch": 0.8002397123451858, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2003 + }, + { + "epoch": 0.8006392329204954, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2004 + }, + { + "epoch": 0.801038753495805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 2005 + }, + { + "epoch": 0.8014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2006 + }, + { + "epoch": 0.8018377946464242, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2007 + }, + { + "epoch": 0.8022373152217339, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2008 + }, + { + "epoch": 0.8026368357970436, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2009 + }, + { + "epoch": 0.8030363563723532, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 2010 + }, + { + "epoch": 0.8034358769476628, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2011 + }, + { + "epoch": 0.8038353975229724, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2012 + }, + { + "epoch": 0.8042349180982821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2013 + }, + { + "epoch": 0.8046344386735917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 2014 + }, + { + "epoch": 0.8050339592489013, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 2015 + }, + { + "epoch": 0.805433479824211, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2016 + }, + { + "epoch": 0.8058330003995205, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2017 + }, + { + "epoch": 0.8062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2018 + }, + { + "epoch": 0.8066320415501398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2019 + }, + { + "epoch": 0.8070315621254495, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2020 + }, + { + "epoch": 0.8074310827007591, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2021 + }, + { + "epoch": 0.8078306032760687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2022 + }, + { + "epoch": 0.8082301238513784, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2023 + }, + { + "epoch": 0.8086296444266879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2024 + }, + { + "epoch": 0.8090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2025 + }, + { + "epoch": 0.8094286855773072, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 2026 + }, + { + "epoch": 0.8098282061526169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2027 + }, + { + "epoch": 0.8102277267279265, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2028 + }, + { + "epoch": 0.8106272473032361, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2029 + }, + { + "epoch": 0.8110267678785458, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2030 + }, + { + "epoch": 0.8114262884538553, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2031 + }, + { + "epoch": 0.811825809029165, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 2032 + }, + { + "epoch": 0.8122253296044746, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2033 + }, + { + "epoch": 0.8126248501797843, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2034 + }, + { + "epoch": 0.8130243707550939, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 2035 + }, + { + "epoch": 0.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2036 + }, + { + "epoch": 0.8138234119057132, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 2037 + }, + { + "epoch": 0.8142229324810227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2038 + }, + { + "epoch": 0.8146224530563324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2039 + }, + { + "epoch": 0.8150219736316421, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2040 + }, + { + "epoch": 0.8154214942069516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2041 + }, + { + "epoch": 0.8158210147822613, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2042 + }, + { + "epoch": 0.8162205353575709, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2043 + }, + { + "epoch": 0.8166200559328806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2044 + }, + { + "epoch": 0.8170195765081901, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2045 + }, + { + "epoch": 0.8174190970834998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2046 + }, + { + "epoch": 0.8178186176588095, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2047 + }, + { + "epoch": 0.818218138234119, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2048 + }, + { + "epoch": 0.8186176588094287, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 2049 + }, + { + "epoch": 0.8190171793847383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2050 + }, + { + "epoch": 0.819416699960048, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2051 + }, + { + "epoch": 0.8198162205353575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2052 + }, + { + "epoch": 0.8202157411106672, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2053 + }, + { + "epoch": 0.8206152616859769, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2054 + }, + { + "epoch": 0.8210147822612864, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2055 + }, + { + "epoch": 0.8214143028365961, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2056 + }, + { + "epoch": 0.8218138234119057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2057 + }, + { + "epoch": 0.8222133439872154, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2058 + }, + { + "epoch": 0.8226128645625249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2059 + }, + { + "epoch": 0.8230123851378346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2060 + }, + { + "epoch": 0.8234119057131443, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2061 + }, + { + "epoch": 0.8238114262884538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2062 + }, + { + "epoch": 0.8242109468637635, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2063 + }, + { + "epoch": 0.8246104674390731, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2064 + }, + { + "epoch": 0.8250099880143827, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2065 + }, + { + "epoch": 0.8254095085896924, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2066 + }, + { + "epoch": 0.825809029165002, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2067 + }, + { + "epoch": 0.8262085497403117, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2068 + }, + { + "epoch": 0.8266080703156212, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2069 + }, + { + "epoch": 0.8270075908909309, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2070 + }, + { + "epoch": 0.8274071114662405, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2071 + }, + { + "epoch": 0.8278066320415501, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2072 + }, + { + "epoch": 0.8282061526168598, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 2073 + }, + { + "epoch": 0.8286056731921694, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 2074 + }, + { + "epoch": 0.829005193767479, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2075 + }, + { + "epoch": 0.8294047143427886, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2076 + }, + { + "epoch": 0.8298042349180983, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2077 + }, + { + "epoch": 0.8302037554934079, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2078 + }, + { + "epoch": 0.8306032760687175, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2079 + }, + { + "epoch": 0.8310027966440272, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2080 + }, + { + "epoch": 0.8314023172193368, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2081 + }, + { + "epoch": 0.8318018377946464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2082 + }, + { + "epoch": 0.832201358369956, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2083 + }, + { + "epoch": 0.8326008789452657, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2084 + }, + { + "epoch": 0.8330003995205754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2085 + }, + { + "epoch": 0.8333999200958849, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2086 + }, + { + "epoch": 0.8337994406711946, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2087 + }, + { + "epoch": 0.8341989612465042, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2088 + }, + { + "epoch": 0.8345984818218138, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2089 + }, + { + "epoch": 0.8349980023971234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 2090 + }, + { + "epoch": 0.8353975229724331, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 2091 + }, + { + "epoch": 0.8357970435477428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2092 + }, + { + "epoch": 0.8361965641230523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2093 + }, + { + "epoch": 0.836596084698362, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2094 + }, + { + "epoch": 0.8369956052736716, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2095 + }, + { + "epoch": 0.8373951258489812, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2096 + }, + { + "epoch": 0.8377946464242908, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2097 + }, + { + "epoch": 0.8381941669996005, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2098 + }, + { + "epoch": 0.8385936875749102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2099 + }, + { + "epoch": 0.8389932081502197, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2100 + }, + { + "epoch": 0.8393927287255294, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 2101 + }, + { + "epoch": 0.839792249300839, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2102 + }, + { + "epoch": 0.8401917698761486, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2103 + }, + { + "epoch": 0.8405912904514582, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 2104 + }, + { + "epoch": 0.8409908110267679, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2105 + }, + { + "epoch": 0.8413903316020775, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2106 + }, + { + "epoch": 0.8417898521773871, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2107 + }, + { + "epoch": 0.8421893727526968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2108 + }, + { + "epoch": 0.8425888933280063, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2109 + }, + { + "epoch": 0.842988413903316, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2110 + }, + { + "epoch": 0.8433879344786257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2111 + }, + { + "epoch": 0.8437874550539353, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2112 + }, + { + "epoch": 0.8441869756292449, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2113 + }, + { + "epoch": 0.8445864962045545, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2114 + }, + { + "epoch": 0.8449860167798642, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2115 + }, + { + "epoch": 0.8453855373551737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2116 + }, + { + "epoch": 0.8457850579304834, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2117 + }, + { + "epoch": 0.8461845785057931, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2118 + }, + { + "epoch": 0.8465840990811027, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2119 + }, + { + "epoch": 0.8469836196564123, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2120 + }, + { + "epoch": 0.8473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2121 + }, + { + "epoch": 0.8477826608070316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2122 + }, + { + "epoch": 0.8481821813823411, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 2123 + }, + { + "epoch": 0.8485817019576508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2124 + }, + { + "epoch": 0.8489812225329605, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 2125 + }, + { + "epoch": 0.84938074310827, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2126 + }, + { + "epoch": 0.8497802636835797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2127 + }, + { + "epoch": 0.8501797842588893, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2128 + }, + { + "epoch": 0.850579304834199, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2129 + }, + { + "epoch": 0.8509788254095086, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2130 + }, + { + "epoch": 0.8513783459848182, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2131 + }, + { + "epoch": 0.8517778665601279, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2132 + }, + { + "epoch": 0.8521773871354374, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 0.8525769077107471, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2134 + }, + { + "epoch": 0.8529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2135 + }, + { + "epoch": 0.8533759488613664, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2136 + }, + { + "epoch": 0.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2137 + }, + { + "epoch": 0.8541749900119856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2138 + }, + { + "epoch": 0.8545745105872953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2139 + }, + { + "epoch": 0.8549740311626048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2140 + }, + { + "epoch": 0.8553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2141 + }, + { + "epoch": 0.8557730723132241, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2142 + }, + { + "epoch": 0.8561725928885338, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 2143 + }, + { + "epoch": 0.8565721134638434, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2144 + }, + { + "epoch": 0.856971634039153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2145 + }, + { + "epoch": 0.8573711546144627, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2146 + }, + { + "epoch": 0.8577706751897722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2147 + }, + { + "epoch": 0.8581701957650819, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 2148 + }, + { + "epoch": 0.8585697163403915, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2149 + }, + { + "epoch": 0.8589692369157012, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2150 + }, + { + "epoch": 0.8593687574910108, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2151 + }, + { + "epoch": 0.8597682780663204, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2152 + }, + { + "epoch": 0.8601677986416301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2153 + }, + { + "epoch": 0.8605673192169396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 2154 + }, + { + "epoch": 0.8609668397922493, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2155 + }, + { + "epoch": 0.861366360367559, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2156 + }, + { + "epoch": 0.8617658809428685, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2157 + }, + { + "epoch": 0.8621654015181782, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2158 + }, + { + "epoch": 0.8625649220934878, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2159 + }, + { + "epoch": 0.8629644426687975, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2160 + }, + { + "epoch": 0.863363963244107, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2161 + }, + { + "epoch": 0.8637634838194167, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2162 + }, + { + "epoch": 0.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2163 + }, + { + "epoch": 0.8645625249700359, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2164 + }, + { + "epoch": 0.8649620455453456, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2165 + }, + { + "epoch": 0.8653615661206552, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2166 + }, + { + "epoch": 0.8657610866959649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2167 + }, + { + "epoch": 0.8661606072712744, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2168 + }, + { + "epoch": 0.8665601278465841, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2169 + }, + { + "epoch": 0.8669596484218938, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2170 + }, + { + "epoch": 0.8673591689972033, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2171 + }, + { + "epoch": 0.867758689572513, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2172 + }, + { + "epoch": 0.8681582101478226, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2173 + }, + { + "epoch": 0.8685577307231322, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2174 + }, + { + "epoch": 0.8689572512984419, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2175 + }, + { + "epoch": 0.8693567718737515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2176 + }, + { + "epoch": 0.8697562924490612, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2177 + }, + { + "epoch": 0.8701558130243707, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2178 + }, + { + "epoch": 0.8705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 2179 + }, + { + "epoch": 0.87095485417499, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2180 + }, + { + "epoch": 0.8713543747502996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2181 + }, + { + "epoch": 0.8717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2182 + }, + { + "epoch": 0.8721534159009189, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2183 + }, + { + "epoch": 0.8725529364762286, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2184 + }, + { + "epoch": 0.8729524570515381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2185 + }, + { + "epoch": 0.8733519776268478, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2186 + }, + { + "epoch": 0.8737514982021574, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2187 + }, + { + "epoch": 0.874151018777467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2188 + }, + { + "epoch": 0.8745505393527767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2189 + }, + { + "epoch": 0.8749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2190 + }, + { + "epoch": 0.875349580503396, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2191 + }, + { + "epoch": 0.8757491010787055, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2192 + }, + { + "epoch": 0.8761486216540152, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2193 + }, + { + "epoch": 0.8765481422293248, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2194 + }, + { + "epoch": 0.8769476628046344, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2195 + }, + { + "epoch": 0.8773471833799441, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2196 + }, + { + "epoch": 0.8777467039552537, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2197 + }, + { + "epoch": 0.8781462245305633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2198 + }, + { + "epoch": 0.8785457451058729, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 2199 + }, + { + "epoch": 0.8789452656811826, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2200 + }, + { + "epoch": 0.8793447862564923, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2201 + }, + { + "epoch": 0.8797443068318018, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2202 + }, + { + "epoch": 0.8801438274071115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2203 + }, + { + "epoch": 0.8805433479824211, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2204 + }, + { + "epoch": 0.8809428685577307, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2205 + }, + { + "epoch": 0.8813423891330403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2206 + }, + { + "epoch": 0.88174190970835, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2207 + }, + { + "epoch": 0.8821414302836597, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2208 + }, + { + "epoch": 0.8825409508589692, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2209 + }, + { + "epoch": 0.8829404714342789, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2210 + }, + { + "epoch": 0.8833399920095885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2211 + }, + { + "epoch": 0.8837395125848981, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 2212 + }, + { + "epoch": 0.8841390331602077, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2213 + }, + { + "epoch": 0.8845385537355174, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2214 + }, + { + "epoch": 0.884938074310827, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2215 + }, + { + "epoch": 0.8853375948861366, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2216 + }, + { + "epoch": 0.8857371154614463, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2217 + }, + { + "epoch": 0.8861366360367559, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2218 + }, + { + "epoch": 0.8865361566120655, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2219 + }, + { + "epoch": 0.8869356771873752, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2220 + }, + { + "epoch": 0.8873351977626848, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2221 + }, + { + "epoch": 0.8877347183379944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2222 + }, + { + "epoch": 0.888134238913304, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2223 + }, + { + "epoch": 0.8885337594886137, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2224 + }, + { + "epoch": 0.8889332800639232, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2225 + }, + { + "epoch": 0.8893328006392329, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2226 + }, + { + "epoch": 0.8897323212145426, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2227 + }, + { + "epoch": 0.8901318417898522, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2228 + }, + { + "epoch": 0.8905313623651618, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2229 + }, + { + "epoch": 0.8909308829404714, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2230 + }, + { + "epoch": 0.8913304035157811, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2231 + }, + { + "epoch": 0.8917299240910906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2232 + }, + { + "epoch": 0.8921294446664003, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2233 + }, + { + "epoch": 0.89252896524171, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2234 + }, + { + "epoch": 0.8929284858170196, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2235 + }, + { + "epoch": 0.8933280063923292, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2236 + }, + { + "epoch": 0.8937275269676388, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2237 + }, + { + "epoch": 0.8941270475429485, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2238 + }, + { + "epoch": 0.894526568118258, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2239 + }, + { + "epoch": 0.8949260886935677, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2240 + }, + { + "epoch": 0.8953256092688774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2241 + }, + { + "epoch": 0.895725129844187, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2242 + }, + { + "epoch": 0.8961246504194966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2243 + }, + { + "epoch": 0.8965241709948062, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2244 + }, + { + "epoch": 0.8969236915701159, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2245 + }, + { + "epoch": 0.8973232121454255, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2246 + }, + { + "epoch": 0.8977227327207351, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2247 + }, + { + "epoch": 0.8981222532960448, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2248 + }, + { + "epoch": 0.8985217738713543, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2249 + }, + { + "epoch": 0.898921294446664, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2250 + }, + { + "epoch": 0.8993208150219736, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2251 + }, + { + "epoch": 0.8997203355972833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2252 + }, + { + "epoch": 0.9001198561725929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2253 + }, + { + "epoch": 0.9005193767479025, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2254 + }, + { + "epoch": 0.9009188973232122, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2255 + }, + { + "epoch": 0.9013184178985217, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2256 + }, + { + "epoch": 0.9017179384738314, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2257 + }, + { + "epoch": 0.902117459049141, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2258 + }, + { + "epoch": 0.9025169796244507, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2259 + }, + { + "epoch": 0.9029165001997603, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2260 + }, + { + "epoch": 0.9033160207750699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2261 + }, + { + "epoch": 0.9037155413503796, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2262 + }, + { + "epoch": 0.9041150619256891, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2263 + }, + { + "epoch": 0.9045145825009988, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2264 + }, + { + "epoch": 0.9049141030763085, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2265 + }, + { + "epoch": 0.905313623651618, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2266 + }, + { + "epoch": 0.9057131442269277, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2267 + }, + { + "epoch": 0.9061126648022373, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2268 + }, + { + "epoch": 0.906512185377547, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 2269 + }, + { + "epoch": 0.9069117059528565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2270 + }, + { + "epoch": 0.9073112265281662, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2271 + }, + { + "epoch": 0.9077107471034759, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2272 + }, + { + "epoch": 0.9081102676787854, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2273 + }, + { + "epoch": 0.9085097882540951, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2274 + }, + { + "epoch": 0.9089093088294047, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2275 + }, + { + "epoch": 0.9093088294047144, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2276 + }, + { + "epoch": 0.9097083499800239, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2277 + }, + { + "epoch": 0.9101078705553336, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2278 + }, + { + "epoch": 0.9105073911306433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2279 + }, + { + "epoch": 0.9109069117059528, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2280 + }, + { + "epoch": 0.9113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2281 + }, + { + "epoch": 0.9117059528565721, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2282 + }, + { + "epoch": 0.9121054734318818, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2283 + }, + { + "epoch": 0.9125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2284 + }, + { + "epoch": 0.912904514582501, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2285 + }, + { + "epoch": 0.9133040351578107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2286 + }, + { + "epoch": 0.9137035557331202, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2287 + }, + { + "epoch": 0.9141030763084299, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2288 + }, + { + "epoch": 0.9145025968837395, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2289 + }, + { + "epoch": 0.9149021174590491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2290 + }, + { + "epoch": 0.9153016380343588, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 2291 + }, + { + "epoch": 0.9157011586096684, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2292 + }, + { + "epoch": 0.9161006791849781, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2293 + }, + { + "epoch": 0.9165001997602876, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2294 + }, + { + "epoch": 0.9168997203355973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2295 + }, + { + "epoch": 0.9172992409109069, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2296 + }, + { + "epoch": 0.9176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2297 + }, + { + "epoch": 0.9180982820615262, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2298 + }, + { + "epoch": 0.9184978026368358, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2299 + }, + { + "epoch": 0.9188973232121455, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2300 + }, + { + "epoch": 0.919296843787455, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2301 + }, + { + "epoch": 0.9196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2302 + }, + { + "epoch": 0.9200958849380743, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 2303 + }, + { + "epoch": 0.9204954055133839, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2304 + }, + { + "epoch": 0.9208949260886936, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2305 + }, + { + "epoch": 0.9212944466640032, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2306 + }, + { + "epoch": 0.9216939672393129, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2307 + }, + { + "epoch": 0.9220934878146224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2308 + }, + { + "epoch": 0.9224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2309 + }, + { + "epoch": 0.9228925289652418, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2310 + }, + { + "epoch": 0.9232920495405513, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2311 + }, + { + "epoch": 0.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2312 + }, + { + "epoch": 0.9240910906911706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2313 + }, + { + "epoch": 0.9244906112664802, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2314 + }, + { + "epoch": 0.9248901318417898, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2315 + }, + { + "epoch": 0.9252896524170995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 2316 + }, + { + "epoch": 0.9256891729924092, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2317 + }, + { + "epoch": 0.9260886935677187, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2318 + }, + { + "epoch": 0.9264882141430284, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2319 + }, + { + "epoch": 0.926887734718338, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2320 + }, + { + "epoch": 0.9272872552936476, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2321 + }, + { + "epoch": 0.9276867758689572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2322 + }, + { + "epoch": 0.9280862964442669, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2323 + }, + { + "epoch": 0.9284858170195766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2324 + }, + { + "epoch": 0.9288853375948861, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2325 + }, + { + "epoch": 0.9292848581701958, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2326 + }, + { + "epoch": 0.9296843787455054, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2327 + }, + { + "epoch": 0.930083899320815, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2328 + }, + { + "epoch": 0.9304834198961246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 2329 + }, + { + "epoch": 0.9308829404714343, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2330 + }, + { + "epoch": 0.931282461046744, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2331 + }, + { + "epoch": 0.9316819816220535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 2332 + }, + { + "epoch": 0.9320815021973632, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2333 + }, + { + "epoch": 0.9324810227726728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2334 + }, + { + "epoch": 0.9328805433479824, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2335 + }, + { + "epoch": 0.9332800639232921, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2336 + }, + { + "epoch": 0.9336795844986017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2337 + }, + { + "epoch": 0.9340791050739113, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2338 + }, + { + "epoch": 0.9344786256492209, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2339 + }, + { + "epoch": 0.9348781462245306, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2340 + }, + { + "epoch": 0.9352776667998401, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2341 + }, + { + "epoch": 0.9356771873751498, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2342 + }, + { + "epoch": 0.9360767079504595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2343 + }, + { + "epoch": 0.9364762285257691, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 2344 + }, + { + "epoch": 0.9368757491010787, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2345 + }, + { + "epoch": 0.9372752696763883, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2346 + }, + { + "epoch": 0.937674790251698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2347 + }, + { + "epoch": 0.9380743108270075, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2348 + }, + { + "epoch": 0.9384738314023172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2349 + }, + { + "epoch": 0.9388733519776269, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2350 + }, + { + "epoch": 0.9392728725529365, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2351 + }, + { + "epoch": 0.9396723931282461, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2352 + }, + { + "epoch": 0.9400719137035557, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2353 + }, + { + "epoch": 0.9404714342788654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2354 + }, + { + "epoch": 0.9408709548541749, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2355 + }, + { + "epoch": 0.9412704754294846, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2356 + }, + { + "epoch": 0.9416699960047943, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2357 + }, + { + "epoch": 0.9420695165801038, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2358 + }, + { + "epoch": 0.9424690371554135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 2359 + }, + { + "epoch": 0.9428685577307231, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2360 + }, + { + "epoch": 0.9432680783060328, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 2361 + }, + { + "epoch": 0.9436675988813424, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 2362 + }, + { + "epoch": 0.944067119456652, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 2363 + }, + { + "epoch": 0.9444666400319617, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2364 + }, + { + "epoch": 0.9448661606072712, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 2365 + }, + { + "epoch": 0.9452656811825809, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2366 + }, + { + "epoch": 0.9456652017578905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2367 + }, + { + "epoch": 0.9460647223332002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2368 + }, + { + "epoch": 0.9464642429085098, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2369 + }, + { + "epoch": 0.9468637634838194, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2370 + }, + { + "epoch": 0.9472632840591291, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2371 + }, + { + "epoch": 0.9476628046344386, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2372 + }, + { + "epoch": 0.9480623252097483, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2373 + }, + { + "epoch": 0.9484618457850579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2374 + }, + { + "epoch": 0.9488613663603676, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2375 + }, + { + "epoch": 0.9492608869356772, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2376 + }, + { + "epoch": 0.9496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2377 + }, + { + "epoch": 0.9500599280862965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2378 + }, + { + "epoch": 0.950459448661606, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2379 + }, + { + "epoch": 0.9508589692369157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2380 + }, + { + "epoch": 0.9512584898122254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2381 + }, + { + "epoch": 0.951658010387535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 2382 + }, + { + "epoch": 0.9520575309628446, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2383 + }, + { + "epoch": 0.9524570515381542, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2384 + }, + { + "epoch": 0.9528565721134639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2385 + }, + { + "epoch": 0.9532560926887734, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2386 + }, + { + "epoch": 0.9536556132640831, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2387 + }, + { + "epoch": 0.9540551338393928, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2388 + }, + { + "epoch": 0.9544546544147023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2389 + }, + { + "epoch": 0.954854174990012, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 2390 + }, + { + "epoch": 0.9552536955653216, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2391 + }, + { + "epoch": 0.9556532161406313, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 2392 + }, + { + "epoch": 0.9560527367159408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2393 + }, + { + "epoch": 0.9564522572912505, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2394 + }, + { + "epoch": 0.9568517778665602, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2395 + }, + { + "epoch": 0.9572512984418697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2396 + }, + { + "epoch": 0.9576508190171794, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2397 + }, + { + "epoch": 0.958050339592489, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2398 + }, + { + "epoch": 0.9584498601677987, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2399 + }, + { + "epoch": 0.9588493807431082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2400 + }, + { + "epoch": 0.9592489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2401 + }, + { + "epoch": 0.9596484218937276, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2402 + }, + { + "epoch": 0.9600479424690371, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2403 + }, + { + "epoch": 0.9604474630443468, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2404 + }, + { + "epoch": 0.9608469836196564, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2405 + }, + { + "epoch": 0.961246504194966, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2406 + }, + { + "epoch": 0.9616460247702757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2407 + }, + { + "epoch": 0.9620455453455853, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 2408 + }, + { + "epoch": 0.962445065920895, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2409 + }, + { + "epoch": 0.9628445864962045, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2410 + }, + { + "epoch": 0.9632441070715142, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2411 + }, + { + "epoch": 0.9636436276468238, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2412 + }, + { + "epoch": 0.9640431482221334, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2413 + }, + { + "epoch": 0.9644426687974431, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2414 + }, + { + "epoch": 0.9648421893727527, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2415 + }, + { + "epoch": 0.9652417099480624, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2416 + }, + { + "epoch": 0.9656412305233719, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2417 + }, + { + "epoch": 0.9660407510986816, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2418 + }, + { + "epoch": 0.9664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2419 + }, + { + "epoch": 0.9668397922493008, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2420 + }, + { + "epoch": 0.9672393128246105, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2421 + }, + { + "epoch": 0.9676388333999201, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2422 + }, + { + "epoch": 0.9680383539752297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2423 + }, + { + "epoch": 0.9684378745505393, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2424 + }, + { + "epoch": 0.968837395125849, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 2425 + }, + { + "epoch": 0.9692369157011587, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2426 + }, + { + "epoch": 0.9696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 2427 + }, + { + "epoch": 0.9700359568517779, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2428 + }, + { + "epoch": 0.9704354774270875, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2429 + }, + { + "epoch": 0.9708349980023971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2430 + }, + { + "epoch": 0.9712345185777067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2431 + }, + { + "epoch": 0.9716340391530164, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2432 + }, + { + "epoch": 0.9720335597283261, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2433 + }, + { + "epoch": 0.9724330803036356, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2434 + }, + { + "epoch": 0.9728326008789453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2435 + }, + { + "epoch": 0.9732321214542549, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2436 + }, + { + "epoch": 0.9736316420295645, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2437 + }, + { + "epoch": 0.9740311626048741, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2438 + }, + { + "epoch": 0.9744306831801838, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2439 + }, + { + "epoch": 0.9748302037554935, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2440 + }, + { + "epoch": 0.975229724330803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2441 + }, + { + "epoch": 0.9756292449061127, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2442 + }, + { + "epoch": 0.9760287654814223, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 2443 + }, + { + "epoch": 0.9764282860567319, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2444 + }, + { + "epoch": 0.9768278066320415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2445 + }, + { + "epoch": 0.9772273272073512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2446 + }, + { + "epoch": 0.9776268477826608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2447 + }, + { + "epoch": 0.9780263683579704, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2448 + }, + { + "epoch": 0.9784258889332801, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2449 + }, + { + "epoch": 0.9788254095085897, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2450 + }, + { + "epoch": 0.9792249300838993, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2451 + }, + { + "epoch": 0.979624450659209, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2452 + }, + { + "epoch": 0.9800239712345186, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2453 + }, + { + "epoch": 0.9804234918098282, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2454 + }, + { + "epoch": 0.9808230123851378, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2455 + }, + { + "epoch": 0.9812225329604475, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2456 + }, + { + "epoch": 0.981622053535757, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2457 + }, + { + "epoch": 0.9820215741110667, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2458 + }, + { + "epoch": 0.9824210946863764, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2459 + }, + { + "epoch": 0.982820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2460 + }, + { + "epoch": 0.9832201358369956, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2461 + }, + { + "epoch": 0.9836196564123052, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2462 + }, + { + "epoch": 0.9840191769876149, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2463 + }, + { + "epoch": 0.9844186975629244, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2464 + }, + { + "epoch": 0.9848182181382341, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2465 + }, + { + "epoch": 0.9852177387135438, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2466 + }, + { + "epoch": 0.9856172592888534, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2467 + }, + { + "epoch": 0.986016779864163, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2468 + }, + { + "epoch": 0.9864163004394726, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2469 + }, + { + "epoch": 0.9868158210147823, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2470 + }, + { + "epoch": 0.9872153415900919, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2471 + }, + { + "epoch": 0.9876148621654015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2472 + }, + { + "epoch": 0.9880143827407112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2473 + }, + { + "epoch": 0.9884139033160207, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2474 + }, + { + "epoch": 0.9888134238913304, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2475 + }, + { + "epoch": 0.98921294446664, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2476 + }, + { + "epoch": 0.9896124650419497, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2477 + }, + { + "epoch": 0.9900119856172593, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2478 + }, + { + "epoch": 0.9904115061925689, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2479 + }, + { + "epoch": 0.9908110267678786, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2480 + }, + { + "epoch": 0.9912105473431881, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2481 + }, + { + "epoch": 0.9916100679184978, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2482 + }, + { + "epoch": 0.9920095884938074, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2483 + }, + { + "epoch": 0.9924091090691171, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2484 + }, + { + "epoch": 0.9928086296444267, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2485 + }, + { + "epoch": 0.9932081502197363, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2486 + }, + { + "epoch": 0.993607670795046, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2487 + }, + { + "epoch": 0.9940071913703555, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 2488 + }, + { + "epoch": 0.9944067119456652, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2489 + }, + { + "epoch": 0.9948062325209748, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2490 + }, + { + "epoch": 0.9952057530962845, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2491 + }, + { + "epoch": 0.9956052736715941, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2492 + }, + { + "epoch": 0.9960047942469037, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2493 + }, + { + "epoch": 0.9964043148222134, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2494 + }, + { + "epoch": 0.9968038353975229, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2495 + }, + { + "epoch": 0.9972033559728326, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2496 + }, + { + "epoch": 0.9976028765481423, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2497 + }, + { + "epoch": 0.9980023971234518, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2498 + }, + { + "epoch": 0.9984019176987615, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2499 + }, + { + "epoch": 0.9988014382740711, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2500 + }, + { + "epoch": 0.9992009588493808, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2501 + }, + { + "epoch": 0.9996004794246903, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2502 + }, + { + "epoch": 1.0, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2503 + }, + { + "epoch": 1.0003995205753096, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2504 + }, + { + "epoch": 1.0007990411506194, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2505 + }, + { + "epoch": 1.001198561725929, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2506 + }, + { + "epoch": 1.0015980823012385, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2507 + }, + { + "epoch": 1.001997602876548, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2508 + }, + { + "epoch": 1.0023971234518578, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2509 + }, + { + "epoch": 1.0027966440271674, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2510 + }, + { + "epoch": 1.003196164602477, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2511 + }, + { + "epoch": 1.0035956851777867, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 2512 + }, + { + "epoch": 1.0039952057530963, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2513 + }, + { + "epoch": 1.0043947263284059, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2514 + }, + { + "epoch": 1.0047942469037154, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2515 + }, + { + "epoch": 1.0051937674790252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2516 + }, + { + "epoch": 1.0055932880543348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2517 + }, + { + "epoch": 1.0059928086296444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2518 + }, + { + "epoch": 1.0063923292049541, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2519 + }, + { + "epoch": 1.0067918497802637, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2520 + }, + { + "epoch": 1.0071913703555733, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2521 + }, + { + "epoch": 1.0075908909308828, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2522 + }, + { + "epoch": 1.0079904115061926, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2523 + }, + { + "epoch": 1.0083899320815022, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2524 + }, + { + "epoch": 1.0087894526568117, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 2525 + }, + { + "epoch": 1.0091889732321215, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2526 + }, + { + "epoch": 1.009588493807431, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2527 + }, + { + "epoch": 1.0099880143827407, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2528 + }, + { + "epoch": 1.0103875349580504, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2529 + }, + { + "epoch": 1.01078705553336, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2530 + }, + { + "epoch": 1.0111865761086696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2531 + }, + { + "epoch": 1.0115860966839791, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2532 + }, + { + "epoch": 1.011985617259289, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2533 + }, + { + "epoch": 1.0123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2534 + }, + { + "epoch": 1.012784658409908, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2535 + }, + { + "epoch": 1.0131841789852178, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 2536 + }, + { + "epoch": 1.0135836995605274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2537 + }, + { + "epoch": 1.013983220135837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2538 + }, + { + "epoch": 1.0143827407111465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2539 + }, + { + "epoch": 1.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2540 + }, + { + "epoch": 1.0151817818617659, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 2541 + }, + { + "epoch": 1.0155813024370755, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2542 + }, + { + "epoch": 1.0159808230123852, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2543 + }, + { + "epoch": 1.0163803435876948, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 2544 + }, + { + "epoch": 1.0167798641630044, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2545 + }, + { + "epoch": 1.017179384738314, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2546 + }, + { + "epoch": 1.0175789053136237, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2547 + }, + { + "epoch": 1.0179784258889333, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2548 + }, + { + "epoch": 1.0183779464642428, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2549 + }, + { + "epoch": 1.0187774670395526, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2550 + }, + { + "epoch": 1.0191769876148622, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2551 + }, + { + "epoch": 1.0195765081901718, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 2552 + }, + { + "epoch": 1.0199760287654813, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2553 + }, + { + "epoch": 1.020375549340791, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2554 + }, + { + "epoch": 1.0207750699161007, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2555 + }, + { + "epoch": 1.0211745904914102, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2556 + }, + { + "epoch": 1.02157411106672, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 2557 + }, + { + "epoch": 1.0219736316420296, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2558 + }, + { + "epoch": 1.0223731522173392, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2559 + }, + { + "epoch": 1.0227726727926487, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2560 + }, + { + "epoch": 1.0231721933679585, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2561 + }, + { + "epoch": 1.023571713943268, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 2562 + }, + { + "epoch": 1.0239712345185776, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2563 + }, + { + "epoch": 1.0243707550938874, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2564 + }, + { + "epoch": 1.024770275669197, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2565 + }, + { + "epoch": 1.0251697962445065, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2566 + }, + { + "epoch": 1.0255693168198161, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2567 + }, + { + "epoch": 1.025968837395126, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2568 + }, + { + "epoch": 1.0263683579704355, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2569 + }, + { + "epoch": 1.026767878545745, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2570 + }, + { + "epoch": 1.0271673991210548, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2571 + }, + { + "epoch": 1.0275669196963644, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2572 + }, + { + "epoch": 1.027966440271674, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2573 + }, + { + "epoch": 1.0283659608469837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2574 + }, + { + "epoch": 1.0287654814222933, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2575 + }, + { + "epoch": 1.0291650019976029, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 2576 + }, + { + "epoch": 1.0295645225729124, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2577 + }, + { + "epoch": 1.0299640431482222, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2578 + }, + { + "epoch": 1.0303635637235318, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 2579 + }, + { + "epoch": 1.0307630842988413, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2580 + }, + { + "epoch": 1.0311626048741511, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2581 + }, + { + "epoch": 1.0315621254494607, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2582 + }, + { + "epoch": 1.0319616460247703, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2583 + }, + { + "epoch": 1.0323611666000798, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2584 + }, + { + "epoch": 1.0327606871753896, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2585 + }, + { + "epoch": 1.0331602077506992, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2586 + }, + { + "epoch": 1.0335597283260087, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2587 + }, + { + "epoch": 1.0339592489013185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2588 + }, + { + "epoch": 1.034358769476628, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2589 + }, + { + "epoch": 1.0347582900519376, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2590 + }, + { + "epoch": 1.0351578106272472, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2591 + }, + { + "epoch": 1.035557331202557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 2592 + }, + { + "epoch": 1.0359568517778666, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2593 + }, + { + "epoch": 1.0363563723531761, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2594 + }, + { + "epoch": 1.036755892928486, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2595 + }, + { + "epoch": 1.0371554135037955, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2596 + }, + { + "epoch": 1.037554934079105, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 2597 + }, + { + "epoch": 1.0379544546544146, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2598 + }, + { + "epoch": 1.0383539752297244, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2599 + }, + { + "epoch": 1.038753495805034, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2600 + }, + { + "epoch": 1.0391530163803435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2601 + }, + { + "epoch": 1.0395525369556533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2602 + }, + { + "epoch": 1.0399520575309629, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2603 + }, + { + "epoch": 1.0403515781062724, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 2604 + }, + { + "epoch": 1.040751098681582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2605 + }, + { + "epoch": 1.0411506192568918, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2606 + }, + { + "epoch": 1.0415501398322013, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2607 + }, + { + "epoch": 1.041949660407511, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2608 + }, + { + "epoch": 1.0423491809828207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 2609 + }, + { + "epoch": 1.0427487015581303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2610 + }, + { + "epoch": 1.0431482221334398, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2611 + }, + { + "epoch": 1.0435477427087494, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2612 + }, + { + "epoch": 1.0439472632840592, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2613 + }, + { + "epoch": 1.0443467838593687, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2614 + }, + { + "epoch": 1.0447463044346783, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2615 + }, + { + "epoch": 1.045145825009988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2616 + }, + { + "epoch": 1.0455453455852977, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2617 + }, + { + "epoch": 1.0459448661606072, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2618 + }, + { + "epoch": 1.046344386735917, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2619 + }, + { + "epoch": 1.0467439073112266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2620 + }, + { + "epoch": 1.0471434278865361, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2621 + }, + { + "epoch": 1.0475429484618457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2622 + }, + { + "epoch": 1.0479424690371555, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2623 + }, + { + "epoch": 1.048341989612465, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2624 + }, + { + "epoch": 1.0487415101877746, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2625 + }, + { + "epoch": 1.0491410307630844, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 2626 + }, + { + "epoch": 1.049540551338394, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2627 + }, + { + "epoch": 1.0499400719137035, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2628 + }, + { + "epoch": 1.050339592489013, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2629 + }, + { + "epoch": 1.0507391130643229, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2630 + }, + { + "epoch": 1.0511386336396324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2631 + }, + { + "epoch": 1.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2632 + }, + { + "epoch": 1.0519376747902518, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2633 + }, + { + "epoch": 1.0523371953655614, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2634 + }, + { + "epoch": 1.052736715940871, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2635 + }, + { + "epoch": 1.0531362365161805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2636 + }, + { + "epoch": 1.0535357570914903, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2637 + }, + { + "epoch": 1.0539352776667998, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2638 + }, + { + "epoch": 1.0543347982421094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2639 + }, + { + "epoch": 1.0547343188174192, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2640 + }, + { + "epoch": 1.0551338393927288, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2641 + }, + { + "epoch": 1.0555333599680383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2642 + }, + { + "epoch": 1.0559328805433479, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2643 + }, + { + "epoch": 1.0563324011186577, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2644 + }, + { + "epoch": 1.0567319216939672, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2645 + }, + { + "epoch": 1.0571314422692768, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2646 + }, + { + "epoch": 1.0575309628445866, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2647 + }, + { + "epoch": 1.0579304834198962, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2648 + }, + { + "epoch": 1.0583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2649 + }, + { + "epoch": 1.0587295245705153, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2650 + }, + { + "epoch": 1.059129045145825, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 2651 + }, + { + "epoch": 1.0595285657211346, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2652 + }, + { + "epoch": 1.0599280862964442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2653 + }, + { + "epoch": 1.060327606871754, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2654 + }, + { + "epoch": 1.0607271274470635, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2655 + }, + { + "epoch": 1.061126648022373, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2656 + }, + { + "epoch": 1.0615261685976827, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2657 + }, + { + "epoch": 1.0619256891729925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2658 + }, + { + "epoch": 1.062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2659 + }, + { + "epoch": 1.0627247303236116, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2660 + }, + { + "epoch": 1.0631242508989214, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2661 + }, + { + "epoch": 1.063523771474231, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2662 + }, + { + "epoch": 1.0639232920495405, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2663 + }, + { + "epoch": 1.0643228126248503, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2664 + }, + { + "epoch": 1.0647223332001599, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2665 + }, + { + "epoch": 1.0651218537754694, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2666 + }, + { + "epoch": 1.065521374350779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2667 + }, + { + "epoch": 1.0659208949260888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2668 + }, + { + "epoch": 1.0663204155013983, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 2669 + }, + { + "epoch": 1.066719936076708, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2670 + }, + { + "epoch": 1.0671194566520175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2671 + }, + { + "epoch": 1.0675189772273272, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2672 + }, + { + "epoch": 1.0679184978026368, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2673 + }, + { + "epoch": 1.0683180183779464, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2674 + }, + { + "epoch": 1.0687175389532562, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2675 + }, + { + "epoch": 1.0691170595285657, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2676 + }, + { + "epoch": 1.0695165801038753, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2677 + }, + { + "epoch": 1.069916100679185, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 2678 + }, + { + "epoch": 1.0703156212544946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2679 + }, + { + "epoch": 1.0707151418298042, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2680 + }, + { + "epoch": 1.0711146624051138, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2681 + }, + { + "epoch": 1.0715141829804236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2682 + }, + { + "epoch": 1.0719137035557331, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2683 + }, + { + "epoch": 1.0723132241310427, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2684 + }, + { + "epoch": 1.0727127447063525, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2685 + }, + { + "epoch": 1.073112265281662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2686 + }, + { + "epoch": 1.0735117858569716, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2687 + }, + { + "epoch": 1.0739113064322812, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2688 + }, + { + "epoch": 1.074310827007591, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2689 + }, + { + "epoch": 1.0747103475829005, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2690 + }, + { + "epoch": 1.07510986815821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2691 + }, + { + "epoch": 1.0755093887335199, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2692 + }, + { + "epoch": 1.0759089093088294, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2693 + }, + { + "epoch": 1.076308429884139, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2694 + }, + { + "epoch": 1.0767079504594486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2695 + }, + { + "epoch": 1.0771074710347583, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2696 + }, + { + "epoch": 1.077506991610068, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2697 + }, + { + "epoch": 1.0779065121853775, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2698 + }, + { + "epoch": 1.0783060327606873, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 1.0787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2700 + }, + { + "epoch": 1.0791050739113064, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2701 + }, + { + "epoch": 1.079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 2702 + }, + { + "epoch": 1.0799041150619257, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2703 + }, + { + "epoch": 1.0803036356372353, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2704 + }, + { + "epoch": 1.0807031562125449, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2705 + }, + { + "epoch": 1.0811026767878547, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2706 + }, + { + "epoch": 1.0815021973631642, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2707 + }, + { + "epoch": 1.0819017179384738, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2708 + }, + { + "epoch": 1.0823012385137836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2709 + }, + { + "epoch": 1.0827007590890931, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2710 + }, + { + "epoch": 1.0831002796644027, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 2711 + }, + { + "epoch": 1.0834998002397123, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2712 + }, + { + "epoch": 1.083899320815022, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2713 + }, + { + "epoch": 1.0842988413903316, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2714 + }, + { + "epoch": 1.0846983619656412, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2715 + }, + { + "epoch": 1.0850978825409507, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2716 + }, + { + "epoch": 1.0854974031162605, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2717 + }, + { + "epoch": 1.08589692369157, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2718 + }, + { + "epoch": 1.0862964442668797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2719 + }, + { + "epoch": 1.0866959648421894, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2720 + }, + { + "epoch": 1.087095485417499, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2721 + }, + { + "epoch": 1.0874950059928086, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2722 + }, + { + "epoch": 1.0878945265681184, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2723 + }, + { + "epoch": 1.088294047143428, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2724 + }, + { + "epoch": 1.0886935677187375, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2725 + }, + { + "epoch": 1.089093088294047, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2726 + }, + { + "epoch": 1.0894926088693568, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2727 + }, + { + "epoch": 1.0898921294446664, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2728 + }, + { + "epoch": 1.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2729 + }, + { + "epoch": 1.0906911705952858, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2730 + }, + { + "epoch": 1.0910906911705953, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2731 + }, + { + "epoch": 1.0914902117459049, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2732 + }, + { + "epoch": 1.0918897323212144, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2733 + }, + { + "epoch": 1.0922892528965242, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2734 + }, + { + "epoch": 1.0926887734718338, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2735 + }, + { + "epoch": 1.0930882940471434, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2736 + }, + { + "epoch": 1.0934878146224531, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2737 + }, + { + "epoch": 1.0938873351977627, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2738 + }, + { + "epoch": 1.0942868557730723, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2739 + }, + { + "epoch": 1.0946863763483818, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2740 + }, + { + "epoch": 1.0950858969236916, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2741 + }, + { + "epoch": 1.0954854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2742 + }, + { + "epoch": 1.0958849380743108, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2743 + }, + { + "epoch": 1.0962844586496205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2744 + }, + { + "epoch": 1.09668397922493, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2745 + }, + { + "epoch": 1.0970834998002397, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2746 + }, + { + "epoch": 1.0974830203755492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2747 + }, + { + "epoch": 1.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2748 + }, + { + "epoch": 1.0982820615261686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2749 + }, + { + "epoch": 1.0986815821014781, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2750 + }, + { + "epoch": 1.099081102676788, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2751 + }, + { + "epoch": 1.0994806232520975, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2752 + }, + { + "epoch": 1.099880143827407, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2753 + }, + { + "epoch": 1.1002796644027169, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2754 + }, + { + "epoch": 1.1006791849780264, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2755 + }, + { + "epoch": 1.101078705553336, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 2756 + }, + { + "epoch": 1.1014782261286455, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2757 + }, + { + "epoch": 1.1018777467039553, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2758 + }, + { + "epoch": 1.102277267279265, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2759 + }, + { + "epoch": 1.1026767878545745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2760 + }, + { + "epoch": 1.103076308429884, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2761 + }, + { + "epoch": 1.1034758290051938, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2762 + }, + { + "epoch": 1.1038753495805034, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 2763 + }, + { + "epoch": 1.104274870155813, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2764 + }, + { + "epoch": 1.1046743907311227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2765 + }, + { + "epoch": 1.1050739113064323, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2766 + }, + { + "epoch": 1.1054734318817419, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2767 + }, + { + "epoch": 1.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2768 + }, + { + "epoch": 1.1062724730323612, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2769 + }, + { + "epoch": 1.1066719936076708, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2770 + }, + { + "epoch": 1.1070715141829803, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2771 + }, + { + "epoch": 1.1074710347582901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2772 + }, + { + "epoch": 1.1078705553335997, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2773 + }, + { + "epoch": 1.1082700759089092, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 2774 + }, + { + "epoch": 1.108669596484219, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2775 + }, + { + "epoch": 1.1090691170595286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2776 + }, + { + "epoch": 1.1094686376348382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2777 + }, + { + "epoch": 1.1098681582101477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2778 + }, + { + "epoch": 1.1102676787854575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2779 + }, + { + "epoch": 1.110667199360767, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2780 + }, + { + "epoch": 1.1110667199360766, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2781 + }, + { + "epoch": 1.1114662405113864, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2782 + }, + { + "epoch": 1.111865761086696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2783 + }, + { + "epoch": 1.1122652816620056, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2784 + }, + { + "epoch": 1.1126648022373151, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2785 + }, + { + "epoch": 1.113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2786 + }, + { + "epoch": 1.1134638433879345, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2787 + }, + { + "epoch": 1.113863363963244, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2788 + }, + { + "epoch": 1.1142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2789 + }, + { + "epoch": 1.1146624051138634, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2790 + }, + { + "epoch": 1.115061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2791 + }, + { + "epoch": 1.1154614462644825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2792 + }, + { + "epoch": 1.1158609668397923, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2793 + }, + { + "epoch": 1.1162604874151019, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2794 + }, + { + "epoch": 1.1166600079904114, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2795 + }, + { + "epoch": 1.1170595285657212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2796 + }, + { + "epoch": 1.1174590491410308, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2797 + }, + { + "epoch": 1.1178585697163403, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2798 + }, + { + "epoch": 1.1182580902916501, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2799 + }, + { + "epoch": 1.1186576108669597, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2800 + }, + { + "epoch": 1.1190571314422693, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2801 + }, + { + "epoch": 1.1194566520175788, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2802 + }, + { + "epoch": 1.1198561725928886, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2803 + }, + { + "epoch": 1.1202556931681982, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2804 + }, + { + "epoch": 1.1206552137435077, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2805 + }, + { + "epoch": 1.1210547343188173, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2806 + }, + { + "epoch": 1.121454254894127, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2807 + }, + { + "epoch": 1.1218537754694367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2808 + }, + { + "epoch": 1.1222532960447462, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2809 + }, + { + "epoch": 1.122652816620056, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2810 + }, + { + "epoch": 1.1230523371953656, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2811 + }, + { + "epoch": 1.1234518577706751, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2812 + }, + { + "epoch": 1.123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2813 + }, + { + "epoch": 1.1242508989212945, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2814 + }, + { + "epoch": 1.124650419496604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2815 + }, + { + "epoch": 1.1250499400719136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2816 + }, + { + "epoch": 1.1254494606472234, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2817 + }, + { + "epoch": 1.125848981222533, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2818 + }, + { + "epoch": 1.1262485017978425, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2819 + }, + { + "epoch": 1.1266480223731523, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2820 + }, + { + "epoch": 1.1270475429484619, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2821 + }, + { + "epoch": 1.1274470635237714, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2822 + }, + { + "epoch": 1.127846584099081, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 2823 + }, + { + "epoch": 1.1282461046743908, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 2824 + }, + { + "epoch": 1.1286456252497004, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2825 + }, + { + "epoch": 1.12904514582501, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2826 + }, + { + "epoch": 1.1294446664003197, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2827 + }, + { + "epoch": 1.1298441869756293, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2828 + }, + { + "epoch": 1.1302437075509388, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2829 + }, + { + "epoch": 1.1306432281262486, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2830 + }, + { + "epoch": 1.1310427487015582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2831 + }, + { + "epoch": 1.1314422692768678, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2832 + }, + { + "epoch": 1.1318417898521773, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2833 + }, + { + "epoch": 1.132241310427487, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2834 + }, + { + "epoch": 1.1326408310027967, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2835 + }, + { + "epoch": 1.1330403515781062, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2836 + }, + { + "epoch": 1.1334398721534158, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2837 + }, + { + "epoch": 1.1338393927287256, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2838 + }, + { + "epoch": 1.1342389133040351, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2839 + }, + { + "epoch": 1.1346384338793447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2840 + }, + { + "epoch": 1.1350379544546545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2841 + }, + { + "epoch": 1.135437475029964, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2842 + }, + { + "epoch": 1.1358369956052736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2843 + }, + { + "epoch": 1.1362365161805834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2844 + }, + { + "epoch": 1.136636036755893, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2845 + }, + { + "epoch": 1.1370355573312025, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2846 + }, + { + "epoch": 1.137435077906512, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2847 + }, + { + "epoch": 1.137834598481822, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2848 + }, + { + "epoch": 1.1382341190571315, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2849 + }, + { + "epoch": 1.138633639632441, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2850 + }, + { + "epoch": 1.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2851 + }, + { + "epoch": 1.1394326807830604, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2852 + }, + { + "epoch": 1.13983220135837, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2853 + }, + { + "epoch": 1.1402317219336795, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2854 + }, + { + "epoch": 1.1406312425089893, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2855 + }, + { + "epoch": 1.1410307630842988, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2856 + }, + { + "epoch": 1.1414302836596084, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2857 + }, + { + "epoch": 1.1418298042349182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2858 + }, + { + "epoch": 1.1422293248102278, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2859 + }, + { + "epoch": 1.1426288453855373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2860 + }, + { + "epoch": 1.143028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2861 + }, + { + "epoch": 1.1434278865361567, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2862 + }, + { + "epoch": 1.1438274071114662, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2863 + }, + { + "epoch": 1.1442269276867758, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2864 + }, + { + "epoch": 1.1446264482620856, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 2865 + }, + { + "epoch": 1.1450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2866 + }, + { + "epoch": 1.1454254894127047, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2867 + }, + { + "epoch": 1.1458250099880143, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 2868 + }, + { + "epoch": 1.146224530563324, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2869 + }, + { + "epoch": 1.1466240511386336, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2870 + }, + { + "epoch": 1.1470235717139432, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2871 + }, + { + "epoch": 1.147423092289253, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 2872 + }, + { + "epoch": 1.1478226128645626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2873 + }, + { + "epoch": 1.1482221334398721, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2874 + }, + { + "epoch": 1.148621654015182, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2875 + }, + { + "epoch": 1.1490211745904915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2876 + }, + { + "epoch": 1.149420695165801, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2877 + }, + { + "epoch": 1.1498202157411106, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 2878 + }, + { + "epoch": 1.1502197363164204, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2879 + }, + { + "epoch": 1.15061925689173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2880 + }, + { + "epoch": 1.1510187774670395, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2881 + }, + { + "epoch": 1.151418298042349, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2882 + }, + { + "epoch": 1.1518178186176589, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2883 + }, + { + "epoch": 1.1522173391929684, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2884 + }, + { + "epoch": 1.152616859768278, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2885 + }, + { + "epoch": 1.1530163803435878, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2886 + }, + { + "epoch": 1.1534159009188973, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2887 + }, + { + "epoch": 1.153815421494207, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 2888 + }, + { + "epoch": 1.1542149420695167, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2889 + }, + { + "epoch": 1.1546144626448263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2890 + }, + { + "epoch": 1.1550139832201358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 2891 + }, + { + "epoch": 1.1554135037954454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2892 + }, + { + "epoch": 1.1558130243707552, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2893 + }, + { + "epoch": 1.1562125449460647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2894 + }, + { + "epoch": 1.1566120655213743, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2895 + }, + { + "epoch": 1.1570115860966839, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2896 + }, + { + "epoch": 1.1574111066719937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2897 + }, + { + "epoch": 1.1578106272473032, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 2898 + }, + { + "epoch": 1.1582101478226128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2899 + }, + { + "epoch": 1.1586096683979226, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2900 + }, + { + "epoch": 1.1590091889732321, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2901 + }, + { + "epoch": 1.1594087095485417, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2902 + }, + { + "epoch": 1.1598082301238515, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2903 + }, + { + "epoch": 1.160207750699161, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2904 + }, + { + "epoch": 1.1606072712744706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2905 + }, + { + "epoch": 1.1610067918497802, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2906 + }, + { + "epoch": 1.16140631242509, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 2907 + }, + { + "epoch": 1.1618058330003995, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2908 + }, + { + "epoch": 1.162205353575709, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2909 + }, + { + "epoch": 1.1626048741510189, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2910 + }, + { + "epoch": 1.1630043947263284, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2911 + }, + { + "epoch": 1.163403915301638, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2912 + }, + { + "epoch": 1.1638034358769476, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2913 + }, + { + "epoch": 1.1642029564522574, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2914 + }, + { + "epoch": 1.164602477027567, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2915 + }, + { + "epoch": 1.1650019976028765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2916 + }, + { + "epoch": 1.1654015181781863, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2917 + }, + { + "epoch": 1.1658010387534958, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2918 + }, + { + "epoch": 1.1662005593288054, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2919 + }, + { + "epoch": 1.1666000799041152, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2920 + }, + { + "epoch": 1.1669996004794247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2921 + }, + { + "epoch": 1.1673991210547343, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2922 + }, + { + "epoch": 1.1677986416300439, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2923 + }, + { + "epoch": 1.1681981622053537, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2924 + }, + { + "epoch": 1.1685976827806632, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2925 + }, + { + "epoch": 1.1689972033559728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2926 + }, + { + "epoch": 1.1693967239312824, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2927 + }, + { + "epoch": 1.1697962445065921, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2928 + }, + { + "epoch": 1.1701957650819017, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2929 + }, + { + "epoch": 1.1705952856572113, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2930 + }, + { + "epoch": 1.170994806232521, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2931 + }, + { + "epoch": 1.1713943268078306, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2932 + }, + { + "epoch": 1.1717938473831402, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2933 + }, + { + "epoch": 1.17219336795845, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 2934 + }, + { + "epoch": 1.1725928885337595, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2935 + }, + { + "epoch": 1.172992409109069, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2936 + }, + { + "epoch": 1.1733919296843787, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2937 + }, + { + "epoch": 1.1737914502596885, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2938 + }, + { + "epoch": 1.174190970834998, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2939 + }, + { + "epoch": 1.1745904914103076, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2940 + }, + { + "epoch": 1.1749900119856171, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2941 + }, + { + "epoch": 1.175389532560927, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2942 + }, + { + "epoch": 1.1757890531362365, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2943 + }, + { + "epoch": 1.176188573711546, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2944 + }, + { + "epoch": 1.1765880942868558, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2945 + }, + { + "epoch": 1.1769876148621654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2946 + }, + { + "epoch": 1.177387135437475, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2947 + }, + { + "epoch": 1.1777866560127848, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 2948 + }, + { + "epoch": 1.1781861765880943, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2949 + }, + { + "epoch": 1.178585697163404, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2950 + }, + { + "epoch": 1.1789852177387135, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2951 + }, + { + "epoch": 1.1793847383140232, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2952 + }, + { + "epoch": 1.1797842588893328, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2953 + }, + { + "epoch": 1.1801837794646424, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2954 + }, + { + "epoch": 1.180583300039952, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2955 + }, + { + "epoch": 1.1809828206152617, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2956 + }, + { + "epoch": 1.1813823411905713, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2957 + }, + { + "epoch": 1.1817818617658808, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2958 + }, + { + "epoch": 1.1821813823411906, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2959 + }, + { + "epoch": 1.1825809029165002, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2960 + }, + { + "epoch": 1.1829804234918098, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2961 + }, + { + "epoch": 1.1833799440671195, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 2962 + }, + { + "epoch": 1.1837794646424291, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2963 + }, + { + "epoch": 1.1841789852177387, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 2964 + }, + { + "epoch": 1.1845785057930485, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2965 + }, + { + "epoch": 1.184978026368358, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2966 + }, + { + "epoch": 1.1853775469436676, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2967 + }, + { + "epoch": 1.1857770675189772, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2968 + }, + { + "epoch": 1.186176588094287, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2969 + }, + { + "epoch": 1.1865761086695965, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 2970 + }, + { + "epoch": 1.186975629244906, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2971 + }, + { + "epoch": 1.1873751498202156, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2972 + }, + { + "epoch": 1.1877746703955254, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 2973 + }, + { + "epoch": 1.188174190970835, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2974 + }, + { + "epoch": 1.1885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2975 + }, + { + "epoch": 1.1889732321214543, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2976 + }, + { + "epoch": 1.189372752696764, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2977 + }, + { + "epoch": 1.1897722732720735, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2978 + }, + { + "epoch": 1.1901717938473833, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2979 + }, + { + "epoch": 1.1905713144226928, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2980 + }, + { + "epoch": 1.1909708349980024, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2981 + }, + { + "epoch": 1.191370355573312, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2982 + }, + { + "epoch": 1.1917698761486217, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2983 + }, + { + "epoch": 1.1921693967239313, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2984 + }, + { + "epoch": 1.1925689172992409, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 2985 + }, + { + "epoch": 1.1929684378745504, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2986 + }, + { + "epoch": 1.1933679584498602, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2987 + }, + { + "epoch": 1.1937674790251698, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2988 + }, + { + "epoch": 1.1941669996004793, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2989 + }, + { + "epoch": 1.1945665201757891, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2990 + }, + { + "epoch": 1.1949660407510987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2991 + }, + { + "epoch": 1.1953655613264083, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2992 + }, + { + "epoch": 1.195765081901718, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2993 + }, + { + "epoch": 1.1961646024770276, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2994 + }, + { + "epoch": 1.1965641230523372, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2995 + }, + { + "epoch": 1.1969636436276467, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2996 + }, + { + "epoch": 1.1973631642029565, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2997 + }, + { + "epoch": 1.197762684778266, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2998 + }, + { + "epoch": 1.1981622053535756, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2999 + }, + { + "epoch": 1.1985617259288852, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3000 + }, + { + "epoch": 1.198961246504195, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 3001 + }, + { + "epoch": 1.1993607670795046, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3002 + }, + { + "epoch": 1.1997602876548141, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3003 + }, + { + "epoch": 1.200159808230124, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 3004 + }, + { + "epoch": 1.2005593288054335, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3005 + }, + { + "epoch": 1.200958849380743, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3006 + }, + { + "epoch": 1.2013583699560528, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3007 + }, + { + "epoch": 1.2017578905313624, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3008 + }, + { + "epoch": 1.202157411106672, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3009 + }, + { + "epoch": 1.2025569316819817, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 3010 + }, + { + "epoch": 1.2029564522572913, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 3011 + }, + { + "epoch": 1.2033559728326009, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3012 + }, + { + "epoch": 1.2037554934079104, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3013 + }, + { + "epoch": 1.2041550139832202, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3014 + }, + { + "epoch": 1.2045545345585298, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3015 + }, + { + "epoch": 1.2049540551338394, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3016 + }, + { + "epoch": 1.205353575709149, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3017 + }, + { + "epoch": 1.2057530962844587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3018 + }, + { + "epoch": 1.2061526168597683, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3019 + }, + { + "epoch": 1.2065521374350778, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3020 + }, + { + "epoch": 1.2069516580103876, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3021 + }, + { + "epoch": 1.2073511785856972, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3022 + }, + { + "epoch": 1.2077506991610067, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3023 + }, + { + "epoch": 1.2081502197363165, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3024 + }, + { + "epoch": 1.208549740311626, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3025 + }, + { + "epoch": 1.2089492608869357, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3026 + }, + { + "epoch": 1.2093487814622452, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3027 + }, + { + "epoch": 1.209748302037555, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3028 + }, + { + "epoch": 1.2101478226128646, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 3029 + }, + { + "epoch": 1.2105473431881741, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3030 + }, + { + "epoch": 1.2109468637634837, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3031 + }, + { + "epoch": 1.2113463843387935, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3032 + }, + { + "epoch": 1.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3033 + }, + { + "epoch": 1.2121454254894126, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3034 + }, + { + "epoch": 1.2125449460647224, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3035 + }, + { + "epoch": 1.212944466640032, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3036 + }, + { + "epoch": 1.2133439872153415, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3037 + }, + { + "epoch": 1.2137435077906513, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3038 + }, + { + "epoch": 1.2141430283659609, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 3039 + }, + { + "epoch": 1.2145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3040 + }, + { + "epoch": 1.21494206951658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3041 + }, + { + "epoch": 1.2153415900918898, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3042 + }, + { + "epoch": 1.2157411106671994, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3043 + }, + { + "epoch": 1.216140631242509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3044 + }, + { + "epoch": 1.2165401518178185, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3045 + }, + { + "epoch": 1.2169396723931283, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3046 + }, + { + "epoch": 1.2173391929684378, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3047 + }, + { + "epoch": 1.2177387135437474, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3048 + }, + { + "epoch": 1.2181382341190572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3049 + }, + { + "epoch": 1.2185377546943668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3050 + }, + { + "epoch": 1.2189372752696763, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 3051 + }, + { + "epoch": 1.219336795844986, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3052 + }, + { + "epoch": 1.2197363164202957, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3053 + }, + { + "epoch": 1.2201358369956052, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3054 + }, + { + "epoch": 1.220535357570915, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 3055 + }, + { + "epoch": 1.2209348781462246, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3056 + }, + { + "epoch": 1.2213343987215342, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3057 + }, + { + "epoch": 1.2217339192968437, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3058 + }, + { + "epoch": 1.2221334398721535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3059 + }, + { + "epoch": 1.222532960447463, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3060 + }, + { + "epoch": 1.2229324810227726, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3061 + }, + { + "epoch": 1.2233320015980822, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3062 + }, + { + "epoch": 1.223731522173392, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3063 + }, + { + "epoch": 1.2241310427487015, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3064 + }, + { + "epoch": 1.2245305633240111, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3065 + }, + { + "epoch": 1.224930083899321, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3066 + }, + { + "epoch": 1.2253296044746305, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3067 + }, + { + "epoch": 1.22572912504994, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3068 + }, + { + "epoch": 1.2261286456252498, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3069 + }, + { + "epoch": 1.2265281662005594, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3070 + }, + { + "epoch": 1.226927686775869, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 3071 + }, + { + "epoch": 1.2273272073511785, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3072 + }, + { + "epoch": 1.2277267279264883, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 3073 + }, + { + "epoch": 1.2281262485017979, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3074 + }, + { + "epoch": 1.2285257690771074, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3075 + }, + { + "epoch": 1.228925289652417, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 3076 + }, + { + "epoch": 1.2293248102277268, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3077 + }, + { + "epoch": 1.2297243308030363, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3078 + }, + { + "epoch": 1.230123851378346, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3079 + }, + { + "epoch": 1.2305233719536557, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3080 + }, + { + "epoch": 1.2309228925289653, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3081 + }, + { + "epoch": 1.2313224131042748, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3082 + }, + { + "epoch": 1.2317219336795846, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3083 + }, + { + "epoch": 1.2321214542548942, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3084 + }, + { + "epoch": 1.2325209748302037, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3085 + }, + { + "epoch": 1.2329204954055133, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3086 + }, + { + "epoch": 1.233320015980823, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3087 + }, + { + "epoch": 1.2337195365561326, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3088 + }, + { + "epoch": 1.2341190571314422, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3089 + }, + { + "epoch": 1.2345185777067518, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3090 + }, + { + "epoch": 1.2349180982820616, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3091 + }, + { + "epoch": 1.2353176188573711, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3092 + }, + { + "epoch": 1.2357171394326807, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3093 + }, + { + "epoch": 1.2361166600079905, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3094 + }, + { + "epoch": 1.2365161805833, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3095 + }, + { + "epoch": 1.2369157011586096, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3096 + }, + { + "epoch": 1.2373152217339194, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3097 + }, + { + "epoch": 1.237714742309229, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3098 + }, + { + "epoch": 1.2381142628845385, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3099 + }, + { + "epoch": 1.2385137834598483, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 3100 + }, + { + "epoch": 1.2389133040351579, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3101 + }, + { + "epoch": 1.2393128246104674, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3102 + }, + { + "epoch": 1.239712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3103 + }, + { + "epoch": 1.2401118657610868, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3104 + }, + { + "epoch": 1.2405113863363963, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3105 + }, + { + "epoch": 1.240910906911706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3106 + }, + { + "epoch": 1.2413104274870155, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3107 + }, + { + "epoch": 1.2417099480623253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3108 + }, + { + "epoch": 1.2421094686376348, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3109 + }, + { + "epoch": 1.2425089892129444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3110 + }, + { + "epoch": 1.2429085097882542, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3111 + }, + { + "epoch": 1.2433080303635637, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3112 + }, + { + "epoch": 1.2437075509388733, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3113 + }, + { + "epoch": 1.244107071514183, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3114 + }, + { + "epoch": 1.2445065920894927, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3115 + }, + { + "epoch": 1.2449061126648022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3116 + }, + { + "epoch": 1.2453056332401118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3117 + }, + { + "epoch": 1.2457051538154216, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3118 + }, + { + "epoch": 1.2461046743907311, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3119 + }, + { + "epoch": 1.2465041949660407, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3120 + }, + { + "epoch": 1.2469037155413503, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3121 + }, + { + "epoch": 1.24730323611666, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3122 + }, + { + "epoch": 1.2477027566919696, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3123 + }, + { + "epoch": 1.2481022772672792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3124 + }, + { + "epoch": 1.248501797842589, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3125 + }, + { + "epoch": 1.2489013184178985, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3126 + }, + { + "epoch": 1.249300838993208, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3127 + }, + { + "epoch": 1.2497003595685179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3128 + }, + { + "epoch": 1.2500998801438274, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 3129 + }, + { + "epoch": 1.250499400719137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3130 + }, + { + "epoch": 1.2508989212944468, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3131 + }, + { + "epoch": 1.2512984418697564, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3132 + }, + { + "epoch": 1.251697962445066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3133 + }, + { + "epoch": 1.2520974830203755, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3134 + }, + { + "epoch": 1.252497003595685, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3135 + }, + { + "epoch": 1.2528965241709948, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3136 + }, + { + "epoch": 1.2532960447463044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3137 + }, + { + "epoch": 1.253695565321614, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3138 + }, + { + "epoch": 1.2540950858969238, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3139 + }, + { + "epoch": 1.2544946064722333, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3140 + }, + { + "epoch": 1.2548941270475429, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3141 + }, + { + "epoch": 1.2552936476228527, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3142 + }, + { + "epoch": 1.2556931681981622, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3143 + }, + { + "epoch": 1.2560926887734718, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3144 + }, + { + "epoch": 1.2564922093487816, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3145 + }, + { + "epoch": 1.2568917299240912, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3146 + }, + { + "epoch": 1.2572912504994007, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3147 + }, + { + "epoch": 1.2576907710747103, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3148 + }, + { + "epoch": 1.2580902916500198, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3149 + }, + { + "epoch": 1.2584898122253296, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3150 + }, + { + "epoch": 1.2588893328006392, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 3151 + }, + { + "epoch": 1.2592888533759488, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3152 + }, + { + "epoch": 1.2596883739512585, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3153 + }, + { + "epoch": 1.260087894526568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3154 + }, + { + "epoch": 1.2604874151018777, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3155 + }, + { + "epoch": 1.2608869356771875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3156 + }, + { + "epoch": 1.261286456252497, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3157 + }, + { + "epoch": 1.2616859768278066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3158 + }, + { + "epoch": 1.2620854974031164, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3159 + }, + { + "epoch": 1.262485017978426, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3160 + }, + { + "epoch": 1.2628845385537355, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3161 + }, + { + "epoch": 1.263284059129045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3162 + }, + { + "epoch": 1.2636835797043549, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3163 + }, + { + "epoch": 1.2640831002796644, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3164 + }, + { + "epoch": 1.264482620854974, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3165 + }, + { + "epoch": 1.2648821414302835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3166 + }, + { + "epoch": 1.2652816620055933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3167 + }, + { + "epoch": 1.265681182580903, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3168 + }, + { + "epoch": 1.2660807031562125, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3169 + }, + { + "epoch": 1.2664802237315222, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3170 + }, + { + "epoch": 1.2668797443068318, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 3171 + }, + { + "epoch": 1.2672792648821414, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3172 + }, + { + "epoch": 1.2676787854574512, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3173 + }, + { + "epoch": 1.2680783060327607, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3174 + }, + { + "epoch": 1.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3175 + }, + { + "epoch": 1.26887734718338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3176 + }, + { + "epoch": 1.2692768677586896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3177 + }, + { + "epoch": 1.2696763883339992, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3178 + }, + { + "epoch": 1.2700759089093088, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3179 + }, + { + "epoch": 1.2704754294846183, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3180 + }, + { + "epoch": 1.2708749500599281, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3181 + }, + { + "epoch": 1.2712744706352377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3182 + }, + { + "epoch": 1.2716739912105472, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3183 + }, + { + "epoch": 1.272073511785857, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3184 + }, + { + "epoch": 1.2724730323611666, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3185 + }, + { + "epoch": 1.2728725529364762, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3186 + }, + { + "epoch": 1.273272073511786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3187 + }, + { + "epoch": 1.2736715940870955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3188 + }, + { + "epoch": 1.274071114662405, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3189 + }, + { + "epoch": 1.2744706352377149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3190 + }, + { + "epoch": 1.2748701558130244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3191 + }, + { + "epoch": 1.275269676388334, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3192 + }, + { + "epoch": 1.2756691969636436, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3193 + }, + { + "epoch": 1.2760687175389531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3194 + }, + { + "epoch": 1.276468238114263, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3195 + }, + { + "epoch": 1.2768677586895725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3196 + }, + { + "epoch": 1.277267279264882, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3197 + }, + { + "epoch": 1.2776667998401918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3198 + }, + { + "epoch": 1.2780663204155014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3199 + }, + { + "epoch": 1.278465840990811, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3200 + }, + { + "epoch": 1.2788653615661207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3201 + }, + { + "epoch": 1.2792648821414303, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3202 + }, + { + "epoch": 1.2796644027167399, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3203 + }, + { + "epoch": 1.2800639232920497, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 3204 + }, + { + "epoch": 1.2804634438673592, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3205 + }, + { + "epoch": 1.2808629644426688, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 3206 + }, + { + "epoch": 1.2812624850179783, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3207 + }, + { + "epoch": 1.2816620055932881, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 3208 + }, + { + "epoch": 1.2820615261685977, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3209 + }, + { + "epoch": 1.2824610467439073, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3210 + }, + { + "epoch": 1.2828605673192168, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3211 + }, + { + "epoch": 1.2832600878945266, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3212 + }, + { + "epoch": 1.2836596084698362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3213 + }, + { + "epoch": 1.2840591290451457, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3214 + }, + { + "epoch": 1.2844586496204555, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3215 + }, + { + "epoch": 1.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3216 + }, + { + "epoch": 1.2852576907710747, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3217 + }, + { + "epoch": 1.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3218 + }, + { + "epoch": 1.286056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3219 + }, + { + "epoch": 1.2864562524970036, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 3220 + }, + { + "epoch": 1.2868557730723134, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3221 + }, + { + "epoch": 1.287255293647623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3222 + }, + { + "epoch": 1.2876548142229325, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3223 + }, + { + "epoch": 1.288054334798242, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3224 + }, + { + "epoch": 1.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3225 + }, + { + "epoch": 1.2888533759488614, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3226 + }, + { + "epoch": 1.289252896524171, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3227 + }, + { + "epoch": 1.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3228 + }, + { + "epoch": 1.2900519376747903, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3229 + }, + { + "epoch": 1.2904514582500999, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3230 + }, + { + "epoch": 1.2908509788254094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3231 + }, + { + "epoch": 1.2912504994007192, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3232 + }, + { + "epoch": 1.2916500199760288, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3233 + }, + { + "epoch": 1.2920495405513384, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3234 + }, + { + "epoch": 1.2924490611266481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3235 + }, + { + "epoch": 1.2928485817019577, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3236 + }, + { + "epoch": 1.2932481022772673, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3237 + }, + { + "epoch": 1.2936476228525768, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3238 + }, + { + "epoch": 1.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3239 + }, + { + "epoch": 1.2944466640031962, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 3240 + }, + { + "epoch": 1.2948461845785058, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3241 + }, + { + "epoch": 1.2952457051538153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3242 + }, + { + "epoch": 1.295645225729125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3243 + }, + { + "epoch": 1.2960447463044347, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3244 + }, + { + "epoch": 1.2964442668797442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 3245 + }, + { + "epoch": 1.296843787455054, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3246 + }, + { + "epoch": 1.2972433080303636, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3247 + }, + { + "epoch": 1.2976428286056731, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3248 + }, + { + "epoch": 1.298042349180983, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3249 + }, + { + "epoch": 1.2984418697562925, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3250 + }, + { + "epoch": 1.298841390331602, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3251 + }, + { + "epoch": 1.2992409109069116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3252 + }, + { + "epoch": 1.2996404314822214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3253 + }, + { + "epoch": 1.300039952057531, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3254 + }, + { + "epoch": 1.3004394726328405, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3255 + }, + { + "epoch": 1.30083899320815, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3256 + }, + { + "epoch": 1.30123851378346, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3257 + }, + { + "epoch": 1.3016380343587695, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3258 + }, + { + "epoch": 1.302037554934079, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3259 + }, + { + "epoch": 1.3024370755093888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3260 + }, + { + "epoch": 1.3028365960846984, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3261 + }, + { + "epoch": 1.303236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3262 + }, + { + "epoch": 1.3036356372353177, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3263 + }, + { + "epoch": 1.3040351578106273, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3264 + }, + { + "epoch": 1.3044346783859369, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3265 + }, + { + "epoch": 1.3048341989612466, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3266 + }, + { + "epoch": 1.3052337195365562, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3267 + }, + { + "epoch": 1.3056332401118658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3268 + }, + { + "epoch": 1.3060327606871753, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3269 + }, + { + "epoch": 1.306432281262485, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3270 + }, + { + "epoch": 1.3068318018377947, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3271 + }, + { + "epoch": 1.3072313224131042, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3272 + }, + { + "epoch": 1.3076308429884138, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3273 + }, + { + "epoch": 1.3080303635637236, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3274 + }, + { + "epoch": 1.3084298841390332, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3275 + }, + { + "epoch": 1.3088294047143427, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3276 + }, + { + "epoch": 1.3092289252896525, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3277 + }, + { + "epoch": 1.309628445864962, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3278 + }, + { + "epoch": 1.3100279664402716, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3279 + }, + { + "epoch": 1.3104274870155814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3280 + }, + { + "epoch": 1.310827007590891, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3281 + }, + { + "epoch": 1.3112265281662006, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 3282 + }, + { + "epoch": 1.3116260487415101, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3283 + }, + { + "epoch": 1.3120255693168197, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3284 + }, + { + "epoch": 1.3124250898921295, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3285 + }, + { + "epoch": 1.312824610467439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3286 + }, + { + "epoch": 1.3132241310427486, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3287 + }, + { + "epoch": 1.3136236516180584, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3288 + }, + { + "epoch": 1.314023172193368, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3289 + }, + { + "epoch": 1.3144226927686775, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3290 + }, + { + "epoch": 1.3148222133439873, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3291 + }, + { + "epoch": 1.3152217339192969, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3292 + }, + { + "epoch": 1.3156212544946064, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3293 + }, + { + "epoch": 1.3160207750699162, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3294 + }, + { + "epoch": 1.3164202956452258, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3295 + }, + { + "epoch": 1.3168198162205353, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3296 + }, + { + "epoch": 1.317219336795845, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3297 + }, + { + "epoch": 1.3176188573711547, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3298 + }, + { + "epoch": 1.3180183779464643, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3299 + }, + { + "epoch": 1.3184178985217738, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3300 + }, + { + "epoch": 1.3188174190970834, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3301 + }, + { + "epoch": 1.3192169396723932, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3302 + }, + { + "epoch": 1.3196164602477027, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3303 + }, + { + "epoch": 1.3200159808230123, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3304 + }, + { + "epoch": 1.320415501398322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3305 + }, + { + "epoch": 1.3208150219736317, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3306 + }, + { + "epoch": 1.3212145425489412, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3307 + }, + { + "epoch": 1.321614063124251, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3308 + }, + { + "epoch": 1.3220135836995606, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3309 + }, + { + "epoch": 1.3224131042748701, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3310 + }, + { + "epoch": 1.32281262485018, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3311 + }, + { + "epoch": 1.3232121454254895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3312 + }, + { + "epoch": 1.323611666000799, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3313 + }, + { + "epoch": 1.3240111865761086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3314 + }, + { + "epoch": 1.3244107071514182, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3315 + }, + { + "epoch": 1.324810227726728, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3316 + }, + { + "epoch": 1.3252097483020375, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3317 + }, + { + "epoch": 1.325609268877347, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3318 + }, + { + "epoch": 1.3260087894526569, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3319 + }, + { + "epoch": 1.3264083100279664, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3320 + }, + { + "epoch": 1.326807830603276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3321 + }, + { + "epoch": 1.3272073511785858, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3322 + }, + { + "epoch": 1.3276068717538954, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3323 + }, + { + "epoch": 1.328006392329205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3324 + }, + { + "epoch": 1.3284059129045147, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3325 + }, + { + "epoch": 1.3288054334798243, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3326 + }, + { + "epoch": 1.3292049540551338, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3327 + }, + { + "epoch": 1.3296044746304434, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3328 + }, + { + "epoch": 1.330003995205753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3329 + }, + { + "epoch": 1.3304035157810628, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3330 + }, + { + "epoch": 1.3308030363563723, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3331 + }, + { + "epoch": 1.3312025569316819, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3332 + }, + { + "epoch": 1.3316020775069917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3333 + }, + { + "epoch": 1.3320015980823012, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3334 + }, + { + "epoch": 1.3324011186576108, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3335 + }, + { + "epoch": 1.3328006392329206, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3336 + }, + { + "epoch": 1.3332001598082301, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3337 + }, + { + "epoch": 1.3335996803835397, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3338 + }, + { + "epoch": 1.3339992009588495, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3339 + }, + { + "epoch": 1.334398721534159, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3340 + }, + { + "epoch": 1.3347982421094686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3341 + }, + { + "epoch": 1.3351977626847782, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3342 + }, + { + "epoch": 1.335597283260088, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 3343 + }, + { + "epoch": 1.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3344 + }, + { + "epoch": 1.336396324410707, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3345 + }, + { + "epoch": 1.3367958449860167, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 3346 + }, + { + "epoch": 1.3371953655613265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3347 + }, + { + "epoch": 1.337594886136636, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3348 + }, + { + "epoch": 1.3379944067119456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3349 + }, + { + "epoch": 1.3383939272872554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3350 + }, + { + "epoch": 1.338793447862565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3351 + }, + { + "epoch": 1.3391929684378745, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3352 + }, + { + "epoch": 1.3395924890131843, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3353 + }, + { + "epoch": 1.3399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3354 + }, + { + "epoch": 1.3403915301638034, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 3355 + }, + { + "epoch": 1.3407910507391132, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3356 + }, + { + "epoch": 1.3411905713144228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3357 + }, + { + "epoch": 1.3415900918897323, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3358 + }, + { + "epoch": 1.341989612465042, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3359 + }, + { + "epoch": 1.3423891330403515, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3360 + }, + { + "epoch": 1.3427886536156612, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3361 + }, + { + "epoch": 1.3431881741909708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3362 + }, + { + "epoch": 1.3435876947662804, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3363 + }, + { + "epoch": 1.3439872153415902, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3364 + }, + { + "epoch": 1.3443867359168997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3365 + }, + { + "epoch": 1.3447862564922093, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3366 + }, + { + "epoch": 1.345185777067519, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3367 + }, + { + "epoch": 1.3455852976428286, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3368 + }, + { + "epoch": 1.3459848182181382, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3369 + }, + { + "epoch": 1.346384338793448, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3370 + }, + { + "epoch": 1.3467838593687576, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 3371 + }, + { + "epoch": 1.3471833799440671, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3372 + }, + { + "epoch": 1.3475829005193767, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3373 + }, + { + "epoch": 1.3479824210946862, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3374 + }, + { + "epoch": 1.348381941669996, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 3375 + }, + { + "epoch": 1.3487814622453056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3376 + }, + { + "epoch": 1.3491809828206152, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3377 + }, + { + "epoch": 1.349580503395925, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 3378 + }, + { + "epoch": 1.3499800239712345, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 3379 + }, + { + "epoch": 1.350379544546544, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3380 + }, + { + "epoch": 1.3507790651218539, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3381 + }, + { + "epoch": 1.3511785856971634, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3382 + }, + { + "epoch": 1.351578106272473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3383 + }, + { + "epoch": 1.3519776268477828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3384 + }, + { + "epoch": 1.3523771474230923, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 3385 + }, + { + "epoch": 1.352776667998402, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3386 + }, + { + "epoch": 1.3531761885737115, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3387 + }, + { + "epoch": 1.3535757091490213, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3388 + }, + { + "epoch": 1.3539752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3389 + }, + { + "epoch": 1.3543747502996404, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3390 + }, + { + "epoch": 1.35477427087495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3391 + }, + { + "epoch": 1.3551737914502597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 3392 + }, + { + "epoch": 1.3555733120255693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3393 + }, + { + "epoch": 1.3559728326008789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3394 + }, + { + "epoch": 1.3563723531761886, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3395 + }, + { + "epoch": 1.3567718737514982, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3396 + }, + { + "epoch": 1.3571713943268078, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3397 + }, + { + "epoch": 1.3575709149021176, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3398 + }, + { + "epoch": 1.3579704354774271, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 3399 + }, + { + "epoch": 1.3583699560527367, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3400 + }, + { + "epoch": 1.3587694766280465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3401 + }, + { + "epoch": 1.359168997203356, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3402 + }, + { + "epoch": 1.3595685177786656, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3403 + }, + { + "epoch": 1.3599680383539752, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3404 + }, + { + "epoch": 1.3603675589292847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3405 + }, + { + "epoch": 1.3607670795045945, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 3406 + }, + { + "epoch": 1.361166600079904, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3407 + }, + { + "epoch": 1.3615661206552137, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3408 + }, + { + "epoch": 1.3619656412305234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3409 + }, + { + "epoch": 1.362365161805833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3410 + }, + { + "epoch": 1.3627646823811426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3411 + }, + { + "epoch": 1.3631642029564524, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 3412 + }, + { + "epoch": 1.363563723531762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3413 + }, + { + "epoch": 1.3639632441070715, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3414 + }, + { + "epoch": 1.3643627646823813, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3415 + }, + { + "epoch": 1.3647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3416 + }, + { + "epoch": 1.3651618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3417 + }, + { + "epoch": 1.36556132640831, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3418 + }, + { + "epoch": 1.3659608469836195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3419 + }, + { + "epoch": 1.3663603675589293, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3420 + }, + { + "epoch": 1.3667598881342389, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3421 + }, + { + "epoch": 1.3671594087095484, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3422 + }, + { + "epoch": 1.3675589292848582, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3423 + }, + { + "epoch": 1.3679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3424 + }, + { + "epoch": 1.3683579704354774, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3425 + }, + { + "epoch": 1.3687574910107871, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3426 + }, + { + "epoch": 1.3691570115860967, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3427 + }, + { + "epoch": 1.3695565321614063, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3428 + }, + { + "epoch": 1.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3429 + }, + { + "epoch": 1.3703555733120256, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3430 + }, + { + "epoch": 1.3707550938873352, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3431 + }, + { + "epoch": 1.3711546144626447, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3432 + }, + { + "epoch": 1.3715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3433 + }, + { + "epoch": 1.371953655613264, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 3434 + }, + { + "epoch": 1.3723531761885737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3435 + }, + { + "epoch": 1.3727526967638832, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3436 + }, + { + "epoch": 1.373152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 3437 + }, + { + "epoch": 1.3735517379145026, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3438 + }, + { + "epoch": 1.3739512584898121, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3439 + }, + { + "epoch": 1.374350779065122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3440 + }, + { + "epoch": 1.3747502996404315, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3441 + }, + { + "epoch": 1.375149820215741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3442 + }, + { + "epoch": 1.3755493407910508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3443 + }, + { + "epoch": 1.3759488613663604, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3444 + }, + { + "epoch": 1.37634838194167, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3445 + }, + { + "epoch": 1.3767479025169798, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3446 + }, + { + "epoch": 1.3771474230922893, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3447 + }, + { + "epoch": 1.3775469436675989, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3448 + }, + { + "epoch": 1.3779464642429085, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3449 + }, + { + "epoch": 1.378345984818218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3450 + }, + { + "epoch": 1.3787455053935278, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3451 + }, + { + "epoch": 1.3791450259688374, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3452 + }, + { + "epoch": 1.379544546544147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3453 + }, + { + "epoch": 1.3799440671194567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3454 + }, + { + "epoch": 1.3803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3455 + }, + { + "epoch": 1.3807431082700758, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3456 + }, + { + "epoch": 1.3811426288453856, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3457 + }, + { + "epoch": 1.3815421494206952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3458 + }, + { + "epoch": 1.3819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3459 + }, + { + "epoch": 1.3823411905713145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3460 + }, + { + "epoch": 1.3827407111466241, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3461 + }, + { + "epoch": 1.3831402317219337, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3462 + }, + { + "epoch": 1.3835397522972432, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3463 + }, + { + "epoch": 1.3839392728725528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3464 + }, + { + "epoch": 1.3843387934478626, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3465 + }, + { + "epoch": 1.3847383140231722, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3466 + }, + { + "epoch": 1.3851378345984817, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3467 + }, + { + "epoch": 1.3855373551737915, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3468 + }, + { + "epoch": 1.385936875749101, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3469 + }, + { + "epoch": 1.3863363963244106, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3470 + }, + { + "epoch": 1.3867359168997204, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3471 + }, + { + "epoch": 1.38713543747503, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3472 + }, + { + "epoch": 1.3875349580503396, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3473 + }, + { + "epoch": 1.3879344786256493, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3474 + }, + { + "epoch": 1.388333999200959, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3475 + }, + { + "epoch": 1.3887335197762685, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3476 + }, + { + "epoch": 1.389133040351578, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3477 + }, + { + "epoch": 1.3895325609268878, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3478 + }, + { + "epoch": 1.3899320815021974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3479 + }, + { + "epoch": 1.390331602077507, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3480 + }, + { + "epoch": 1.3907311226528165, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3481 + }, + { + "epoch": 1.3911306432281263, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3482 + }, + { + "epoch": 1.3915301638034359, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3483 + }, + { + "epoch": 1.3919296843787454, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3484 + }, + { + "epoch": 1.3923292049540552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3485 + }, + { + "epoch": 1.3927287255293648, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3486 + }, + { + "epoch": 1.3931282461046743, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3487 + }, + { + "epoch": 1.3935277666799841, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3488 + }, + { + "epoch": 1.3939272872552937, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3489 + }, + { + "epoch": 1.3943268078306033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3490 + }, + { + "epoch": 1.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3491 + }, + { + "epoch": 1.3951258489812226, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3492 + }, + { + "epoch": 1.3955253695565322, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3493 + }, + { + "epoch": 1.3959248901318417, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3494 + }, + { + "epoch": 1.3963244107071513, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3495 + }, + { + "epoch": 1.396723931282461, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3496 + }, + { + "epoch": 1.3971234518577706, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3497 + }, + { + "epoch": 1.3975229724330802, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3498 + }, + { + "epoch": 1.39792249300839, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3499 + }, + { + "epoch": 1.3983220135836996, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3500 + }, + { + "epoch": 1.3987215341590091, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3501 + }, + { + "epoch": 1.399121054734319, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3502 + }, + { + "epoch": 1.3995205753096285, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3503 + }, + { + "epoch": 1.399920095884938, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3504 + }, + { + "epoch": 1.4003196164602478, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3505 + }, + { + "epoch": 1.4007191370355574, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3506 + }, + { + "epoch": 1.401118657610867, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3507 + }, + { + "epoch": 1.4015181781861765, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3508 + }, + { + "epoch": 1.401917698761486, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3509 + }, + { + "epoch": 1.4023172193367959, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3510 + }, + { + "epoch": 1.4027167399121054, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3511 + }, + { + "epoch": 1.403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3512 + }, + { + "epoch": 1.4035157810627248, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3513 + }, + { + "epoch": 1.4039153016380344, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3514 + }, + { + "epoch": 1.404314822213344, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3515 + }, + { + "epoch": 1.4047143427886537, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3516 + }, + { + "epoch": 1.4051138633639633, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3517 + }, + { + "epoch": 1.4055133839392728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3518 + }, + { + "epoch": 1.4059129045145826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3519 + }, + { + "epoch": 1.4063124250898922, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3520 + }, + { + "epoch": 1.4067119456652017, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3521 + }, + { + "epoch": 1.4071114662405113, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3522 + }, + { + "epoch": 1.407510986815821, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3523 + }, + { + "epoch": 1.4079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3524 + }, + { + "epoch": 1.4083100279664402, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3525 + }, + { + "epoch": 1.4087095485417498, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3526 + }, + { + "epoch": 1.4091090691170596, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3527 + }, + { + "epoch": 1.4095085896923691, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3528 + }, + { + "epoch": 1.4099081102676787, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3529 + }, + { + "epoch": 1.4103076308429885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 3530 + }, + { + "epoch": 1.410707151418298, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3531 + }, + { + "epoch": 1.4111066719936076, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 3532 + }, + { + "epoch": 1.4115061925689174, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3533 + }, + { + "epoch": 1.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3534 + }, + { + "epoch": 1.4123052337195365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3535 + }, + { + "epoch": 1.4127047542948463, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3536 + }, + { + "epoch": 1.4131042748701559, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3537 + }, + { + "epoch": 1.4135037954454654, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3538 + }, + { + "epoch": 1.413903316020775, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3539 + }, + { + "epoch": 1.4143028365960846, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3540 + }, + { + "epoch": 1.4147023571713944, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3541 + }, + { + "epoch": 1.415101877746704, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3542 + }, + { + "epoch": 1.4155013983220135, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3543 + }, + { + "epoch": 1.4159009188973233, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3544 + }, + { + "epoch": 1.4163004394726328, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3545 + }, + { + "epoch": 1.4166999600479424, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3546 + }, + { + "epoch": 1.4170994806232522, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3547 + }, + { + "epoch": 1.4174990011985618, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3548 + }, + { + "epoch": 1.4178985217738713, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3549 + }, + { + "epoch": 1.418298042349181, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 3550 + }, + { + "epoch": 1.4186975629244907, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3551 + }, + { + "epoch": 1.4190970834998002, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3552 + }, + { + "epoch": 1.4194966040751098, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3553 + }, + { + "epoch": 1.4198961246504194, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3554 + }, + { + "epoch": 1.4202956452257292, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3555 + }, + { + "epoch": 1.4206951658010387, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3556 + }, + { + "epoch": 1.4210946863763483, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3557 + }, + { + "epoch": 1.421494206951658, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3558 + }, + { + "epoch": 1.4218937275269676, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3559 + }, + { + "epoch": 1.4222932481022772, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3560 + }, + { + "epoch": 1.422692768677587, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3561 + }, + { + "epoch": 1.4230922892528965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3562 + }, + { + "epoch": 1.4234918098282061, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3563 + }, + { + "epoch": 1.423891330403516, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3564 + }, + { + "epoch": 1.4242908509788255, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3565 + }, + { + "epoch": 1.424690371554135, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3566 + }, + { + "epoch": 1.4250898921294446, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3567 + }, + { + "epoch": 1.4254894127047544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3568 + }, + { + "epoch": 1.425888933280064, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3569 + }, + { + "epoch": 1.4262884538553735, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3570 + }, + { + "epoch": 1.426687974430683, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3571 + }, + { + "epoch": 1.4270874950059929, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3572 + }, + { + "epoch": 1.4274870155813024, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3573 + }, + { + "epoch": 1.427886536156612, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3574 + }, + { + "epoch": 1.4282860567319218, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3575 + }, + { + "epoch": 1.4286855773072313, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3576 + }, + { + "epoch": 1.429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 3577 + }, + { + "epoch": 1.4294846184578507, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3578 + }, + { + "epoch": 1.4298841390331603, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3579 + }, + { + "epoch": 1.4302836596084698, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3580 + }, + { + "epoch": 1.4306831801837796, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3581 + }, + { + "epoch": 1.4310827007590892, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3582 + }, + { + "epoch": 1.4314822213343987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3583 + }, + { + "epoch": 1.4318817419097083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3584 + }, + { + "epoch": 1.4322812624850179, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3585 + }, + { + "epoch": 1.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3586 + }, + { + "epoch": 1.4330803036356372, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3587 + }, + { + "epoch": 1.4334798242109468, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3588 + }, + { + "epoch": 1.4338793447862566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3589 + }, + { + "epoch": 1.4342788653615661, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3590 + }, + { + "epoch": 1.4346783859368757, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3591 + }, + { + "epoch": 1.4350779065121855, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3592 + }, + { + "epoch": 1.435477427087495, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3593 + }, + { + "epoch": 1.4358769476628046, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3594 + }, + { + "epoch": 1.4362764682381144, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 3595 + }, + { + "epoch": 1.436675988813424, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3596 + }, + { + "epoch": 1.4370755093887335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3597 + }, + { + "epoch": 1.437475029964043, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3598 + }, + { + "epoch": 1.4378745505393526, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3599 + }, + { + "epoch": 1.4382740711146624, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3600 + }, + { + "epoch": 1.438673591689972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3601 + }, + { + "epoch": 1.4390731122652816, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3602 + }, + { + "epoch": 1.4394726328405913, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3603 + }, + { + "epoch": 1.439872153415901, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3604 + }, + { + "epoch": 1.4402716739912105, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3605 + }, + { + "epoch": 1.4406711945665203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3606 + }, + { + "epoch": 1.4410707151418298, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3607 + }, + { + "epoch": 1.4414702357171394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3608 + }, + { + "epoch": 1.4418697562924492, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3609 + }, + { + "epoch": 1.4422692768677587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3610 + }, + { + "epoch": 1.4426687974430683, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3611 + }, + { + "epoch": 1.4430683180183779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3612 + }, + { + "epoch": 1.4434678385936877, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3613 + }, + { + "epoch": 1.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3614 + }, + { + "epoch": 1.4442668797443068, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3615 + }, + { + "epoch": 1.4446664003196164, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 3616 + }, + { + "epoch": 1.4450659208949261, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3617 + }, + { + "epoch": 1.4454654414702357, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3618 + }, + { + "epoch": 1.4458649620455453, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3619 + }, + { + "epoch": 1.446264482620855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3620 + }, + { + "epoch": 1.4466640031961646, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3621 + }, + { + "epoch": 1.4470635237714742, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3622 + }, + { + "epoch": 1.447463044346784, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3623 + }, + { + "epoch": 1.4478625649220935, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3624 + }, + { + "epoch": 1.448262085497403, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3625 + }, + { + "epoch": 1.4486616060727129, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3626 + }, + { + "epoch": 1.4490611266480224, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 3627 + }, + { + "epoch": 1.449460647223332, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3628 + }, + { + "epoch": 1.4498601677986416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3629 + }, + { + "epoch": 1.4502596883739511, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3630 + }, + { + "epoch": 1.450659208949261, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3631 + }, + { + "epoch": 1.4510587295245705, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3632 + }, + { + "epoch": 1.45145825009988, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3633 + }, + { + "epoch": 1.4518577706751898, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3634 + }, + { + "epoch": 1.4522572912504994, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3635 + }, + { + "epoch": 1.452656811825809, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3636 + }, + { + "epoch": 1.4530563324011188, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3637 + }, + { + "epoch": 1.4534558529764283, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3638 + }, + { + "epoch": 1.4538553735517379, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3639 + }, + { + "epoch": 1.4542548941270477, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3640 + }, + { + "epoch": 1.4546544147023572, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3641 + }, + { + "epoch": 1.4550539352776668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3642 + }, + { + "epoch": 1.4554534558529764, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3643 + }, + { + "epoch": 1.455852976428286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3644 + }, + { + "epoch": 1.4562524970035957, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3645 + }, + { + "epoch": 1.4566520175789053, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 3646 + }, + { + "epoch": 1.4570515381542148, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3647 + }, + { + "epoch": 1.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3648 + }, + { + "epoch": 1.4578505793048342, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3649 + }, + { + "epoch": 1.4582500998801438, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3650 + }, + { + "epoch": 1.4586496204554535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3651 + }, + { + "epoch": 1.459049141030763, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3652 + }, + { + "epoch": 1.4594486616060727, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3653 + }, + { + "epoch": 1.4598481821813825, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3654 + }, + { + "epoch": 1.460247702756692, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 3655 + }, + { + "epoch": 1.4606472233320016, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3656 + }, + { + "epoch": 1.4610467439073112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3657 + }, + { + "epoch": 1.461446264482621, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3658 + }, + { + "epoch": 1.4618457850579305, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3659 + }, + { + "epoch": 1.46224530563324, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3660 + }, + { + "epoch": 1.4626448262085496, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3661 + }, + { + "epoch": 1.4630443467838594, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3662 + }, + { + "epoch": 1.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3663 + }, + { + "epoch": 1.4638433879344785, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3664 + }, + { + "epoch": 1.4642429085097883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3665 + }, + { + "epoch": 1.464642429085098, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3666 + }, + { + "epoch": 1.4650419496604075, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3667 + }, + { + "epoch": 1.4654414702357172, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 3668 + }, + { + "epoch": 1.4658409908110268, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3669 + }, + { + "epoch": 1.4662405113863364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3670 + }, + { + "epoch": 1.4666400319616462, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3671 + }, + { + "epoch": 1.4670395525369557, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3672 + }, + { + "epoch": 1.4674390731122653, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3673 + }, + { + "epoch": 1.4678385936875749, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3674 + }, + { + "epoch": 1.4682381142628844, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3675 + }, + { + "epoch": 1.4686376348381942, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 3676 + }, + { + "epoch": 1.4690371554135038, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 3677 + }, + { + "epoch": 1.4694366759888133, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3678 + }, + { + "epoch": 1.4698361965641231, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3679 + }, + { + "epoch": 1.4702357171394327, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3680 + }, + { + "epoch": 1.4706352377147422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 3681 + }, + { + "epoch": 1.471034758290052, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3682 + }, + { + "epoch": 1.4714342788653616, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3683 + }, + { + "epoch": 1.4718337994406712, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3684 + }, + { + "epoch": 1.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3685 + }, + { + "epoch": 1.4726328405912905, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3686 + }, + { + "epoch": 1.4730323611666, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3687 + }, + { + "epoch": 1.4734318817419096, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3688 + }, + { + "epoch": 1.4738314023172192, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3689 + }, + { + "epoch": 1.474230922892529, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3690 + }, + { + "epoch": 1.4746304434678386, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3691 + }, + { + "epoch": 1.4750299640431481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3692 + }, + { + "epoch": 1.475429484618458, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3693 + }, + { + "epoch": 1.4758290051937675, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3694 + }, + { + "epoch": 1.476228525769077, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3695 + }, + { + "epoch": 1.4766280463443868, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3696 + }, + { + "epoch": 1.4770275669196964, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3697 + }, + { + "epoch": 1.477427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3698 + }, + { + "epoch": 1.4778266080703157, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3699 + }, + { + "epoch": 1.4782261286456253, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3700 + }, + { + "epoch": 1.4786256492209349, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3701 + }, + { + "epoch": 1.4790251697962444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3702 + }, + { + "epoch": 1.479424690371554, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3703 + }, + { + "epoch": 1.4798242109468638, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3704 + }, + { + "epoch": 1.4802237315221733, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3705 + }, + { + "epoch": 1.480623252097483, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3706 + }, + { + "epoch": 1.4810227726727927, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3707 + }, + { + "epoch": 1.4814222932481023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3708 + }, + { + "epoch": 1.4818218138234118, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 3709 + }, + { + "epoch": 1.4822213343987216, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 3710 + }, + { + "epoch": 1.4826208549740312, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3711 + }, + { + "epoch": 1.4830203755493407, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 3712 + }, + { + "epoch": 1.4834198961246505, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3713 + }, + { + "epoch": 1.48381941669996, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3714 + }, + { + "epoch": 1.4842189372752697, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3715 + }, + { + "epoch": 1.4846184578505794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3716 + }, + { + "epoch": 1.485017978425889, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3717 + }, + { + "epoch": 1.4854174990011986, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3718 + }, + { + "epoch": 1.4858170195765081, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3719 + }, + { + "epoch": 1.4862165401518177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3720 + }, + { + "epoch": 1.4866160607271275, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3721 + }, + { + "epoch": 1.487015581302437, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3722 + }, + { + "epoch": 1.4874151018777466, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3723 + }, + { + "epoch": 1.4878146224530564, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3724 + }, + { + "epoch": 1.488214143028366, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3725 + }, + { + "epoch": 1.4886136636036755, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3726 + }, + { + "epoch": 1.4890131841789853, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3727 + }, + { + "epoch": 1.4894127047542949, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3728 + }, + { + "epoch": 1.4898122253296044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3729 + }, + { + "epoch": 1.4902117459049142, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3730 + }, + { + "epoch": 1.4906112664802238, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3731 + }, + { + "epoch": 1.4910107870555334, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3732 + }, + { + "epoch": 1.491410307630843, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3733 + }, + { + "epoch": 1.4918098282061525, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3734 + }, + { + "epoch": 1.4922093487814623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3735 + }, + { + "epoch": 1.4926088693567718, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3736 + }, + { + "epoch": 1.4930083899320814, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3737 + }, + { + "epoch": 1.4934079105073912, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3738 + }, + { + "epoch": 1.4938074310827008, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3739 + }, + { + "epoch": 1.4942069516580103, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3740 + }, + { + "epoch": 1.49460647223332, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3741 + }, + { + "epoch": 1.4950059928086297, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3742 + }, + { + "epoch": 1.4954055133839392, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3743 + }, + { + "epoch": 1.495805033959249, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3744 + }, + { + "epoch": 1.4962045545345586, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3745 + }, + { + "epoch": 1.4966040751098681, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3746 + }, + { + "epoch": 1.4970035956851777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3747 + }, + { + "epoch": 1.4974031162604873, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3748 + }, + { + "epoch": 1.497802636835797, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3749 + }, + { + "epoch": 1.4982021574111066, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3750 + }, + { + "epoch": 1.4986016779864162, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3751 + }, + { + "epoch": 1.499001198561726, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3752 + }, + { + "epoch": 1.4994007191370355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3753 + }, + { + "epoch": 1.499800239712345, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3754 + }, + { + "epoch": 1.500199760287655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3755 + }, + { + "epoch": 1.5005992808629645, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3756 + }, + { + "epoch": 1.500998801438274, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3757 + }, + { + "epoch": 1.5013983220135838, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3758 + }, + { + "epoch": 1.5017978425888934, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3759 + }, + { + "epoch": 1.502197363164203, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 3760 + }, + { + "epoch": 1.5025968837395127, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3761 + }, + { + "epoch": 1.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3762 + }, + { + "epoch": 1.5033959248901319, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3763 + }, + { + "epoch": 1.5037954454654414, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3764 + }, + { + "epoch": 1.504194966040751, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3765 + }, + { + "epoch": 1.5045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3766 + }, + { + "epoch": 1.5049940071913703, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3767 + }, + { + "epoch": 1.50539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3768 + }, + { + "epoch": 1.5057930483419897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3769 + }, + { + "epoch": 1.5061925689172992, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3770 + }, + { + "epoch": 1.5065920894926088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3771 + }, + { + "epoch": 1.5069916100679186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3772 + }, + { + "epoch": 1.5073911306432282, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3773 + }, + { + "epoch": 1.5077906512185377, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3774 + }, + { + "epoch": 1.5081901717938475, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3775 + }, + { + "epoch": 1.5085896923691569, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3776 + }, + { + "epoch": 1.5089892129444666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3777 + }, + { + "epoch": 1.5093887335197764, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3778 + }, + { + "epoch": 1.5097882540950858, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3779 + }, + { + "epoch": 1.5101877746703956, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3780 + }, + { + "epoch": 1.5105872952457051, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3781 + }, + { + "epoch": 1.5109868158210147, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3782 + }, + { + "epoch": 1.5113863363963245, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3783 + }, + { + "epoch": 1.511785856971634, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3784 + }, + { + "epoch": 1.5121853775469436, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3785 + }, + { + "epoch": 1.5125848981222534, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3786 + }, + { + "epoch": 1.512984418697563, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3787 + }, + { + "epoch": 1.5133839392728725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3788 + }, + { + "epoch": 1.5137834598481823, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3789 + }, + { + "epoch": 1.5141829804234919, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3790 + }, + { + "epoch": 1.5145825009988014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3791 + }, + { + "epoch": 1.5149820215741112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3792 + }, + { + "epoch": 1.5153815421494206, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3793 + }, + { + "epoch": 1.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3794 + }, + { + "epoch": 1.51618058330004, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3795 + }, + { + "epoch": 1.5165801038753495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3796 + }, + { + "epoch": 1.5169796244506593, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3797 + }, + { + "epoch": 1.5173791450259688, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3798 + }, + { + "epoch": 1.5177786656012784, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3799 + }, + { + "epoch": 1.5181781861765882, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3800 + }, + { + "epoch": 1.5185777067518977, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3801 + }, + { + "epoch": 1.5189772273272073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3802 + }, + { + "epoch": 1.519376747902517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3803 + }, + { + "epoch": 1.5197762684778267, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3804 + }, + { + "epoch": 1.5201757890531362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3805 + }, + { + "epoch": 1.520575309628446, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 3806 + }, + { + "epoch": 1.5209748302037553, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3807 + }, + { + "epoch": 1.5213743507790651, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3808 + }, + { + "epoch": 1.5217738713543747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3809 + }, + { + "epoch": 1.5221733919296843, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3810 + }, + { + "epoch": 1.522572912504994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3811 + }, + { + "epoch": 1.5229724330803036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3812 + }, + { + "epoch": 1.5233719536556132, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3813 + }, + { + "epoch": 1.523771474230923, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 3814 + }, + { + "epoch": 1.5241709948062325, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3815 + }, + { + "epoch": 1.524570515381542, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3816 + }, + { + "epoch": 1.5249700359568519, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3817 + }, + { + "epoch": 1.5253695565321614, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3818 + }, + { + "epoch": 1.525769077107471, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3819 + }, + { + "epoch": 1.5261685976827808, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3820 + }, + { + "epoch": 1.5265681182580901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 3821 + }, + { + "epoch": 1.5269676388334, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3822 + }, + { + "epoch": 1.5273671594087097, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3823 + }, + { + "epoch": 1.527766679984019, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 3824 + }, + { + "epoch": 1.5281662005593288, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 3825 + }, + { + "epoch": 1.5285657211346384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3826 + }, + { + "epoch": 1.528965241709948, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3827 + }, + { + "epoch": 1.5293647622852578, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3828 + }, + { + "epoch": 1.5297642828605673, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3829 + }, + { + "epoch": 1.5301638034358769, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3830 + }, + { + "epoch": 1.5305633240111867, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3831 + }, + { + "epoch": 1.5309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 3832 + }, + { + "epoch": 1.5313623651618058, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3833 + }, + { + "epoch": 1.5317618857371156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 3834 + }, + { + "epoch": 1.5321614063124251, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3835 + }, + { + "epoch": 1.5325609268877347, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3836 + }, + { + "epoch": 1.5329604474630445, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3837 + }, + { + "epoch": 1.5333599680383538, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3838 + }, + { + "epoch": 1.5337594886136636, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3839 + }, + { + "epoch": 1.5341590091889732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 3840 + }, + { + "epoch": 1.5345585297642828, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3841 + }, + { + "epoch": 1.5349580503395925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3842 + }, + { + "epoch": 1.535357570914902, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3843 + }, + { + "epoch": 1.5357570914902117, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3844 + }, + { + "epoch": 1.5361566120655215, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 3845 + }, + { + "epoch": 1.536556132640831, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3846 + }, + { + "epoch": 1.5369556532161406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3847 + }, + { + "epoch": 1.5373551737914504, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3848 + }, + { + "epoch": 1.53775469436676, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3849 + }, + { + "epoch": 1.5381542149420695, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3850 + }, + { + "epoch": 1.5385537355173793, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3851 + }, + { + "epoch": 1.5389532560926886, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3852 + }, + { + "epoch": 1.5393527766679984, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3853 + }, + { + "epoch": 1.539752297243308, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3854 + }, + { + "epoch": 1.5401518178186175, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 3855 + }, + { + "epoch": 1.5405513383939273, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3856 + }, + { + "epoch": 1.540950858969237, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3857 + }, + { + "epoch": 1.5413503795445465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3858 + }, + { + "epoch": 1.5417499001198562, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3859 + }, + { + "epoch": 1.5421494206951658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3860 + }, + { + "epoch": 1.5425489412704754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3861 + }, + { + "epoch": 1.5429484618457852, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3862 + }, + { + "epoch": 1.5433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3863 + }, + { + "epoch": 1.5437475029964043, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3864 + }, + { + "epoch": 1.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3865 + }, + { + "epoch": 1.5445465441470234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3866 + }, + { + "epoch": 1.5449460647223332, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3867 + }, + { + "epoch": 1.545345585297643, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3868 + }, + { + "epoch": 1.5457451058729523, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3869 + }, + { + "epoch": 1.5461446264482621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 3870 + }, + { + "epoch": 1.5465441470235717, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3871 + }, + { + "epoch": 1.5469436675988812, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3872 + }, + { + "epoch": 1.547343188174191, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3873 + }, + { + "epoch": 1.5477427087495006, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3874 + }, + { + "epoch": 1.5481422293248102, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3875 + }, + { + "epoch": 1.54854174990012, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 3876 + }, + { + "epoch": 1.5489412704754295, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3877 + }, + { + "epoch": 1.549340791050739, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3878 + }, + { + "epoch": 1.5497403116260489, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3879 + }, + { + "epoch": 1.5501398322013584, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3880 + }, + { + "epoch": 1.550539352776668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3881 + }, + { + "epoch": 1.5509388733519778, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3882 + }, + { + "epoch": 1.5513383939272871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3883 + }, + { + "epoch": 1.551737914502597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 3884 + }, + { + "epoch": 1.5521374350779065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3885 + }, + { + "epoch": 1.552536955653216, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 3886 + }, + { + "epoch": 1.5529364762285258, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3887 + }, + { + "epoch": 1.5533359968038354, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3888 + }, + { + "epoch": 1.553735517379145, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3889 + }, + { + "epoch": 1.5541350379544547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3890 + }, + { + "epoch": 1.5545345585297643, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3891 + }, + { + "epoch": 1.5549340791050739, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3892 + }, + { + "epoch": 1.5553335996803836, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3893 + }, + { + "epoch": 1.5557331202556932, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3894 + }, + { + "epoch": 1.5561326408310028, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3895 + }, + { + "epoch": 1.5565321614063126, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3896 + }, + { + "epoch": 1.556931681981622, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3897 + }, + { + "epoch": 1.5573312025569317, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3898 + }, + { + "epoch": 1.5577307231322413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3899 + }, + { + "epoch": 1.5581302437075508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3900 + }, + { + "epoch": 1.5585297642828606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3901 + }, + { + "epoch": 1.5589292848581702, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3902 + }, + { + "epoch": 1.5593288054334797, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3903 + }, + { + "epoch": 1.5597283260087895, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3904 + }, + { + "epoch": 1.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3905 + }, + { + "epoch": 1.5605273671594087, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3906 + }, + { + "epoch": 1.5609268877347184, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3907 + }, + { + "epoch": 1.561326408310028, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3908 + }, + { + "epoch": 1.5617259288853376, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3909 + }, + { + "epoch": 1.5621254494606474, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3910 + }, + { + "epoch": 1.5625249700359567, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3911 + }, + { + "epoch": 1.5629244906112665, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3912 + }, + { + "epoch": 1.5633240111865763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3913 + }, + { + "epoch": 1.5637235317618856, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3914 + }, + { + "epoch": 1.5641230523371954, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3915 + }, + { + "epoch": 1.564522572912505, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3916 + }, + { + "epoch": 1.5649220934878145, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3917 + }, + { + "epoch": 1.5653216140631243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3918 + }, + { + "epoch": 1.5657211346384339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 3919 + }, + { + "epoch": 1.5661206552137434, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3920 + }, + { + "epoch": 1.5665201757890532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3921 + }, + { + "epoch": 1.5669196963643628, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 3922 + }, + { + "epoch": 1.5673192169396724, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3923 + }, + { + "epoch": 1.5677187375149821, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3924 + }, + { + "epoch": 1.5681182580902915, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3925 + }, + { + "epoch": 1.5685177786656013, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3926 + }, + { + "epoch": 1.568917299240911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3927 + }, + { + "epoch": 1.5693168198162204, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3928 + }, + { + "epoch": 1.5697163403915302, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3929 + }, + { + "epoch": 1.5701158609668397, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3930 + }, + { + "epoch": 1.5705153815421493, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3931 + }, + { + "epoch": 1.570914902117459, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3932 + }, + { + "epoch": 1.5713144226927687, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3933 + }, + { + "epoch": 1.5717139432680782, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3934 + }, + { + "epoch": 1.572113463843388, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3935 + }, + { + "epoch": 1.5725129844186976, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3936 + }, + { + "epoch": 1.5729125049940071, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3937 + }, + { + "epoch": 1.573312025569317, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 3938 + }, + { + "epoch": 1.5737115461446265, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3939 + }, + { + "epoch": 1.574111066719936, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3940 + }, + { + "epoch": 1.5745105872952458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3941 + }, + { + "epoch": 1.5749101078705552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3942 + }, + { + "epoch": 1.575309628445865, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3943 + }, + { + "epoch": 1.5757091490211745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3944 + }, + { + "epoch": 1.576108669596484, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3945 + }, + { + "epoch": 1.5765081901717939, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3946 + }, + { + "epoch": 1.5769077107471035, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3947 + }, + { + "epoch": 1.577307231322413, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3948 + }, + { + "epoch": 1.5777067518977228, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3949 + }, + { + "epoch": 1.5781062724730324, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3950 + }, + { + "epoch": 1.578505793048342, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3951 + }, + { + "epoch": 1.5789053136236517, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3952 + }, + { + "epoch": 1.5793048341989613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3953 + }, + { + "epoch": 1.5797043547742708, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3954 + }, + { + "epoch": 1.5801038753495806, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3955 + }, + { + "epoch": 1.58050339592489, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3956 + }, + { + "epoch": 1.5809029165001998, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 3957 + }, + { + "epoch": 1.5813024370755095, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3958 + }, + { + "epoch": 1.581701957650819, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 3959 + }, + { + "epoch": 1.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3960 + }, + { + "epoch": 1.5825009988014382, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3961 + }, + { + "epoch": 1.5829005193767478, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3962 + }, + { + "epoch": 1.5833000399520576, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3963 + }, + { + "epoch": 1.5836995605273672, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 3964 + }, + { + "epoch": 1.5840990811026767, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 3965 + }, + { + "epoch": 1.5844986016779865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3966 + }, + { + "epoch": 1.584898122253296, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 3967 + }, + { + "epoch": 1.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3968 + }, + { + "epoch": 1.5856971634039154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3969 + }, + { + "epoch": 1.5860966839792248, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3970 + }, + { + "epoch": 1.5864962045545346, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3971 + }, + { + "epoch": 1.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3972 + }, + { + "epoch": 1.5872952457051537, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3973 + }, + { + "epoch": 1.5876947662804635, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3974 + }, + { + "epoch": 1.588094286855773, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3975 + }, + { + "epoch": 1.5884938074310826, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3976 + }, + { + "epoch": 1.5888933280063924, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3977 + }, + { + "epoch": 1.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3978 + }, + { + "epoch": 1.5896923691570115, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 3979 + }, + { + "epoch": 1.5900918897323213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3980 + }, + { + "epoch": 1.5904914103076309, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3981 + }, + { + "epoch": 1.5908909308829404, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3982 + }, + { + "epoch": 1.5912904514582502, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3983 + }, + { + "epoch": 1.5916899720335598, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3984 + }, + { + "epoch": 1.5920894926088693, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3985 + }, + { + "epoch": 1.5924890131841791, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3986 + }, + { + "epoch": 1.5928885337594885, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 3987 + }, + { + "epoch": 1.5932880543347983, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3988 + }, + { + "epoch": 1.5936875749101078, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3989 + }, + { + "epoch": 1.5940870954854174, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3990 + }, + { + "epoch": 1.5944866160607272, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3991 + }, + { + "epoch": 1.5948861366360367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3992 + }, + { + "epoch": 1.5952856572113463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3993 + }, + { + "epoch": 1.595685177786656, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3994 + }, + { + "epoch": 1.5960846983619656, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3995 + }, + { + "epoch": 1.5964842189372752, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3996 + }, + { + "epoch": 1.596883739512585, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3997 + }, + { + "epoch": 1.5972832600878946, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3998 + }, + { + "epoch": 1.5976827806632041, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3999 + }, + { + "epoch": 1.598082301238514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4000 + }, + { + "epoch": 1.5984818218138233, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4001 + }, + { + "epoch": 1.598881342389133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4002 + }, + { + "epoch": 1.5992808629644428, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4003 + }, + { + "epoch": 1.5996803835397522, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4004 + }, + { + "epoch": 1.600079904115062, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4005 + }, + { + "epoch": 1.6004794246903715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4006 + }, + { + "epoch": 1.600878945265681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4007 + }, + { + "epoch": 1.6012784658409909, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4008 + }, + { + "epoch": 1.6016779864163004, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4009 + }, + { + "epoch": 1.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4010 + }, + { + "epoch": 1.6024770275669198, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4011 + }, + { + "epoch": 1.6028765481422294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4012 + }, + { + "epoch": 1.603276068717539, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4013 + }, + { + "epoch": 1.6036755892928487, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4014 + }, + { + "epoch": 1.604075109868158, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4015 + }, + { + "epoch": 1.6044746304434678, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 4016 + }, + { + "epoch": 1.6048741510187776, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4017 + }, + { + "epoch": 1.605273671594087, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4018 + }, + { + "epoch": 1.6056731921693967, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4019 + }, + { + "epoch": 1.6060727127447063, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4020 + }, + { + "epoch": 1.6064722333200159, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4021 + }, + { + "epoch": 1.6068717538953257, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4022 + }, + { + "epoch": 1.6072712744706352, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4023 + }, + { + "epoch": 1.6076707950459448, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4024 + }, + { + "epoch": 1.6080703156212546, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4025 + }, + { + "epoch": 1.6084698361965641, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4026 + }, + { + "epoch": 1.6088693567718737, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4027 + }, + { + "epoch": 1.6092688773471835, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4028 + }, + { + "epoch": 1.609668397922493, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4029 + }, + { + "epoch": 1.6100679184978026, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4030 + }, + { + "epoch": 1.6104674390731124, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 4031 + }, + { + "epoch": 1.6108669596484217, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4032 + }, + { + "epoch": 1.6112664802237315, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4033 + }, + { + "epoch": 1.611666000799041, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 4034 + }, + { + "epoch": 1.6120655213743507, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4035 + }, + { + "epoch": 1.6124650419496604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4036 + }, + { + "epoch": 1.61286456252497, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4037 + }, + { + "epoch": 1.6132640831002796, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 4038 + }, + { + "epoch": 1.6136636036755894, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4039 + }, + { + "epoch": 1.614063124250899, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4040 + }, + { + "epoch": 1.6144626448262085, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4041 + }, + { + "epoch": 1.6148621654015183, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4042 + }, + { + "epoch": 1.6152616859768278, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4043 + }, + { + "epoch": 1.6156612065521374, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4044 + }, + { + "epoch": 1.6160607271274472, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4045 + }, + { + "epoch": 1.6164602477027565, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4046 + }, + { + "epoch": 1.6168597682780663, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4047 + }, + { + "epoch": 1.617259288853376, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4048 + }, + { + "epoch": 1.6176588094286855, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4049 + }, + { + "epoch": 1.6180583300039952, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4050 + }, + { + "epoch": 1.6184578505793048, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4051 + }, + { + "epoch": 1.6188573711546144, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4052 + }, + { + "epoch": 1.6192568917299242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4053 + }, + { + "epoch": 1.6196564123052337, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4054 + }, + { + "epoch": 1.6200559328805433, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4055 + }, + { + "epoch": 1.620455453455853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4056 + }, + { + "epoch": 1.6208549740311626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4057 + }, + { + "epoch": 1.6212544946064722, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4058 + }, + { + "epoch": 1.621654015181782, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4059 + }, + { + "epoch": 1.6220535357570913, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4060 + }, + { + "epoch": 1.6224530563324011, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4061 + }, + { + "epoch": 1.622852576907711, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4062 + }, + { + "epoch": 1.6232520974830202, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4063 + }, + { + "epoch": 1.62365161805833, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4064 + }, + { + "epoch": 1.6240511386336396, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4065 + }, + { + "epoch": 1.6244506592089492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4066 + }, + { + "epoch": 1.624850179784259, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4067 + }, + { + "epoch": 1.6252497003595685, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4068 + }, + { + "epoch": 1.625649220934878, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4069 + }, + { + "epoch": 1.6260487415101879, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4070 + }, + { + "epoch": 1.6264482620854974, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4071 + }, + { + "epoch": 1.626847782660807, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4072 + }, + { + "epoch": 1.6272473032361168, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4073 + }, + { + "epoch": 1.6276468238114263, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4074 + }, + { + "epoch": 1.628046344386736, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 4075 + }, + { + "epoch": 1.6284458649620457, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4076 + }, + { + "epoch": 1.628845385537355, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4077 + }, + { + "epoch": 1.6292449061126648, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4078 + }, + { + "epoch": 1.6296444266879744, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 4079 + }, + { + "epoch": 1.630043947263284, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4080 + }, + { + "epoch": 1.6304434678385937, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4081 + }, + { + "epoch": 1.6308429884139033, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4082 + }, + { + "epoch": 1.6312425089892129, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4083 + }, + { + "epoch": 1.6316420295645226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4084 + }, + { + "epoch": 1.6320415501398322, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4085 + }, + { + "epoch": 1.6324410707151418, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4086 + }, + { + "epoch": 1.6328405912904516, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4087 + }, + { + "epoch": 1.6332401118657611, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4088 + }, + { + "epoch": 1.6336396324410707, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4089 + }, + { + "epoch": 1.6340391530163805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4090 + }, + { + "epoch": 1.6344386735916898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4091 + }, + { + "epoch": 1.6348381941669996, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4092 + }, + { + "epoch": 1.6352377147423094, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4093 + }, + { + "epoch": 1.6356372353176187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4094 + }, + { + "epoch": 1.6360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4095 + }, + { + "epoch": 1.636436276468238, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4096 + }, + { + "epoch": 1.6368357970435476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 4097 + }, + { + "epoch": 1.6372353176188574, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4098 + }, + { + "epoch": 1.637634838194167, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4099 + }, + { + "epoch": 1.6380343587694766, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 4100 + }, + { + "epoch": 1.6384338793447863, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4101 + }, + { + "epoch": 1.638833399920096, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4102 + }, + { + "epoch": 1.6392329204954055, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4103 + }, + { + "epoch": 1.6396324410707153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4104 + }, + { + "epoch": 1.6400319616460246, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4105 + }, + { + "epoch": 1.6404314822213344, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4106 + }, + { + "epoch": 1.6408310027966442, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4107 + }, + { + "epoch": 1.6412305233719535, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4108 + }, + { + "epoch": 1.6416300439472633, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4109 + }, + { + "epoch": 1.6420295645225729, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4110 + }, + { + "epoch": 1.6424290850978824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 4111 + }, + { + "epoch": 1.6428286056731922, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4112 + }, + { + "epoch": 1.6432281262485018, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4113 + }, + { + "epoch": 1.6436276468238114, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4114 + }, + { + "epoch": 1.6440271673991211, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4115 + }, + { + "epoch": 1.6444266879744307, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4116 + }, + { + "epoch": 1.6448262085497403, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4117 + }, + { + "epoch": 1.64522572912505, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4118 + }, + { + "epoch": 1.6456252497003596, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4119 + }, + { + "epoch": 1.6460247702756692, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4120 + }, + { + "epoch": 1.646424290850979, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4121 + }, + { + "epoch": 1.6468238114262883, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4122 + }, + { + "epoch": 1.647223332001598, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4123 + }, + { + "epoch": 1.6476228525769077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4124 + }, + { + "epoch": 1.6480223731522172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4125 + }, + { + "epoch": 1.648421893727527, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4126 + }, + { + "epoch": 1.6488214143028366, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4127 + }, + { + "epoch": 1.6492209348781461, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4128 + }, + { + "epoch": 1.649620455453456, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4129 + }, + { + "epoch": 1.6500199760287655, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4130 + }, + { + "epoch": 1.650419496604075, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 4131 + }, + { + "epoch": 1.6508190171793848, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4132 + }, + { + "epoch": 1.6512185377546944, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4133 + }, + { + "epoch": 1.651618058330004, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4134 + }, + { + "epoch": 1.6520175789053138, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4135 + }, + { + "epoch": 1.652417099480623, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 4136 + }, + { + "epoch": 1.6528166200559329, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4137 + }, + { + "epoch": 1.6532161406312427, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4138 + }, + { + "epoch": 1.653615661206552, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4139 + }, + { + "epoch": 1.6540151817818618, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4140 + }, + { + "epoch": 1.6544147023571714, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4141 + }, + { + "epoch": 1.654814222932481, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4142 + }, + { + "epoch": 1.6552137435077907, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4143 + }, + { + "epoch": 1.6556132640831003, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4144 + }, + { + "epoch": 1.6560127846584098, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4145 + }, + { + "epoch": 1.6564123052337196, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4146 + }, + { + "epoch": 1.6568118258090292, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4147 + }, + { + "epoch": 1.6572113463843388, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4148 + }, + { + "epoch": 1.6576108669596485, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4149 + }, + { + "epoch": 1.6580103875349579, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4150 + }, + { + "epoch": 1.6584099081102677, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4151 + }, + { + "epoch": 1.6588094286855775, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4152 + }, + { + "epoch": 1.6592089492608868, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4153 + }, + { + "epoch": 1.6596084698361966, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4154 + }, + { + "epoch": 1.6600079904115062, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4155 + }, + { + "epoch": 1.6604075109868157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4156 + }, + { + "epoch": 1.6608070315621255, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4157 + }, + { + "epoch": 1.661206552137435, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4158 + }, + { + "epoch": 1.6616060727127446, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4159 + }, + { + "epoch": 1.6620055932880544, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4160 + }, + { + "epoch": 1.662405113863364, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4161 + }, + { + "epoch": 1.6628046344386735, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4162 + }, + { + "epoch": 1.6632041550139833, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4163 + }, + { + "epoch": 1.663603675589293, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4164 + }, + { + "epoch": 1.6640031961646025, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4165 + }, + { + "epoch": 1.6644027167399122, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4166 + }, + { + "epoch": 1.6648022373152216, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4167 + }, + { + "epoch": 1.6652017578905314, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4168 + }, + { + "epoch": 1.665601278465841, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 4169 + }, + { + "epoch": 1.6660007990411505, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4170 + }, + { + "epoch": 1.6664003196164603, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4171 + }, + { + "epoch": 1.6667998401917699, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 4172 + }, + { + "epoch": 1.6671993607670794, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4173 + }, + { + "epoch": 1.6675988813423892, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4174 + }, + { + "epoch": 1.6679984019176988, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4175 + }, + { + "epoch": 1.6683979224930083, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 4176 + }, + { + "epoch": 1.6687974430683181, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4177 + }, + { + "epoch": 1.6691969636436277, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4178 + }, + { + "epoch": 1.6695964842189372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4179 + }, + { + "epoch": 1.669996004794247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4180 + }, + { + "epoch": 1.6703955253695564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4181 + }, + { + "epoch": 1.6707950459448662, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4182 + }, + { + "epoch": 1.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4183 + }, + { + "epoch": 1.6715940870954853, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4184 + }, + { + "epoch": 1.671993607670795, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4185 + }, + { + "epoch": 1.6723931282461046, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 4186 + }, + { + "epoch": 1.6727926488214142, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4187 + }, + { + "epoch": 1.673192169396724, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4188 + }, + { + "epoch": 1.6735916899720336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4189 + }, + { + "epoch": 1.6739912105473431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4190 + }, + { + "epoch": 1.674390731122653, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4191 + }, + { + "epoch": 1.6747902516979625, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4192 + }, + { + "epoch": 1.675189772273272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4193 + }, + { + "epoch": 1.6755892928485818, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4194 + }, + { + "epoch": 1.6759888134238912, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4195 + }, + { + "epoch": 1.676388333999201, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4196 + }, + { + "epoch": 1.6767878545745107, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4197 + }, + { + "epoch": 1.67718737514982, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4198 + }, + { + "epoch": 1.6775868957251299, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4199 + }, + { + "epoch": 1.6779864163004394, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4200 + }, + { + "epoch": 1.678385936875749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 4201 + }, + { + "epoch": 1.6787854574510588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4202 + }, + { + "epoch": 1.6791849780263683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4203 + }, + { + "epoch": 1.679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 4204 + }, + { + "epoch": 1.6799840191769877, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4205 + }, + { + "epoch": 1.6803835397522973, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4206 + }, + { + "epoch": 1.6807830603276068, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4207 + }, + { + "epoch": 1.6811825809029166, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4208 + }, + { + "epoch": 1.6815821014782262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4209 + }, + { + "epoch": 1.6819816220535357, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4210 + }, + { + "epoch": 1.6823811426288455, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4211 + }, + { + "epoch": 1.6827806632041549, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4212 + }, + { + "epoch": 1.6831801837794647, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 4213 + }, + { + "epoch": 1.6835797043547742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4214 + }, + { + "epoch": 1.6839792249300838, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 4215 + }, + { + "epoch": 1.6843787455053936, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4216 + }, + { + "epoch": 1.6847782660807031, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 4217 + }, + { + "epoch": 1.6851777866560127, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4218 + }, + { + "epoch": 1.6855773072313225, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4219 + }, + { + "epoch": 1.685976827806632, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 4220 + }, + { + "epoch": 1.6863763483819416, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 4221 + }, + { + "epoch": 1.6867758689572514, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4222 + }, + { + "epoch": 1.687175389532561, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4223 + }, + { + "epoch": 1.6875749101078705, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4224 + }, + { + "epoch": 1.6879744306831803, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4225 + }, + { + "epoch": 1.6883739512584897, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4226 + }, + { + "epoch": 1.6887734718337994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4227 + }, + { + "epoch": 1.6891729924091092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4228 + }, + { + "epoch": 1.6895725129844186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4229 + }, + { + "epoch": 1.6899720335597284, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4230 + }, + { + "epoch": 1.690371554135038, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4231 + }, + { + "epoch": 1.6907710747103475, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4232 + }, + { + "epoch": 1.6911705952856573, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4233 + }, + { + "epoch": 1.6915701158609668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4234 + }, + { + "epoch": 1.6919696364362764, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4235 + }, + { + "epoch": 1.6923691570115862, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4236 + }, + { + "epoch": 1.6927686775868958, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4237 + }, + { + "epoch": 1.6931681981622053, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4238 + }, + { + "epoch": 1.693567718737515, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4239 + }, + { + "epoch": 1.6939672393128244, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4240 + }, + { + "epoch": 1.6943667598881342, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4241 + }, + { + "epoch": 1.694766280463444, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4242 + }, + { + "epoch": 1.6951658010387534, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4243 + }, + { + "epoch": 1.6955653216140631, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4244 + }, + { + "epoch": 1.6959648421893727, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4245 + }, + { + "epoch": 1.6963643627646823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4246 + }, + { + "epoch": 1.696763883339992, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4247 + }, + { + "epoch": 1.6971634039153016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4248 + }, + { + "epoch": 1.6975629244906112, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4249 + }, + { + "epoch": 1.697962445065921, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4250 + }, + { + "epoch": 1.6983619656412305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4251 + }, + { + "epoch": 1.69876148621654, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4252 + }, + { + "epoch": 1.69916100679185, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4253 + }, + { + "epoch": 1.6995605273671595, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 4254 + }, + { + "epoch": 1.699960047942469, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4255 + }, + { + "epoch": 1.7003595685177788, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4256 + }, + { + "epoch": 1.7007590890930881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4257 + }, + { + "epoch": 1.701158609668398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4258 + }, + { + "epoch": 1.7015581302437075, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4259 + }, + { + "epoch": 1.701957650819017, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4260 + }, + { + "epoch": 1.7023571713943269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4261 + }, + { + "epoch": 1.7027566919696364, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4262 + }, + { + "epoch": 1.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4263 + }, + { + "epoch": 1.7035557331202558, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4264 + }, + { + "epoch": 1.7039552536955653, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4265 + }, + { + "epoch": 1.704354774270875, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4266 + }, + { + "epoch": 1.7047542948461847, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4267 + }, + { + "epoch": 1.7051538154214942, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4268 + }, + { + "epoch": 1.7055533359968038, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4269 + }, + { + "epoch": 1.7059528565721136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4270 + }, + { + "epoch": 1.706352377147423, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4271 + }, + { + "epoch": 1.7067518977227327, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4272 + }, + { + "epoch": 1.7071514182980425, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4273 + }, + { + "epoch": 1.7075509388733519, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4274 + }, + { + "epoch": 1.7079504594486616, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4275 + }, + { + "epoch": 1.7083499800239712, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4276 + }, + { + "epoch": 1.7087495005992808, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4277 + }, + { + "epoch": 1.7091490211745906, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4278 + }, + { + "epoch": 1.7095485417499001, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4279 + }, + { + "epoch": 1.7099480623252097, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4280 + }, + { + "epoch": 1.7103475829005195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4281 + }, + { + "epoch": 1.710747103475829, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4282 + }, + { + "epoch": 1.7111466240511386, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4283 + }, + { + "epoch": 1.7115461446264484, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4284 + }, + { + "epoch": 1.7119456652017577, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4285 + }, + { + "epoch": 1.7123451857770675, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4286 + }, + { + "epoch": 1.7127447063523773, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4287 + }, + { + "epoch": 1.7131442269276866, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4288 + }, + { + "epoch": 1.7135437475029964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4289 + }, + { + "epoch": 1.713943268078306, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4290 + }, + { + "epoch": 1.7143427886536156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4291 + }, + { + "epoch": 1.7147423092289253, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4292 + }, + { + "epoch": 1.715141829804235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4293 + }, + { + "epoch": 1.7155413503795445, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4294 + }, + { + "epoch": 1.7159408709548543, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4295 + }, + { + "epoch": 1.7163403915301638, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4296 + }, + { + "epoch": 1.7167399121054734, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4297 + }, + { + "epoch": 1.7171394326807832, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4298 + }, + { + "epoch": 1.7175389532560927, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4299 + }, + { + "epoch": 1.7179384738314023, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 4300 + }, + { + "epoch": 1.718337994406712, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4301 + }, + { + "epoch": 1.7187375149820214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4302 + }, + { + "epoch": 1.7191370355573312, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4303 + }, + { + "epoch": 1.7195365561326408, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4304 + }, + { + "epoch": 1.7199360767079503, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4305 + }, + { + "epoch": 1.7203355972832601, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4306 + }, + { + "epoch": 1.7207351178585697, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4307 + }, + { + "epoch": 1.7211346384338793, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 4308 + }, + { + "epoch": 1.721534159009189, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4309 + }, + { + "epoch": 1.7219336795844986, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4310 + }, + { + "epoch": 1.7223332001598082, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4311 + }, + { + "epoch": 1.722732720735118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4312 + }, + { + "epoch": 1.7231322413104275, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4313 + }, + { + "epoch": 1.723531761885737, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4314 + }, + { + "epoch": 1.7239312824610469, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4315 + }, + { + "epoch": 1.7243308030363562, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4316 + }, + { + "epoch": 1.724730323611666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4317 + }, + { + "epoch": 1.7251298441869758, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4318 + }, + { + "epoch": 1.7255293647622851, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4319 + }, + { + "epoch": 1.725928885337595, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4320 + }, + { + "epoch": 1.7263284059129045, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4321 + }, + { + "epoch": 1.726727926488214, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4322 + }, + { + "epoch": 1.7271274470635238, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4323 + }, + { + "epoch": 1.7275269676388334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4324 + }, + { + "epoch": 1.727926488214143, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4325 + }, + { + "epoch": 1.7283260087894528, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4326 + }, + { + "epoch": 1.7287255293647623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4327 + }, + { + "epoch": 1.7291250499400719, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4328 + }, + { + "epoch": 1.7295245705153817, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4329 + }, + { + "epoch": 1.729924091090691, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4330 + }, + { + "epoch": 1.7303236116660008, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4331 + }, + { + "epoch": 1.7307231322413106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4332 + }, + { + "epoch": 1.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4333 + }, + { + "epoch": 1.7315221733919297, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4334 + }, + { + "epoch": 1.7319216939672393, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4335 + }, + { + "epoch": 1.7323212145425488, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4336 + }, + { + "epoch": 1.7327207351178586, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4337 + }, + { + "epoch": 1.7331202556931682, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4338 + }, + { + "epoch": 1.7335197762684778, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4339 + }, + { + "epoch": 1.7339192968437875, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4340 + }, + { + "epoch": 1.734318817419097, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4341 + }, + { + "epoch": 1.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4342 + }, + { + "epoch": 1.7351178585697165, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4343 + }, + { + "epoch": 1.735517379145026, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4344 + }, + { + "epoch": 1.7359168997203356, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4345 + }, + { + "epoch": 1.7363164202956454, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4346 + }, + { + "epoch": 1.7367159408709547, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4347 + }, + { + "epoch": 1.7371154614462645, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4348 + }, + { + "epoch": 1.737514982021574, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4349 + }, + { + "epoch": 1.7379145025968836, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4350 + }, + { + "epoch": 1.7383140231721934, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4351 + }, + { + "epoch": 1.738713543747503, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4352 + }, + { + "epoch": 1.7391130643228125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4353 + }, + { + "epoch": 1.7395125848981223, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4354 + }, + { + "epoch": 1.739912105473432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4355 + }, + { + "epoch": 1.7403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4356 + }, + { + "epoch": 1.7407111466240512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 4357 + }, + { + "epoch": 1.7411106671993608, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4358 + }, + { + "epoch": 1.7415101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4359 + }, + { + "epoch": 1.7419097083499802, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4360 + }, + { + "epoch": 1.7423092289252895, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4361 + }, + { + "epoch": 1.7427087495005993, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4362 + }, + { + "epoch": 1.743108270075909, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4363 + }, + { + "epoch": 1.7435077906512184, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4364 + }, + { + "epoch": 1.7439073112265282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4365 + }, + { + "epoch": 1.7443068318018378, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4366 + }, + { + "epoch": 1.7447063523771473, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4367 + }, + { + "epoch": 1.7451058729524571, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4368 + }, + { + "epoch": 1.7455053935277667, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4369 + }, + { + "epoch": 1.7459049141030762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4370 + }, + { + "epoch": 1.746304434678386, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4371 + }, + { + "epoch": 1.7467039552536956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4372 + }, + { + "epoch": 1.7471034758290052, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4373 + }, + { + "epoch": 1.747502996404315, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4374 + }, + { + "epoch": 1.7479025169796243, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4375 + }, + { + "epoch": 1.748302037554934, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4376 + }, + { + "epoch": 1.7487015581302439, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4377 + }, + { + "epoch": 1.7491010787055532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4378 + }, + { + "epoch": 1.749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4379 + }, + { + "epoch": 1.7499001198561726, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4380 + }, + { + "epoch": 1.7502996404314821, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4381 + }, + { + "epoch": 1.750699161006792, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4382 + }, + { + "epoch": 1.7510986815821015, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4383 + }, + { + "epoch": 1.751498202157411, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4384 + }, + { + "epoch": 1.7518977227327208, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4385 + }, + { + "epoch": 1.7522972433080304, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4386 + }, + { + "epoch": 1.75269676388334, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4387 + }, + { + "epoch": 1.7530962844586497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4388 + }, + { + "epoch": 1.7534958050339593, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4389 + }, + { + "epoch": 1.7538953256092689, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4390 + }, + { + "epoch": 1.7542948461845786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4391 + }, + { + "epoch": 1.754694366759888, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4392 + }, + { + "epoch": 1.7550938873351978, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 4393 + }, + { + "epoch": 1.7554934079105073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4394 + }, + { + "epoch": 1.755892928485817, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4395 + }, + { + "epoch": 1.7562924490611267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4396 + }, + { + "epoch": 1.7566919696364363, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4397 + }, + { + "epoch": 1.7570914902117458, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4398 + }, + { + "epoch": 1.7574910107870556, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4399 + }, + { + "epoch": 1.7578905313623652, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4400 + }, + { + "epoch": 1.7582900519376747, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4401 + }, + { + "epoch": 1.7586895725129845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4402 + }, + { + "epoch": 1.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4403 + }, + { + "epoch": 1.7594886136636037, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4404 + }, + { + "epoch": 1.7598881342389134, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4405 + }, + { + "epoch": 1.7602876548142228, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 4406 + }, + { + "epoch": 1.7606871753895326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 4407 + }, + { + "epoch": 1.7610866959648424, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4408 + }, + { + "epoch": 1.7614862165401517, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 4409 + }, + { + "epoch": 1.7618857371154615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4410 + }, + { + "epoch": 1.762285257690771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4411 + }, + { + "epoch": 1.7626847782660806, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4412 + }, + { + "epoch": 1.7630842988413904, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4413 + }, + { + "epoch": 1.7634838194167, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4414 + }, + { + "epoch": 1.7638833399920095, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4415 + }, + { + "epoch": 1.7642828605673193, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4416 + }, + { + "epoch": 1.7646823811426289, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4417 + }, + { + "epoch": 1.7650819017179384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4418 + }, + { + "epoch": 1.7654814222932482, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4419 + }, + { + "epoch": 1.7658809428685576, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4420 + }, + { + "epoch": 1.7662804634438674, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4421 + }, + { + "epoch": 1.7666799840191771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4422 + }, + { + "epoch": 1.7670795045944865, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4423 + }, + { + "epoch": 1.7674790251697963, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4424 + }, + { + "epoch": 1.7678785457451058, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4425 + }, + { + "epoch": 1.7682780663204154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4426 + }, + { + "epoch": 1.7686775868957252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4427 + }, + { + "epoch": 1.7690771074710347, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4428 + }, + { + "epoch": 1.7694766280463443, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4429 + }, + { + "epoch": 1.769876148621654, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4430 + }, + { + "epoch": 1.7702756691969637, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4431 + }, + { + "epoch": 1.7706751897722732, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4432 + }, + { + "epoch": 1.771074710347583, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4433 + }, + { + "epoch": 1.7714742309228926, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4434 + }, + { + "epoch": 1.7718737514982021, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4435 + }, + { + "epoch": 1.772273272073512, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4436 + }, + { + "epoch": 1.7726727926488213, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4437 + }, + { + "epoch": 1.773072313224131, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4438 + }, + { + "epoch": 1.7734718337994406, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4439 + }, + { + "epoch": 1.7738713543747502, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4440 + }, + { + "epoch": 1.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4441 + }, + { + "epoch": 1.7746703955253695, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4442 + }, + { + "epoch": 1.775069916100679, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 4443 + }, + { + "epoch": 1.7754694366759889, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 4444 + }, + { + "epoch": 1.7758689572512985, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4445 + }, + { + "epoch": 1.776268477826608, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 4446 + }, + { + "epoch": 1.7766679984019178, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 4447 + }, + { + "epoch": 1.7770675189772274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4448 + }, + { + "epoch": 1.777467039552537, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 4449 + }, + { + "epoch": 1.7778665601278467, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 4450 + }, + { + "epoch": 1.778266080703156, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4451 + }, + { + "epoch": 1.7786656012784658, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4452 + }, + { + "epoch": 1.7790651218537756, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4453 + }, + { + "epoch": 1.779464642429085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4454 + }, + { + "epoch": 1.7798641630043948, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4455 + }, + { + "epoch": 1.7802636835797043, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4456 + }, + { + "epoch": 1.780663204155014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4457 + }, + { + "epoch": 1.7810627247303237, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4458 + }, + { + "epoch": 1.7814622453056332, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4459 + }, + { + "epoch": 1.7818617658809428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4460 + }, + { + "epoch": 1.7822612864562526, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4461 + }, + { + "epoch": 1.7826608070315622, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4462 + }, + { + "epoch": 1.7830603276068717, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4463 + }, + { + "epoch": 1.7834598481821815, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4464 + }, + { + "epoch": 1.7838593687574908, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4465 + }, + { + "epoch": 1.7842588893328006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4466 + }, + { + "epoch": 1.7846584099081104, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4467 + }, + { + "epoch": 1.7850579304834198, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4468 + }, + { + "epoch": 1.7854574510587295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4469 + }, + { + "epoch": 1.7858569716340391, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4470 + }, + { + "epoch": 1.7862564922093487, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4471 + }, + { + "epoch": 1.7866560127846585, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4472 + }, + { + "epoch": 1.787055533359968, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4473 + }, + { + "epoch": 1.7874550539352776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4474 + }, + { + "epoch": 1.7878545745105874, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4475 + }, + { + "epoch": 1.788254095085897, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4476 + }, + { + "epoch": 1.7886536156612065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4477 + }, + { + "epoch": 1.7890531362365163, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4478 + }, + { + "epoch": 1.7894526568118259, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4479 + }, + { + "epoch": 1.7898521773871354, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4480 + }, + { + "epoch": 1.7902516979624452, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4481 + }, + { + "epoch": 1.7906512185377546, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 4482 + }, + { + "epoch": 1.7910507391130643, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4483 + }, + { + "epoch": 1.791450259688374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4484 + }, + { + "epoch": 1.7918497802636835, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4485 + }, + { + "epoch": 1.7922493008389933, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4486 + }, + { + "epoch": 1.7926488214143028, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4487 + }, + { + "epoch": 1.7930483419896124, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4488 + }, + { + "epoch": 1.7934478625649222, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4489 + }, + { + "epoch": 1.7938473831402317, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4490 + }, + { + "epoch": 1.7942469037155413, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4491 + }, + { + "epoch": 1.794646424290851, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4492 + }, + { + "epoch": 1.7950459448661606, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4493 + }, + { + "epoch": 1.7954454654414702, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4494 + }, + { + "epoch": 1.79584498601678, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4495 + }, + { + "epoch": 1.7962445065920893, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4496 + }, + { + "epoch": 1.7966440271673991, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4497 + }, + { + "epoch": 1.797043547742709, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4498 + }, + { + "epoch": 1.7974430683180183, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4499 + }, + { + "epoch": 1.797842588893328, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4500 + }, + { + "epoch": 1.7982421094686376, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4501 + }, + { + "epoch": 1.7986416300439472, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4502 + }, + { + "epoch": 1.799041150619257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4503 + }, + { + "epoch": 1.7994406711945665, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4504 + }, + { + "epoch": 1.799840191769876, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4505 + }, + { + "epoch": 1.8002397123451859, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4506 + }, + { + "epoch": 1.8006392329204954, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4507 + }, + { + "epoch": 1.801038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4508 + }, + { + "epoch": 1.8014382740711148, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4509 + }, + { + "epoch": 1.8018377946464241, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4510 + }, + { + "epoch": 1.802237315221734, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4511 + }, + { + "epoch": 1.8026368357970437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4512 + }, + { + "epoch": 1.803036356372353, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4513 + }, + { + "epoch": 1.8034358769476628, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4514 + }, + { + "epoch": 1.8038353975229724, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4515 + }, + { + "epoch": 1.804234918098282, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 4516 + }, + { + "epoch": 1.8046344386735917, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4517 + }, + { + "epoch": 1.8050339592489013, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4518 + }, + { + "epoch": 1.8054334798242109, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4519 + }, + { + "epoch": 1.8058330003995207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4520 + }, + { + "epoch": 1.8062325209748302, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4521 + }, + { + "epoch": 1.8066320415501398, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 4522 + }, + { + "epoch": 1.8070315621254496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4523 + }, + { + "epoch": 1.8074310827007591, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4524 + }, + { + "epoch": 1.8078306032760687, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4525 + }, + { + "epoch": 1.8082301238513785, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4526 + }, + { + "epoch": 1.8086296444266878, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4527 + }, + { + "epoch": 1.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4528 + }, + { + "epoch": 1.8094286855773072, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4529 + }, + { + "epoch": 1.8098282061526167, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4530 + }, + { + "epoch": 1.8102277267279265, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4531 + }, + { + "epoch": 1.810627247303236, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4532 + }, + { + "epoch": 1.8110267678785457, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4533 + }, + { + "epoch": 1.8114262884538554, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4534 + }, + { + "epoch": 1.811825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4535 + }, + { + "epoch": 1.8122253296044746, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4536 + }, + { + "epoch": 1.8126248501797844, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4537 + }, + { + "epoch": 1.813024370755094, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 4538 + }, + { + "epoch": 1.8134238913304035, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4539 + }, + { + "epoch": 1.8138234119057133, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 4540 + }, + { + "epoch": 1.8142229324810226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 4541 + }, + { + "epoch": 1.8146224530563324, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4542 + }, + { + "epoch": 1.8150219736316422, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4543 + }, + { + "epoch": 1.8154214942069515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4544 + }, + { + "epoch": 1.8158210147822613, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4545 + }, + { + "epoch": 1.8162205353575709, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4546 + }, + { + "epoch": 1.8166200559328805, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4547 + }, + { + "epoch": 1.8170195765081902, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4548 + }, + { + "epoch": 1.8174190970834998, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4549 + }, + { + "epoch": 1.8178186176588094, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4550 + }, + { + "epoch": 1.8182181382341192, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4551 + }, + { + "epoch": 1.8186176588094287, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4552 + }, + { + "epoch": 1.8190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4553 + }, + { + "epoch": 1.819416699960048, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4554 + }, + { + "epoch": 1.8198162205353574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4555 + }, + { + "epoch": 1.8202157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4556 + }, + { + "epoch": 1.820615261685977, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4557 + }, + { + "epoch": 1.8210147822612863, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4558 + }, + { + "epoch": 1.821414302836596, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4559 + }, + { + "epoch": 1.8218138234119057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4560 + }, + { + "epoch": 1.8222133439872152, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4561 + }, + { + "epoch": 1.822612864562525, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4562 + }, + { + "epoch": 1.8230123851378346, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4563 + }, + { + "epoch": 1.8234119057131442, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4564 + }, + { + "epoch": 1.823811426288454, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4565 + }, + { + "epoch": 1.8242109468637635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4566 + }, + { + "epoch": 1.824610467439073, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4567 + }, + { + "epoch": 1.8250099880143829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4568 + }, + { + "epoch": 1.8254095085896924, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4569 + }, + { + "epoch": 1.825809029165002, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4570 + }, + { + "epoch": 1.8262085497403118, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4571 + }, + { + "epoch": 1.8266080703156211, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4572 + }, + { + "epoch": 1.827007590890931, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4573 + }, + { + "epoch": 1.8274071114662405, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4574 + }, + { + "epoch": 1.82780663204155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4575 + }, + { + "epoch": 1.8282061526168598, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4576 + }, + { + "epoch": 1.8286056731921694, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4577 + }, + { + "epoch": 1.829005193767479, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4578 + }, + { + "epoch": 1.8294047143427887, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4579 + }, + { + "epoch": 1.8298042349180983, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4580 + }, + { + "epoch": 1.8302037554934079, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4581 + }, + { + "epoch": 1.8306032760687176, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4582 + }, + { + "epoch": 1.8310027966440272, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4583 + }, + { + "epoch": 1.8314023172193368, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4584 + }, + { + "epoch": 1.8318018377946466, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4585 + }, + { + "epoch": 1.832201358369956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4586 + }, + { + "epoch": 1.8326008789452657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4587 + }, + { + "epoch": 1.8330003995205755, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4588 + }, + { + "epoch": 1.8333999200958848, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4589 + }, + { + "epoch": 1.8337994406711946, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4590 + }, + { + "epoch": 1.8341989612465042, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4591 + }, + { + "epoch": 1.8345984818218137, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4592 + }, + { + "epoch": 1.8349980023971235, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4593 + }, + { + "epoch": 1.835397522972433, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4594 + }, + { + "epoch": 1.8357970435477426, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4595 + }, + { + "epoch": 1.8361965641230524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4596 + }, + { + "epoch": 1.836596084698362, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4597 + }, + { + "epoch": 1.8369956052736716, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4598 + }, + { + "epoch": 1.8373951258489813, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4599 + }, + { + "epoch": 1.8377946464242907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4600 + }, + { + "epoch": 1.8381941669996005, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4601 + }, + { + "epoch": 1.8385936875749103, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4602 + }, + { + "epoch": 1.8389932081502196, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4603 + }, + { + "epoch": 1.8393927287255294, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4604 + }, + { + "epoch": 1.839792249300839, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4605 + }, + { + "epoch": 1.8401917698761485, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4606 + }, + { + "epoch": 1.8405912904514583, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4607 + }, + { + "epoch": 1.8409908110267679, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 4608 + }, + { + "epoch": 1.8413903316020774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4609 + }, + { + "epoch": 1.8417898521773872, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4610 + }, + { + "epoch": 1.8421893727526968, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4611 + }, + { + "epoch": 1.8425888933280063, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4612 + }, + { + "epoch": 1.8429884139033161, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4613 + }, + { + "epoch": 1.8433879344786257, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4614 + }, + { + "epoch": 1.8437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4615 + }, + { + "epoch": 1.844186975629245, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4616 + }, + { + "epoch": 1.8445864962045544, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4617 + }, + { + "epoch": 1.8449860167798642, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4618 + }, + { + "epoch": 1.8453855373551737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4619 + }, + { + "epoch": 1.8457850579304833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4620 + }, + { + "epoch": 1.846184578505793, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4621 + }, + { + "epoch": 1.8465840990811027, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4622 + }, + { + "epoch": 1.8469836196564122, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4623 + }, + { + "epoch": 1.847383140231722, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4624 + }, + { + "epoch": 1.8477826608070316, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4625 + }, + { + "epoch": 1.8481821813823411, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4626 + }, + { + "epoch": 1.848581701957651, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4627 + }, + { + "epoch": 1.8489812225329605, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4628 + }, + { + "epoch": 1.84938074310827, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4629 + }, + { + "epoch": 1.8497802636835798, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4630 + }, + { + "epoch": 1.8501797842588892, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4631 + }, + { + "epoch": 1.850579304834199, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4632 + }, + { + "epoch": 1.8509788254095088, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 4633 + }, + { + "epoch": 1.851378345984818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4634 + }, + { + "epoch": 1.8517778665601279, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4635 + }, + { + "epoch": 1.8521773871354374, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4636 + }, + { + "epoch": 1.852576907710747, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4637 + }, + { + "epoch": 1.8529764282860568, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4638 + }, + { + "epoch": 1.8533759488613664, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4639 + }, + { + "epoch": 1.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4640 + }, + { + "epoch": 1.8541749900119857, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4641 + }, + { + "epoch": 1.8545745105872953, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 4642 + }, + { + "epoch": 1.8549740311626048, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4643 + }, + { + "epoch": 1.8553735517379146, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4644 + }, + { + "epoch": 1.855773072313224, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4645 + }, + { + "epoch": 1.8561725928885338, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4646 + }, + { + "epoch": 1.8565721134638435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4647 + }, + { + "epoch": 1.8569716340391529, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4648 + }, + { + "epoch": 1.8573711546144627, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4649 + }, + { + "epoch": 1.8577706751897722, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4650 + }, + { + "epoch": 1.8581701957650818, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4651 + }, + { + "epoch": 1.8585697163403916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4652 + }, + { + "epoch": 1.8589692369157012, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4653 + }, + { + "epoch": 1.8593687574910107, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4654 + }, + { + "epoch": 1.8597682780663205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4655 + }, + { + "epoch": 1.86016779864163, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4656 + }, + { + "epoch": 1.8605673192169396, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4657 + }, + { + "epoch": 1.8609668397922494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4658 + }, + { + "epoch": 1.861366360367559, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4659 + }, + { + "epoch": 1.8617658809428685, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4660 + }, + { + "epoch": 1.8621654015181783, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4661 + }, + { + "epoch": 1.8625649220934877, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4662 + }, + { + "epoch": 1.8629644426687975, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4663 + }, + { + "epoch": 1.863363963244107, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4664 + }, + { + "epoch": 1.8637634838194166, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4665 + }, + { + "epoch": 1.8641630043947264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4666 + }, + { + "epoch": 1.864562524970036, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4667 + }, + { + "epoch": 1.8649620455453455, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4668 + }, + { + "epoch": 1.8653615661206553, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4669 + }, + { + "epoch": 1.8657610866959649, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4670 + }, + { + "epoch": 1.8661606072712744, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4671 + }, + { + "epoch": 1.8665601278465842, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4672 + }, + { + "epoch": 1.8669596484218938, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4673 + }, + { + "epoch": 1.8673591689972033, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4674 + }, + { + "epoch": 1.8677586895725131, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4675 + }, + { + "epoch": 1.8681582101478225, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4676 + }, + { + "epoch": 1.8685577307231322, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4677 + }, + { + "epoch": 1.868957251298442, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4678 + }, + { + "epoch": 1.8693567718737514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4679 + }, + { + "epoch": 1.8697562924490612, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4680 + }, + { + "epoch": 1.8701558130243707, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4681 + }, + { + "epoch": 1.8705553335996803, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4682 + }, + { + "epoch": 1.87095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4683 + }, + { + "epoch": 1.8713543747502996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4684 + }, + { + "epoch": 1.8717538953256092, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4685 + }, + { + "epoch": 1.872153415900919, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4686 + }, + { + "epoch": 1.8725529364762286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4687 + }, + { + "epoch": 1.8729524570515381, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4688 + }, + { + "epoch": 1.873351977626848, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 4689 + }, + { + "epoch": 1.8737514982021573, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4690 + }, + { + "epoch": 1.874151018777467, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4691 + }, + { + "epoch": 1.8745505393527768, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4692 + }, + { + "epoch": 1.8749500599280862, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4693 + }, + { + "epoch": 1.875349580503396, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4694 + }, + { + "epoch": 1.8757491010787055, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4695 + }, + { + "epoch": 1.876148621654015, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4696 + }, + { + "epoch": 1.8765481422293249, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4697 + }, + { + "epoch": 1.8769476628046344, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4698 + }, + { + "epoch": 1.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4699 + }, + { + "epoch": 1.8777467039552538, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4700 + }, + { + "epoch": 1.8781462245305633, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4701 + }, + { + "epoch": 1.878545745105873, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4702 + }, + { + "epoch": 1.8789452656811827, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4703 + }, + { + "epoch": 1.8793447862564923, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4704 + }, + { + "epoch": 1.8797443068318018, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4705 + }, + { + "epoch": 1.8801438274071116, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4706 + }, + { + "epoch": 1.880543347982421, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4707 + }, + { + "epoch": 1.8809428685577307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4708 + }, + { + "epoch": 1.8813423891330403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4709 + }, + { + "epoch": 1.8817419097083499, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4710 + }, + { + "epoch": 1.8821414302836597, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4711 + }, + { + "epoch": 1.8825409508589692, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4712 + }, + { + "epoch": 1.8829404714342788, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4713 + }, + { + "epoch": 1.8833399920095886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4714 + }, + { + "epoch": 1.8837395125848981, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4715 + }, + { + "epoch": 1.8841390331602077, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4716 + }, + { + "epoch": 1.8845385537355175, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4717 + }, + { + "epoch": 1.884938074310827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4718 + }, + { + "epoch": 1.8853375948861366, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4719 + }, + { + "epoch": 1.8857371154614464, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4720 + }, + { + "epoch": 1.8861366360367557, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4721 + }, + { + "epoch": 1.8865361566120655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4722 + }, + { + "epoch": 1.8869356771873753, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4723 + }, + { + "epoch": 1.8873351977626847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4724 + }, + { + "epoch": 1.8877347183379944, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4725 + }, + { + "epoch": 1.888134238913304, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4726 + }, + { + "epoch": 1.8885337594886136, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4727 + }, + { + "epoch": 1.8889332800639234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4728 + }, + { + "epoch": 1.889332800639233, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4729 + }, + { + "epoch": 1.8897323212145425, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4730 + }, + { + "epoch": 1.8901318417898523, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4731 + }, + { + "epoch": 1.8905313623651618, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4732 + }, + { + "epoch": 1.8909308829404714, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 4733 + }, + { + "epoch": 1.8913304035157812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4734 + }, + { + "epoch": 1.8917299240910905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4735 + }, + { + "epoch": 1.8921294446664003, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 4736 + }, + { + "epoch": 1.89252896524171, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4737 + }, + { + "epoch": 1.8929284858170194, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4738 + }, + { + "epoch": 1.8933280063923292, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4739 + }, + { + "epoch": 1.8937275269676388, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4740 + }, + { + "epoch": 1.8941270475429484, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4741 + }, + { + "epoch": 1.8945265681182581, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4742 + }, + { + "epoch": 1.8949260886935677, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4743 + }, + { + "epoch": 1.8953256092688773, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4744 + }, + { + "epoch": 1.895725129844187, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4745 + }, + { + "epoch": 1.8961246504194966, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4746 + }, + { + "epoch": 1.8965241709948062, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4747 + }, + { + "epoch": 1.896923691570116, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4748 + }, + { + "epoch": 1.8973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4749 + }, + { + "epoch": 1.897722732720735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4750 + }, + { + "epoch": 1.898122253296045, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4751 + }, + { + "epoch": 1.8985217738713542, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 4752 + }, + { + "epoch": 1.898921294446664, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4753 + }, + { + "epoch": 1.8993208150219736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4754 + }, + { + "epoch": 1.8997203355972831, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4755 + }, + { + "epoch": 1.900119856172593, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4756 + }, + { + "epoch": 1.9005193767479025, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4757 + }, + { + "epoch": 1.900918897323212, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4758 + }, + { + "epoch": 1.9013184178985219, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4759 + }, + { + "epoch": 1.9017179384738314, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4760 + }, + { + "epoch": 1.902117459049141, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4761 + }, + { + "epoch": 1.9025169796244508, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4762 + }, + { + "epoch": 1.9029165001997603, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4763 + }, + { + "epoch": 1.90331602077507, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4764 + }, + { + "epoch": 1.9037155413503797, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4765 + }, + { + "epoch": 1.904115061925689, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4766 + }, + { + "epoch": 1.9045145825009988, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4767 + }, + { + "epoch": 1.9049141030763086, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4768 + }, + { + "epoch": 1.905313623651618, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4769 + }, + { + "epoch": 1.9057131442269277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4770 + }, + { + "epoch": 1.9061126648022373, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4771 + }, + { + "epoch": 1.9065121853775469, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4772 + }, + { + "epoch": 1.9069117059528566, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4773 + }, + { + "epoch": 1.9073112265281662, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4774 + }, + { + "epoch": 1.9077107471034758, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4775 + }, + { + "epoch": 1.9081102676787856, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4776 + }, + { + "epoch": 1.9085097882540951, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4777 + }, + { + "epoch": 1.9089093088294047, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4778 + }, + { + "epoch": 1.9093088294047145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4779 + }, + { + "epoch": 1.9097083499800238, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4780 + }, + { + "epoch": 1.9101078705553336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4781 + }, + { + "epoch": 1.9105073911306434, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4782 + }, + { + "epoch": 1.9109069117059527, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4783 + }, + { + "epoch": 1.9113064322812625, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4784 + }, + { + "epoch": 1.911705952856572, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4785 + }, + { + "epoch": 1.9121054734318816, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4786 + }, + { + "epoch": 1.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4787 + }, + { + "epoch": 1.912904514582501, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4788 + }, + { + "epoch": 1.9133040351578106, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4789 + }, + { + "epoch": 1.9137035557331203, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 4790 + }, + { + "epoch": 1.91410307630843, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4791 + }, + { + "epoch": 1.9145025968837395, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4792 + }, + { + "epoch": 1.9149021174590493, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4793 + }, + { + "epoch": 1.9153016380343588, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4794 + }, + { + "epoch": 1.9157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4795 + }, + { + "epoch": 1.9161006791849782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4796 + }, + { + "epoch": 1.9165001997602875, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4797 + }, + { + "epoch": 1.9168997203355973, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4798 + }, + { + "epoch": 1.9172992409109069, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4799 + }, + { + "epoch": 1.9176987614862164, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 4800 + }, + { + "epoch": 1.9180982820615262, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4801 + }, + { + "epoch": 1.9184978026368358, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4802 + }, + { + "epoch": 1.9188973232121453, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 4803 + }, + { + "epoch": 1.9192968437874551, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 4804 + }, + { + "epoch": 1.9196963643627647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4805 + }, + { + "epoch": 1.9200958849380743, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4806 + }, + { + "epoch": 1.920495405513384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4807 + }, + { + "epoch": 1.9208949260886936, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4808 + }, + { + "epoch": 1.9212944466640032, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4809 + }, + { + "epoch": 1.921693967239313, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4810 + }, + { + "epoch": 1.9220934878146223, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4811 + }, + { + "epoch": 1.922493008389932, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4812 + }, + { + "epoch": 1.9228925289652419, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4813 + }, + { + "epoch": 1.9232920495405512, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4814 + }, + { + "epoch": 1.923691570115861, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 4815 + }, + { + "epoch": 1.9240910906911706, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4816 + }, + { + "epoch": 1.9244906112664801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4817 + }, + { + "epoch": 1.92489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4818 + }, + { + "epoch": 1.9252896524170995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4819 + }, + { + "epoch": 1.925689172992409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4820 + }, + { + "epoch": 1.9260886935677188, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4821 + }, + { + "epoch": 1.9264882141430284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4822 + }, + { + "epoch": 1.926887734718338, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4823 + }, + { + "epoch": 1.9272872552936477, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4824 + }, + { + "epoch": 1.927686775868957, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4825 + }, + { + "epoch": 1.9280862964442669, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4826 + }, + { + "epoch": 1.9284858170195767, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4827 + }, + { + "epoch": 1.928885337594886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4828 + }, + { + "epoch": 1.9292848581701958, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4829 + }, + { + "epoch": 1.9296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4830 + }, + { + "epoch": 1.930083899320815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4831 + }, + { + "epoch": 1.9304834198961247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4832 + }, + { + "epoch": 1.9308829404714343, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4833 + }, + { + "epoch": 1.9312824610467438, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4834 + }, + { + "epoch": 1.9316819816220536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4835 + }, + { + "epoch": 1.9320815021973632, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4836 + }, + { + "epoch": 1.9324810227726728, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4837 + }, + { + "epoch": 1.9328805433479825, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4838 + }, + { + "epoch": 1.933280063923292, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4839 + }, + { + "epoch": 1.9336795844986017, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4840 + }, + { + "epoch": 1.9340791050739115, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4841 + }, + { + "epoch": 1.9344786256492208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4842 + }, + { + "epoch": 1.9348781462245306, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4843 + }, + { + "epoch": 1.9352776667998401, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4844 + }, + { + "epoch": 1.9356771873751497, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4845 + }, + { + "epoch": 1.9360767079504595, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4846 + }, + { + "epoch": 1.936476228525769, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4847 + }, + { + "epoch": 1.9368757491010786, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4848 + }, + { + "epoch": 1.9372752696763884, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4849 + }, + { + "epoch": 1.937674790251698, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4850 + }, + { + "epoch": 1.9380743108270075, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 4851 + }, + { + "epoch": 1.9384738314023173, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4852 + }, + { + "epoch": 1.938873351977627, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4853 + }, + { + "epoch": 1.9392728725529365, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4854 + }, + { + "epoch": 1.9396723931282462, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4855 + }, + { + "epoch": 1.9400719137035556, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4856 + }, + { + "epoch": 1.9404714342788654, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4857 + }, + { + "epoch": 1.940870954854175, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4858 + }, + { + "epoch": 1.9412704754294845, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4859 + }, + { + "epoch": 1.9416699960047943, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4860 + }, + { + "epoch": 1.9420695165801038, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4861 + }, + { + "epoch": 1.9424690371554134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4862 + }, + { + "epoch": 1.9428685577307232, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4863 + }, + { + "epoch": 1.9432680783060328, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4864 + }, + { + "epoch": 1.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4865 + }, + { + "epoch": 1.9440671194566521, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4866 + }, + { + "epoch": 1.9444666400319617, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4867 + }, + { + "epoch": 1.9448661606072712, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4868 + }, + { + "epoch": 1.945265681182581, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4869 + }, + { + "epoch": 1.9456652017578904, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4870 + }, + { + "epoch": 1.9460647223332002, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4871 + }, + { + "epoch": 1.94646424290851, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4872 + }, + { + "epoch": 1.9468637634838193, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4873 + }, + { + "epoch": 1.947263284059129, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4874 + }, + { + "epoch": 1.9476628046344386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4875 + }, + { + "epoch": 1.9480623252097482, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4876 + }, + { + "epoch": 1.948461845785058, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4877 + }, + { + "epoch": 1.9488613663603676, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4878 + }, + { + "epoch": 1.9492608869356771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4879 + }, + { + "epoch": 1.949660407510987, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4880 + }, + { + "epoch": 1.9500599280862965, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4881 + }, + { + "epoch": 1.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4882 + }, + { + "epoch": 1.9508589692369158, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4883 + }, + { + "epoch": 1.9512584898122254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4884 + }, + { + "epoch": 1.951658010387535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4885 + }, + { + "epoch": 1.9520575309628447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4886 + }, + { + "epoch": 1.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4887 + }, + { + "epoch": 1.9528565721134639, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4888 + }, + { + "epoch": 1.9532560926887734, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4889 + }, + { + "epoch": 1.953655613264083, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 4890 + }, + { + "epoch": 1.9540551338393928, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4891 + }, + { + "epoch": 1.9544546544147023, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4892 + }, + { + "epoch": 1.954854174990012, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4893 + }, + { + "epoch": 1.9552536955653217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4894 + }, + { + "epoch": 1.9556532161406313, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4895 + }, + { + "epoch": 1.9560527367159408, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4896 + }, + { + "epoch": 1.9564522572912506, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4897 + }, + { + "epoch": 1.9568517778665602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4898 + }, + { + "epoch": 1.9572512984418697, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4899 + }, + { + "epoch": 1.9576508190171795, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4900 + }, + { + "epoch": 1.9580503395924889, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4901 + }, + { + "epoch": 1.9584498601677987, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4902 + }, + { + "epoch": 1.9588493807431082, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4903 + }, + { + "epoch": 1.9592489013184178, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4904 + }, + { + "epoch": 1.9596484218937276, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4905 + }, + { + "epoch": 1.9600479424690371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4906 + }, + { + "epoch": 1.9604474630443467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4907 + }, + { + "epoch": 1.9608469836196565, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4908 + }, + { + "epoch": 1.961246504194966, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4909 + }, + { + "epoch": 1.9616460247702756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4910 + }, + { + "epoch": 1.9620455453455854, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4911 + }, + { + "epoch": 1.962445065920895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4912 + }, + { + "epoch": 1.9628445864962045, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 4913 + }, + { + "epoch": 1.9632441070715143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4914 + }, + { + "epoch": 1.9636436276468237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 4915 + }, + { + "epoch": 1.9640431482221334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4916 + }, + { + "epoch": 1.9644426687974432, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 4917 + }, + { + "epoch": 1.9648421893727526, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4918 + }, + { + "epoch": 1.9652417099480624, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4919 + }, + { + "epoch": 1.965641230523372, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4920 + }, + { + "epoch": 1.9660407510986815, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4921 + }, + { + "epoch": 1.9664402716739913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4922 + }, + { + "epoch": 1.9668397922493008, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4923 + }, + { + "epoch": 1.9672393128246104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4924 + }, + { + "epoch": 1.9676388333999202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 4925 + }, + { + "epoch": 1.9680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4926 + }, + { + "epoch": 1.9684378745505393, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4927 + }, + { + "epoch": 1.968837395125849, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4928 + }, + { + "epoch": 1.9692369157011587, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4929 + }, + { + "epoch": 1.9696364362764682, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4930 + }, + { + "epoch": 1.970035956851778, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4931 + }, + { + "epoch": 1.9704354774270874, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4932 + }, + { + "epoch": 1.9708349980023971, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4933 + }, + { + "epoch": 1.9712345185777067, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 4934 + }, + { + "epoch": 1.9716340391530163, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4935 + }, + { + "epoch": 1.972033559728326, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4936 + }, + { + "epoch": 1.9724330803036356, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4937 + }, + { + "epoch": 1.9728326008789452, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4938 + }, + { + "epoch": 1.973232121454255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4939 + }, + { + "epoch": 1.9736316420295645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4940 + }, + { + "epoch": 1.974031162604874, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4941 + }, + { + "epoch": 1.9744306831801839, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4942 + }, + { + "epoch": 1.9748302037554935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4943 + }, + { + "epoch": 1.975229724330803, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4944 + }, + { + "epoch": 1.9756292449061128, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4945 + }, + { + "epoch": 1.9760287654814221, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4946 + }, + { + "epoch": 1.976428286056732, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4947 + }, + { + "epoch": 1.9768278066320415, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4948 + }, + { + "epoch": 1.977227327207351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4949 + }, + { + "epoch": 1.9776268477826608, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4950 + }, + { + "epoch": 1.9780263683579704, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4951 + }, + { + "epoch": 1.97842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 4952 + }, + { + "epoch": 1.9788254095085898, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4953 + }, + { + "epoch": 1.9792249300838993, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4954 + }, + { + "epoch": 1.979624450659209, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4955 + }, + { + "epoch": 1.9800239712345187, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4956 + }, + { + "epoch": 1.9804234918098282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4957 + }, + { + "epoch": 1.9808230123851378, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4958 + }, + { + "epoch": 1.9812225329604476, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4959 + }, + { + "epoch": 1.981622053535757, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4960 + }, + { + "epoch": 1.9820215741110667, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4961 + }, + { + "epoch": 1.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4962 + }, + { + "epoch": 1.9828206152616858, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 4963 + }, + { + "epoch": 1.9832201358369956, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4964 + }, + { + "epoch": 1.9836196564123052, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4965 + }, + { + "epoch": 1.9840191769876148, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4966 + }, + { + "epoch": 1.9844186975629245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4967 + }, + { + "epoch": 1.9848182181382341, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4968 + }, + { + "epoch": 1.9852177387135437, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4969 + }, + { + "epoch": 1.9856172592888535, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4970 + }, + { + "epoch": 1.986016779864163, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4971 + }, + { + "epoch": 1.9864163004394726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4972 + }, + { + "epoch": 1.9868158210147824, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4973 + }, + { + "epoch": 1.987215341590092, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4974 + }, + { + "epoch": 1.9876148621654015, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4975 + }, + { + "epoch": 1.9880143827407113, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4976 + }, + { + "epoch": 1.9884139033160206, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4977 + }, + { + "epoch": 1.9888134238913304, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4978 + }, + { + "epoch": 1.98921294446664, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4979 + }, + { + "epoch": 1.9896124650419496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4980 + }, + { + "epoch": 1.9900119856172593, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4981 + }, + { + "epoch": 1.990411506192569, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4982 + }, + { + "epoch": 1.9908110267678785, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4983 + }, + { + "epoch": 1.9912105473431883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4984 + }, + { + "epoch": 1.9916100679184978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4985 + }, + { + "epoch": 1.9920095884938074, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4986 + }, + { + "epoch": 1.9924091090691172, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4987 + }, + { + "epoch": 1.9928086296444267, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4988 + }, + { + "epoch": 1.9932081502197363, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4989 + }, + { + "epoch": 1.993607670795046, + "grad_norm": 0.04541015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4990 + }, + { + "epoch": 1.9940071913703554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4991 + }, + { + "epoch": 1.9944067119456652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4992 + }, + { + "epoch": 1.9948062325209748, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 4993 + }, + { + "epoch": 1.9952057530962843, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4994 + }, + { + "epoch": 1.9956052736715941, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4995 + }, + { + "epoch": 1.9960047942469037, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4996 + }, + { + "epoch": 1.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4997 + }, + { + "epoch": 1.996803835397523, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4998 + }, + { + "epoch": 1.9972033559728326, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4999 + }, + { + "epoch": 1.9976028765481422, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5000 + }, + { + "epoch": 1.998002397123452, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5001 + }, + { + "epoch": 1.9984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5002 + }, + { + "epoch": 1.998801438274071, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5003 + }, + { + "epoch": 1.9992009588493809, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5004 + }, + { + "epoch": 1.9996004794246902, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5005 + }, + { + "epoch": 2.0, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5006 + }, + { + "epoch": 2.00039952057531, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5007 + }, + { + "epoch": 2.000799041150619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5008 + }, + { + "epoch": 2.001198561725929, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5009 + }, + { + "epoch": 2.0015980823012387, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5010 + }, + { + "epoch": 2.001997602876548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5011 + }, + { + "epoch": 2.002397123451858, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5012 + }, + { + "epoch": 2.0027966440271676, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5013 + }, + { + "epoch": 2.003196164602477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 5014 + }, + { + "epoch": 2.0035956851777867, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5015 + }, + { + "epoch": 2.003995205753096, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5016 + }, + { + "epoch": 2.004394726328406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5017 + }, + { + "epoch": 2.0047942469037157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5018 + }, + { + "epoch": 2.005193767479025, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5019 + }, + { + "epoch": 2.005593288054335, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5020 + }, + { + "epoch": 2.0059928086296446, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5021 + }, + { + "epoch": 2.006392329204954, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5022 + }, + { + "epoch": 2.0067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5023 + }, + { + "epoch": 2.0071913703555735, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 5024 + }, + { + "epoch": 2.007590890930883, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5025 + }, + { + "epoch": 2.0079904115061926, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5026 + }, + { + "epoch": 2.0083899320815024, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5027 + }, + { + "epoch": 2.0087894526568117, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 5028 + }, + { + "epoch": 2.0091889732321215, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 5029 + }, + { + "epoch": 2.009588493807431, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5030 + }, + { + "epoch": 2.0099880143827407, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5031 + }, + { + "epoch": 2.0103875349580504, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5032 + }, + { + "epoch": 2.01078705553336, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5033 + }, + { + "epoch": 2.0111865761086696, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5034 + }, + { + "epoch": 2.0115860966839794, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5035 + }, + { + "epoch": 2.0119856172592887, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5036 + }, + { + "epoch": 2.0123851378345985, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5037 + }, + { + "epoch": 2.0127846584099083, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5038 + }, + { + "epoch": 2.0131841789852176, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5039 + }, + { + "epoch": 2.0135836995605274, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5040 + }, + { + "epoch": 2.013983220135837, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 5041 + }, + { + "epoch": 2.0143827407111465, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 5042 + }, + { + "epoch": 2.0147822612864563, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 5043 + }, + { + "epoch": 2.0151817818617657, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5044 + }, + { + "epoch": 2.0155813024370755, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5045 + }, + { + "epoch": 2.0159808230123852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5046 + }, + { + "epoch": 2.0163803435876946, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5047 + }, + { + "epoch": 2.0167798641630044, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5048 + }, + { + "epoch": 2.017179384738314, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5049 + }, + { + "epoch": 2.0175789053136235, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5050 + }, + { + "epoch": 2.0179784258889333, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5051 + }, + { + "epoch": 2.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5052 + }, + { + "epoch": 2.0187774670395524, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5053 + }, + { + "epoch": 2.019176987614862, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5054 + }, + { + "epoch": 2.019576508190172, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5055 + }, + { + "epoch": 2.0199760287654813, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5056 + }, + { + "epoch": 2.020375549340791, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5057 + }, + { + "epoch": 2.020775069916101, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5058 + }, + { + "epoch": 2.0211745904914102, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5059 + }, + { + "epoch": 2.02157411106672, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5060 + }, + { + "epoch": 2.0219736316420294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5061 + }, + { + "epoch": 2.022373152217339, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5062 + }, + { + "epoch": 2.022772672792649, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5063 + }, + { + "epoch": 2.0231721933679583, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5064 + }, + { + "epoch": 2.023571713943268, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5065 + }, + { + "epoch": 2.023971234518578, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5066 + }, + { + "epoch": 2.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5067 + }, + { + "epoch": 2.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5068 + }, + { + "epoch": 2.0251697962445068, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5069 + }, + { + "epoch": 2.025569316819816, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5070 + }, + { + "epoch": 2.025968837395126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5071 + }, + { + "epoch": 2.0263683579704357, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5072 + }, + { + "epoch": 2.026767878545745, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5073 + }, + { + "epoch": 2.027167399121055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5074 + }, + { + "epoch": 2.027566919696364, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5075 + }, + { + "epoch": 2.027966440271674, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5076 + }, + { + "epoch": 2.0283659608469837, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5077 + }, + { + "epoch": 2.028765481422293, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5078 + }, + { + "epoch": 2.029165001997603, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5079 + }, + { + "epoch": 2.0295645225729126, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5080 + }, + { + "epoch": 2.029964043148222, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5081 + }, + { + "epoch": 2.0303635637235318, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5082 + }, + { + "epoch": 2.0307630842988416, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5083 + }, + { + "epoch": 2.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5084 + }, + { + "epoch": 2.0315621254494607, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 5085 + }, + { + "epoch": 2.0319616460247705, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5086 + }, + { + "epoch": 2.03236116660008, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5087 + }, + { + "epoch": 2.0327606871753896, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5088 + }, + { + "epoch": 2.033160207750699, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5089 + }, + { + "epoch": 2.0335597283260087, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5090 + }, + { + "epoch": 2.0339592489013185, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5091 + }, + { + "epoch": 2.034358769476628, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5092 + }, + { + "epoch": 2.0347582900519376, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5093 + }, + { + "epoch": 2.0351578106272474, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5094 + }, + { + "epoch": 2.0355573312025568, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5095 + }, + { + "epoch": 2.0359568517778666, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5096 + }, + { + "epoch": 2.0363563723531763, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5097 + }, + { + "epoch": 2.0367558929284857, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5098 + }, + { + "epoch": 2.0371554135037955, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5099 + }, + { + "epoch": 2.0375549340791053, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 5100 + }, + { + "epoch": 2.0379544546544146, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5101 + }, + { + "epoch": 2.0383539752297244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5102 + }, + { + "epoch": 2.038753495805034, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5103 + }, + { + "epoch": 2.0391530163803435, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5104 + }, + { + "epoch": 2.0395525369556533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5105 + }, + { + "epoch": 2.0399520575309626, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5106 + }, + { + "epoch": 2.0403515781062724, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5107 + }, + { + "epoch": 2.040751098681582, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 5108 + }, + { + "epoch": 2.0411506192568916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 5109 + }, + { + "epoch": 2.0415501398322013, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5110 + }, + { + "epoch": 2.041949660407511, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5111 + }, + { + "epoch": 2.0423491809828205, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5112 + }, + { + "epoch": 2.0427487015581303, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5113 + }, + { + "epoch": 2.04314822213344, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5114 + }, + { + "epoch": 2.0435477427087494, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5115 + }, + { + "epoch": 2.043947263284059, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 5116 + }, + { + "epoch": 2.044346783859369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5117 + }, + { + "epoch": 2.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5118 + }, + { + "epoch": 2.045145825009988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5119 + }, + { + "epoch": 2.0455453455852974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 5120 + }, + { + "epoch": 2.0459448661606072, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5121 + }, + { + "epoch": 2.046344386735917, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5122 + }, + { + "epoch": 2.0467439073112264, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5123 + }, + { + "epoch": 2.047143427886536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5124 + }, + { + "epoch": 2.047542948461846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5125 + }, + { + "epoch": 2.0479424690371553, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5126 + }, + { + "epoch": 2.048341989612465, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5127 + }, + { + "epoch": 2.048741510187775, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5128 + }, + { + "epoch": 2.049141030763084, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5129 + }, + { + "epoch": 2.049540551338394, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5130 + }, + { + "epoch": 2.0499400719137038, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5131 + }, + { + "epoch": 2.050339592489013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5132 + }, + { + "epoch": 2.050739113064323, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5133 + }, + { + "epoch": 2.0511386336396322, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5134 + }, + { + "epoch": 2.051538154214942, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5135 + }, + { + "epoch": 2.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 5136 + }, + { + "epoch": 2.052337195365561, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5137 + }, + { + "epoch": 2.052736715940871, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 5138 + }, + { + "epoch": 2.0531362365161807, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5139 + }, + { + "epoch": 2.05353575709149, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5140 + }, + { + "epoch": 2.0539352776668, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5141 + }, + { + "epoch": 2.0543347982421096, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5142 + }, + { + "epoch": 2.054734318817419, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5143 + }, + { + "epoch": 2.0551338393927288, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5144 + }, + { + "epoch": 2.0555333599680385, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 5145 + }, + { + "epoch": 2.055932880543348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5146 + }, + { + "epoch": 2.0563324011186577, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5147 + }, + { + "epoch": 2.0567319216939675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5148 + }, + { + "epoch": 2.057131442269277, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5149 + }, + { + "epoch": 2.0575309628445866, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5150 + }, + { + "epoch": 2.057930483419896, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5151 + }, + { + "epoch": 2.0583300039952057, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5152 + }, + { + "epoch": 2.0587295245705155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5153 + }, + { + "epoch": 2.059129045145825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5154 + }, + { + "epoch": 2.0595285657211346, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5155 + }, + { + "epoch": 2.0599280862964444, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5156 + }, + { + "epoch": 2.0603276068717538, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5157 + }, + { + "epoch": 2.0607271274470635, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5158 + }, + { + "epoch": 2.0611266480223733, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5159 + }, + { + "epoch": 2.0615261685976827, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5160 + }, + { + "epoch": 2.0619256891729925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5161 + }, + { + "epoch": 2.0623252097483022, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5162 + }, + { + "epoch": 2.0627247303236116, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5163 + }, + { + "epoch": 2.0631242508989214, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5164 + }, + { + "epoch": 2.0635237714742307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5165 + }, + { + "epoch": 2.0639232920495405, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5166 + }, + { + "epoch": 2.0643228126248503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5167 + }, + { + "epoch": 2.0647223332001596, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5168 + }, + { + "epoch": 2.0651218537754694, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5169 + }, + { + "epoch": 2.065521374350779, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5170 + }, + { + "epoch": 2.0659208949260885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5171 + }, + { + "epoch": 2.0663204155013983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5172 + }, + { + "epoch": 2.066719936076708, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5173 + }, + { + "epoch": 2.0671194566520175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 5174 + }, + { + "epoch": 2.0675189772273272, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5175 + }, + { + "epoch": 2.067918497802637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5176 + }, + { + "epoch": 2.0683180183779464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 5177 + }, + { + "epoch": 2.068717538953256, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5178 + }, + { + "epoch": 2.0691170595285655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5179 + }, + { + "epoch": 2.0695165801038753, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5180 + }, + { + "epoch": 2.069916100679185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5181 + }, + { + "epoch": 2.0703156212544944, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5182 + }, + { + "epoch": 2.070715141829804, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5183 + }, + { + "epoch": 2.071114662405114, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5184 + }, + { + "epoch": 2.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5185 + }, + { + "epoch": 2.071913703555733, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5186 + }, + { + "epoch": 2.072313224131043, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5187 + }, + { + "epoch": 2.0727127447063523, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5188 + }, + { + "epoch": 2.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5189 + }, + { + "epoch": 2.073511785856972, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 5190 + }, + { + "epoch": 2.073911306432281, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5191 + }, + { + "epoch": 2.074310827007591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5192 + }, + { + "epoch": 2.0747103475829007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5193 + }, + { + "epoch": 2.07510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5194 + }, + { + "epoch": 2.07550938873352, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5195 + }, + { + "epoch": 2.075908909308829, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5196 + }, + { + "epoch": 2.076308429884139, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5197 + }, + { + "epoch": 2.076707950459449, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5198 + }, + { + "epoch": 2.077107471034758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5199 + }, + { + "epoch": 2.077506991610068, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5200 + }, + { + "epoch": 2.0779065121853777, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5201 + }, + { + "epoch": 2.078306032760687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5202 + }, + { + "epoch": 2.078705553335997, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5203 + }, + { + "epoch": 2.0791050739113066, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5204 + }, + { + "epoch": 2.079504594486616, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 5205 + }, + { + "epoch": 2.0799041150619257, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5206 + }, + { + "epoch": 2.0803036356372355, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5207 + }, + { + "epoch": 2.080703156212545, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5208 + }, + { + "epoch": 2.0811026767878547, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5209 + }, + { + "epoch": 2.081502197363164, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5210 + }, + { + "epoch": 2.081901717938474, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5211 + }, + { + "epoch": 2.0823012385137836, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5212 + }, + { + "epoch": 2.082700759089093, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5213 + }, + { + "epoch": 2.0831002796644027, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 5214 + }, + { + "epoch": 2.0834998002397125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5215 + }, + { + "epoch": 2.083899320815022, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5216 + }, + { + "epoch": 2.0842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5217 + }, + { + "epoch": 2.0846983619656414, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5218 + }, + { + "epoch": 2.0850978825409507, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5219 + }, + { + "epoch": 2.0854974031162605, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5220 + }, + { + "epoch": 2.0858969236915703, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5221 + }, + { + "epoch": 2.0862964442668797, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 5222 + }, + { + "epoch": 2.0866959648421894, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5223 + }, + { + "epoch": 2.087095485417499, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5224 + }, + { + "epoch": 2.0874950059928086, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5225 + }, + { + "epoch": 2.0878945265681184, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5226 + }, + { + "epoch": 2.0882940471434277, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5227 + }, + { + "epoch": 2.0886935677187375, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5228 + }, + { + "epoch": 2.0890930882940473, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5229 + }, + { + "epoch": 2.0894926088693566, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5230 + }, + { + "epoch": 2.0898921294446664, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5231 + }, + { + "epoch": 2.090291650019976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5232 + }, + { + "epoch": 2.0906911705952855, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5233 + }, + { + "epoch": 2.0910906911705953, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5234 + }, + { + "epoch": 2.091490211745905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5235 + }, + { + "epoch": 2.0918897323212144, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5236 + }, + { + "epoch": 2.0922892528965242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5237 + }, + { + "epoch": 2.092688773471834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5238 + }, + { + "epoch": 2.0930882940471434, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5239 + }, + { + "epoch": 2.093487814622453, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5240 + }, + { + "epoch": 2.0938873351977625, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5241 + }, + { + "epoch": 2.0942868557730723, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5242 + }, + { + "epoch": 2.094686376348382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5243 + }, + { + "epoch": 2.0950858969236914, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5244 + }, + { + "epoch": 2.095485417499001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5245 + }, + { + "epoch": 2.095884938074311, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5246 + }, + { + "epoch": 2.0962844586496203, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5247 + }, + { + "epoch": 2.09668397922493, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5248 + }, + { + "epoch": 2.09708349980024, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5249 + }, + { + "epoch": 2.0974830203755492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5250 + }, + { + "epoch": 2.097882540950859, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5251 + }, + { + "epoch": 2.098282061526169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 5252 + }, + { + "epoch": 2.098681582101478, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5253 + }, + { + "epoch": 2.099081102676788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5254 + }, + { + "epoch": 2.0994806232520973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5255 + }, + { + "epoch": 2.099880143827407, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5256 + }, + { + "epoch": 2.100279664402717, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5257 + }, + { + "epoch": 2.100679184978026, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5258 + }, + { + "epoch": 2.101078705553336, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5259 + }, + { + "epoch": 2.1014782261286458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5260 + }, + { + "epoch": 2.101877746703955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5261 + }, + { + "epoch": 2.102277267279265, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5262 + }, + { + "epoch": 2.1026767878545747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5263 + }, + { + "epoch": 2.103076308429884, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5264 + }, + { + "epoch": 2.103475829005194, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5265 + }, + { + "epoch": 2.1038753495805036, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5266 + }, + { + "epoch": 2.104274870155813, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5267 + }, + { + "epoch": 2.1046743907311227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5268 + }, + { + "epoch": 2.105073911306432, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5269 + }, + { + "epoch": 2.105473431881742, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5270 + }, + { + "epoch": 2.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5271 + }, + { + "epoch": 2.106272473032361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5272 + }, + { + "epoch": 2.1066719936076708, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5273 + }, + { + "epoch": 2.1070715141829806, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5274 + }, + { + "epoch": 2.10747103475829, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5275 + }, + { + "epoch": 2.1078705553335997, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5276 + }, + { + "epoch": 2.1082700759089095, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5277 + }, + { + "epoch": 2.108669596484219, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5278 + }, + { + "epoch": 2.1090691170595286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5279 + }, + { + "epoch": 2.1094686376348384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5280 + }, + { + "epoch": 2.1098681582101477, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5281 + }, + { + "epoch": 2.1102676787854575, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5282 + }, + { + "epoch": 2.1106671993607673, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5283 + }, + { + "epoch": 2.1110667199360766, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5284 + }, + { + "epoch": 2.1114662405113864, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5285 + }, + { + "epoch": 2.1118657610866958, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5286 + }, + { + "epoch": 2.1122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5287 + }, + { + "epoch": 2.1126648022373153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5288 + }, + { + "epoch": 2.1130643228126247, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5289 + }, + { + "epoch": 2.1134638433879345, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5290 + }, + { + "epoch": 2.1138633639632443, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5291 + }, + { + "epoch": 2.1142628845385536, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5292 + }, + { + "epoch": 2.1146624051138634, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5293 + }, + { + "epoch": 2.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5294 + }, + { + "epoch": 2.1154614462644825, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5295 + }, + { + "epoch": 2.1158609668397923, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5296 + }, + { + "epoch": 2.116260487415102, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 5297 + }, + { + "epoch": 2.1166600079904114, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5298 + }, + { + "epoch": 2.117059528565721, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5299 + }, + { + "epoch": 2.1174590491410306, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5300 + }, + { + "epoch": 2.1178585697163403, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5301 + }, + { + "epoch": 2.11825809029165, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5302 + }, + { + "epoch": 2.1186576108669595, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5303 + }, + { + "epoch": 2.1190571314422693, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5304 + }, + { + "epoch": 2.119456652017579, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5305 + }, + { + "epoch": 2.1198561725928884, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5306 + }, + { + "epoch": 2.120255693168198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5307 + }, + { + "epoch": 2.120655213743508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5308 + }, + { + "epoch": 2.1210547343188173, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5309 + }, + { + "epoch": 2.121454254894127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5310 + }, + { + "epoch": 2.121853775469437, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5311 + }, + { + "epoch": 2.122253296044746, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5312 + }, + { + "epoch": 2.122652816620056, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5313 + }, + { + "epoch": 2.1230523371953653, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5314 + }, + { + "epoch": 2.123451857770675, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5315 + }, + { + "epoch": 2.123851378345985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5316 + }, + { + "epoch": 2.1242508989212943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5317 + }, + { + "epoch": 2.124650419496604, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5318 + }, + { + "epoch": 2.125049940071914, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5319 + }, + { + "epoch": 2.125449460647223, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5320 + }, + { + "epoch": 2.125848981222533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5321 + }, + { + "epoch": 2.1262485017978427, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5322 + }, + { + "epoch": 2.126648022373152, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5323 + }, + { + "epoch": 2.127047542948462, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5324 + }, + { + "epoch": 2.1274470635237717, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5325 + }, + { + "epoch": 2.127846584099081, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5326 + }, + { + "epoch": 2.128246104674391, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5327 + }, + { + "epoch": 2.1286456252497006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5328 + }, + { + "epoch": 2.12904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5329 + }, + { + "epoch": 2.1294446664003197, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5330 + }, + { + "epoch": 2.129844186975629, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5331 + }, + { + "epoch": 2.130243707550939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5332 + }, + { + "epoch": 2.1306432281262486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5333 + }, + { + "epoch": 2.131042748701558, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5334 + }, + { + "epoch": 2.1314422692768678, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5335 + }, + { + "epoch": 2.1318417898521775, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5336 + }, + { + "epoch": 2.132241310427487, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5337 + }, + { + "epoch": 2.1326408310027967, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5338 + }, + { + "epoch": 2.1330403515781065, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5339 + }, + { + "epoch": 2.133439872153416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5340 + }, + { + "epoch": 2.1338393927287256, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5341 + }, + { + "epoch": 2.134238913304035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5342 + }, + { + "epoch": 2.1346384338793447, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5343 + }, + { + "epoch": 2.1350379544546545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5344 + }, + { + "epoch": 2.135437475029964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5345 + }, + { + "epoch": 2.1358369956052736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5346 + }, + { + "epoch": 2.1362365161805834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5347 + }, + { + "epoch": 2.1366360367558928, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5348 + }, + { + "epoch": 2.1370355573312025, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5349 + }, + { + "epoch": 2.1374350779065123, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5350 + }, + { + "epoch": 2.1378345984818217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 5351 + }, + { + "epoch": 2.1382341190571315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5352 + }, + { + "epoch": 2.1386336396324412, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5353 + }, + { + "epoch": 2.1390331602077506, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5354 + }, + { + "epoch": 2.1394326807830604, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5355 + }, + { + "epoch": 2.13983220135837, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5356 + }, + { + "epoch": 2.1402317219336795, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 5357 + }, + { + "epoch": 2.1406312425089893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5358 + }, + { + "epoch": 2.1410307630842986, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5359 + }, + { + "epoch": 2.1414302836596084, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5360 + }, + { + "epoch": 2.141829804234918, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5361 + }, + { + "epoch": 2.1422293248102275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5362 + }, + { + "epoch": 2.1426288453855373, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5363 + }, + { + "epoch": 2.143028365960847, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5364 + }, + { + "epoch": 2.1434278865361565, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5365 + }, + { + "epoch": 2.1438274071114662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5366 + }, + { + "epoch": 2.144226927686776, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5367 + }, + { + "epoch": 2.1446264482620854, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5368 + }, + { + "epoch": 2.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5369 + }, + { + "epoch": 2.145425489412705, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5370 + }, + { + "epoch": 2.1458250099880143, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5371 + }, + { + "epoch": 2.146224530563324, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5372 + }, + { + "epoch": 2.146624051138634, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5373 + }, + { + "epoch": 2.147023571713943, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5374 + }, + { + "epoch": 2.147423092289253, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5375 + }, + { + "epoch": 2.1478226128645623, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5376 + }, + { + "epoch": 2.148222133439872, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5377 + }, + { + "epoch": 2.148621654015182, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5378 + }, + { + "epoch": 2.1490211745904912, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5379 + }, + { + "epoch": 2.149420695165801, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5380 + }, + { + "epoch": 2.149820215741111, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5381 + }, + { + "epoch": 2.15021973631642, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5382 + }, + { + "epoch": 2.15061925689173, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5383 + }, + { + "epoch": 2.1510187774670397, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5384 + }, + { + "epoch": 2.151418298042349, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5385 + }, + { + "epoch": 2.151817818617659, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5386 + }, + { + "epoch": 2.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5387 + }, + { + "epoch": 2.152616859768278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5388 + }, + { + "epoch": 2.1530163803435878, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5389 + }, + { + "epoch": 2.153415900918897, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5390 + }, + { + "epoch": 2.153815421494207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5391 + }, + { + "epoch": 2.1542149420695167, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5392 + }, + { + "epoch": 2.154614462644826, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5393 + }, + { + "epoch": 2.155013983220136, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5394 + }, + { + "epoch": 2.1554135037954456, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5395 + }, + { + "epoch": 2.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5396 + }, + { + "epoch": 2.1562125449460647, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5397 + }, + { + "epoch": 2.1566120655213745, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5398 + }, + { + "epoch": 2.157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5399 + }, + { + "epoch": 2.1574111066719937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5400 + }, + { + "epoch": 2.1578106272473034, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5401 + }, + { + "epoch": 2.158210147822613, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5402 + }, + { + "epoch": 2.1586096683979226, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5403 + }, + { + "epoch": 2.159009188973232, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5404 + }, + { + "epoch": 2.1594087095485417, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5405 + }, + { + "epoch": 2.1598082301238515, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5406 + }, + { + "epoch": 2.160207750699161, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5407 + }, + { + "epoch": 2.1606072712744706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5408 + }, + { + "epoch": 2.1610067918497804, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5409 + }, + { + "epoch": 2.1614063124250897, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5410 + }, + { + "epoch": 2.1618058330003995, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5411 + }, + { + "epoch": 2.1622053535757093, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5412 + }, + { + "epoch": 2.1626048741510187, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5413 + }, + { + "epoch": 2.1630043947263284, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5414 + }, + { + "epoch": 2.1634039153016382, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5415 + }, + { + "epoch": 2.1638034358769476, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5416 + }, + { + "epoch": 2.1642029564522574, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5417 + }, + { + "epoch": 2.164602477027567, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5418 + }, + { + "epoch": 2.1650019976028765, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5419 + }, + { + "epoch": 2.1654015181781863, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5420 + }, + { + "epoch": 2.1658010387534956, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5421 + }, + { + "epoch": 2.1662005593288054, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5422 + }, + { + "epoch": 2.166600079904115, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5423 + }, + { + "epoch": 2.1669996004794245, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5424 + }, + { + "epoch": 2.1673991210547343, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5425 + }, + { + "epoch": 2.167798641630044, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5426 + }, + { + "epoch": 2.1681981622053534, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5427 + }, + { + "epoch": 2.1685976827806632, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5428 + }, + { + "epoch": 2.168997203355973, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5429 + }, + { + "epoch": 2.1693967239312824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5430 + }, + { + "epoch": 2.169796244506592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5431 + }, + { + "epoch": 2.1701957650819015, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5432 + }, + { + "epoch": 2.1705952856572113, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5433 + }, + { + "epoch": 2.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5434 + }, + { + "epoch": 2.1713943268078304, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5435 + }, + { + "epoch": 2.17179384738314, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5436 + }, + { + "epoch": 2.17219336795845, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 5437 + }, + { + "epoch": 2.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5438 + }, + { + "epoch": 2.172992409109069, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5439 + }, + { + "epoch": 2.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5440 + }, + { + "epoch": 2.1737914502596882, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5441 + }, + { + "epoch": 2.174190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5442 + }, + { + "epoch": 2.174590491410308, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5443 + }, + { + "epoch": 2.174990011985617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5444 + }, + { + "epoch": 2.175389532560927, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5445 + }, + { + "epoch": 2.1757890531362367, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5446 + }, + { + "epoch": 2.176188573711546, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5447 + }, + { + "epoch": 2.176588094286856, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5448 + }, + { + "epoch": 2.176987614862165, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5449 + }, + { + "epoch": 2.177387135437475, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5450 + }, + { + "epoch": 2.1777866560127848, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5451 + }, + { + "epoch": 2.178186176588094, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 5452 + }, + { + "epoch": 2.178585697163404, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5453 + }, + { + "epoch": 2.1789852177387137, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5454 + }, + { + "epoch": 2.179384738314023, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5455 + }, + { + "epoch": 2.179784258889333, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5456 + }, + { + "epoch": 2.1801837794646426, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5457 + }, + { + "epoch": 2.180583300039952, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5458 + }, + { + "epoch": 2.1809828206152617, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5459 + }, + { + "epoch": 2.1813823411905715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5460 + }, + { + "epoch": 2.181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5461 + }, + { + "epoch": 2.1821813823411906, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5462 + }, + { + "epoch": 2.1825809029165004, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5463 + }, + { + "epoch": 2.1829804234918098, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5464 + }, + { + "epoch": 2.1833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5465 + }, + { + "epoch": 2.183779464642429, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 5466 + }, + { + "epoch": 2.1841789852177387, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 5467 + }, + { + "epoch": 2.1845785057930485, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5468 + }, + { + "epoch": 2.184978026368358, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5469 + }, + { + "epoch": 2.1853775469436676, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5470 + }, + { + "epoch": 2.1857770675189774, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5471 + }, + { + "epoch": 2.1861765880942867, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5472 + }, + { + "epoch": 2.1865761086695965, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 5473 + }, + { + "epoch": 2.1869756292449063, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5474 + }, + { + "epoch": 2.1873751498202156, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5475 + }, + { + "epoch": 2.1877746703955254, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5476 + }, + { + "epoch": 2.1881741909708348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5477 + }, + { + "epoch": 2.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5478 + }, + { + "epoch": 2.1889732321214543, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5479 + }, + { + "epoch": 2.1893727526967637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5480 + }, + { + "epoch": 2.1897722732720735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 5481 + }, + { + "epoch": 2.1901717938473833, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5482 + }, + { + "epoch": 2.1905713144226926, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5483 + }, + { + "epoch": 2.1909708349980024, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5484 + }, + { + "epoch": 2.191370355573312, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5485 + }, + { + "epoch": 2.1917698761486215, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5486 + }, + { + "epoch": 2.1921693967239313, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 5487 + }, + { + "epoch": 2.192568917299241, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5488 + }, + { + "epoch": 2.1929684378745504, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 5489 + }, + { + "epoch": 2.19336795844986, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5490 + }, + { + "epoch": 2.19376747902517, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5491 + }, + { + "epoch": 2.1941669996004793, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5492 + }, + { + "epoch": 2.194566520175789, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5493 + }, + { + "epoch": 2.1949660407510985, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5494 + }, + { + "epoch": 2.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5495 + }, + { + "epoch": 2.195765081901718, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5496 + }, + { + "epoch": 2.1961646024770274, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5497 + }, + { + "epoch": 2.196564123052337, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5498 + }, + { + "epoch": 2.196963643627647, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5499 + }, + { + "epoch": 2.1973631642029563, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5500 + }, + { + "epoch": 2.197762684778266, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5501 + }, + { + "epoch": 2.198162205353576, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5502 + }, + { + "epoch": 2.198561725928885, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5503 + }, + { + "epoch": 2.198961246504195, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5504 + }, + { + "epoch": 2.199360767079505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5505 + }, + { + "epoch": 2.199760287654814, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5506 + }, + { + "epoch": 2.200159808230124, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5507 + }, + { + "epoch": 2.2005593288054337, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5508 + }, + { + "epoch": 2.200958849380743, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5509 + }, + { + "epoch": 2.201358369956053, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5510 + }, + { + "epoch": 2.201757890531362, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5511 + }, + { + "epoch": 2.202157411106672, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5512 + }, + { + "epoch": 2.2025569316819817, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5513 + }, + { + "epoch": 2.202956452257291, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5514 + }, + { + "epoch": 2.203355972832601, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5515 + }, + { + "epoch": 2.2037554934079107, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5516 + }, + { + "epoch": 2.20415501398322, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5517 + }, + { + "epoch": 2.20455453455853, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5518 + }, + { + "epoch": 2.2049540551338396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5519 + }, + { + "epoch": 2.205353575709149, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5520 + }, + { + "epoch": 2.2057530962844587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5521 + }, + { + "epoch": 2.206152616859768, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5522 + }, + { + "epoch": 2.206552137435078, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5523 + }, + { + "epoch": 2.2069516580103876, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5524 + }, + { + "epoch": 2.207351178585697, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5525 + }, + { + "epoch": 2.2077506991610067, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5526 + }, + { + "epoch": 2.2081502197363165, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5527 + }, + { + "epoch": 2.208549740311626, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5528 + }, + { + "epoch": 2.2089492608869357, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5529 + }, + { + "epoch": 2.2093487814622454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5530 + }, + { + "epoch": 2.209748302037555, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5531 + }, + { + "epoch": 2.2101478226128646, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5532 + }, + { + "epoch": 2.2105473431881744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5533 + }, + { + "epoch": 2.2109468637634837, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5534 + }, + { + "epoch": 2.2113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5535 + }, + { + "epoch": 2.2117459049141033, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5536 + }, + { + "epoch": 2.2121454254894126, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5537 + }, + { + "epoch": 2.2125449460647224, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5538 + }, + { + "epoch": 2.2129444666400317, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5539 + }, + { + "epoch": 2.2133439872153415, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5540 + }, + { + "epoch": 2.2137435077906513, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5541 + }, + { + "epoch": 2.2141430283659607, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5542 + }, + { + "epoch": 2.2145425489412705, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5543 + }, + { + "epoch": 2.2149420695165802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5544 + }, + { + "epoch": 2.2153415900918896, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5545 + }, + { + "epoch": 2.2157411106671994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5546 + }, + { + "epoch": 2.216140631242509, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5547 + }, + { + "epoch": 2.2165401518178185, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5548 + }, + { + "epoch": 2.2169396723931283, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5549 + }, + { + "epoch": 2.217339192968438, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5550 + }, + { + "epoch": 2.2177387135437474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5551 + }, + { + "epoch": 2.218138234119057, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5552 + }, + { + "epoch": 2.218537754694367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5553 + }, + { + "epoch": 2.2189372752696763, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 5554 + }, + { + "epoch": 2.219336795844986, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5555 + }, + { + "epoch": 2.2197363164202955, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5556 + }, + { + "epoch": 2.2201358369956052, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5557 + }, + { + "epoch": 2.220535357570915, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5558 + }, + { + "epoch": 2.2209348781462244, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5559 + }, + { + "epoch": 2.221334398721534, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5560 + }, + { + "epoch": 2.221733919296844, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5561 + }, + { + "epoch": 2.2221334398721533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5562 + }, + { + "epoch": 2.222532960447463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5563 + }, + { + "epoch": 2.222932481022773, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5564 + }, + { + "epoch": 2.223332001598082, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5565 + }, + { + "epoch": 2.223731522173392, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5566 + }, + { + "epoch": 2.2241310427487013, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5567 + }, + { + "epoch": 2.224530563324011, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5568 + }, + { + "epoch": 2.224930083899321, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5569 + }, + { + "epoch": 2.2253296044746302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5570 + }, + { + "epoch": 2.22572912504994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5571 + }, + { + "epoch": 2.22612864562525, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5572 + }, + { + "epoch": 2.226528166200559, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5573 + }, + { + "epoch": 2.226927686775869, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5574 + }, + { + "epoch": 2.2273272073511787, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5575 + }, + { + "epoch": 2.227726727926488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5576 + }, + { + "epoch": 2.228126248501798, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5577 + }, + { + "epoch": 2.2285257690771076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5578 + }, + { + "epoch": 2.228925289652417, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5579 + }, + { + "epoch": 2.2293248102277268, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5580 + }, + { + "epoch": 2.2297243308030366, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5581 + }, + { + "epoch": 2.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5582 + }, + { + "epoch": 2.2305233719536557, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5583 + }, + { + "epoch": 2.230922892528965, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5584 + }, + { + "epoch": 2.231322413104275, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5585 + }, + { + "epoch": 2.2317219336795846, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5586 + }, + { + "epoch": 2.232121454254894, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 5587 + }, + { + "epoch": 2.2325209748302037, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5588 + }, + { + "epoch": 2.2329204954055135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 5589 + }, + { + "epoch": 2.233320015980823, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5590 + }, + { + "epoch": 2.2337195365561326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5591 + }, + { + "epoch": 2.2341190571314424, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5592 + }, + { + "epoch": 2.2345185777067518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 5593 + }, + { + "epoch": 2.2349180982820616, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5594 + }, + { + "epoch": 2.2353176188573713, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5595 + }, + { + "epoch": 2.2357171394326807, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5596 + }, + { + "epoch": 2.2361166600079905, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5597 + }, + { + "epoch": 2.2365161805833003, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5598 + }, + { + "epoch": 2.2369157011586096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5599 + }, + { + "epoch": 2.2373152217339194, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5600 + }, + { + "epoch": 2.2377147423092287, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5601 + }, + { + "epoch": 2.2381142628845385, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5602 + }, + { + "epoch": 2.2385137834598483, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5603 + }, + { + "epoch": 2.2389133040351576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5604 + }, + { + "epoch": 2.2393128246104674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 5605 + }, + { + "epoch": 2.239712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5606 + }, + { + "epoch": 2.2401118657610866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5607 + }, + { + "epoch": 2.2405113863363963, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5608 + }, + { + "epoch": 2.240910906911706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5609 + }, + { + "epoch": 2.2413104274870155, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5610 + }, + { + "epoch": 2.2417099480623253, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5611 + }, + { + "epoch": 2.2421094686376346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5612 + }, + { + "epoch": 2.2425089892129444, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5613 + }, + { + "epoch": 2.242908509788254, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5614 + }, + { + "epoch": 2.2433080303635635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5615 + }, + { + "epoch": 2.2437075509388733, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5616 + }, + { + "epoch": 2.244107071514183, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5617 + }, + { + "epoch": 2.2445065920894924, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 5618 + }, + { + "epoch": 2.2449061126648022, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5619 + }, + { + "epoch": 2.245305633240112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5620 + }, + { + "epoch": 2.2457051538154214, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5621 + }, + { + "epoch": 2.246104674390731, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5622 + }, + { + "epoch": 2.246504194966041, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5623 + }, + { + "epoch": 2.2469037155413503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5624 + }, + { + "epoch": 2.24730323611666, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5625 + }, + { + "epoch": 2.24770275669197, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5626 + }, + { + "epoch": 2.248102277267279, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5627 + }, + { + "epoch": 2.248501797842589, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5628 + }, + { + "epoch": 2.2489013184178983, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5629 + }, + { + "epoch": 2.249300838993208, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5630 + }, + { + "epoch": 2.249700359568518, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5631 + }, + { + "epoch": 2.2500998801438272, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 5632 + }, + { + "epoch": 2.250499400719137, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5633 + }, + { + "epoch": 2.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5634 + }, + { + "epoch": 2.251298441869756, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5635 + }, + { + "epoch": 2.251697962445066, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5636 + }, + { + "epoch": 2.2520974830203757, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5637 + }, + { + "epoch": 2.252497003595685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 5638 + }, + { + "epoch": 2.252896524170995, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5639 + }, + { + "epoch": 2.2532960447463046, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 5640 + }, + { + "epoch": 2.253695565321614, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5641 + }, + { + "epoch": 2.2540950858969238, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5642 + }, + { + "epoch": 2.2544946064722335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5643 + }, + { + "epoch": 2.254894127047543, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5644 + }, + { + "epoch": 2.2552936476228527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5645 + }, + { + "epoch": 2.255693168198162, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5646 + }, + { + "epoch": 2.256092688773472, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5647 + }, + { + "epoch": 2.2564922093487816, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5648 + }, + { + "epoch": 2.256891729924091, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5649 + }, + { + "epoch": 2.2572912504994007, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5650 + }, + { + "epoch": 2.2576907710747105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 5651 + }, + { + "epoch": 2.25809029165002, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5652 + }, + { + "epoch": 2.2584898122253296, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5653 + }, + { + "epoch": 2.2588893328006394, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5654 + }, + { + "epoch": 2.2592888533759488, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5655 + }, + { + "epoch": 2.2596883739512585, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5656 + }, + { + "epoch": 2.260087894526568, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5657 + }, + { + "epoch": 2.2604874151018777, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5658 + }, + { + "epoch": 2.2608869356771875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5659 + }, + { + "epoch": 2.2612864562524972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5660 + }, + { + "epoch": 2.2616859768278066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5661 + }, + { + "epoch": 2.2620854974031164, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5662 + }, + { + "epoch": 2.2624850179784257, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5663 + }, + { + "epoch": 2.2628845385537355, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5664 + }, + { + "epoch": 2.2632840591290453, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5665 + }, + { + "epoch": 2.2636835797043546, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5666 + }, + { + "epoch": 2.2640831002796644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 5667 + }, + { + "epoch": 2.264482620854974, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5668 + }, + { + "epoch": 2.2648821414302835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5669 + }, + { + "epoch": 2.2652816620055933, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5670 + }, + { + "epoch": 2.265681182580903, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5671 + }, + { + "epoch": 2.2660807031562125, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5672 + }, + { + "epoch": 2.2664802237315222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5673 + }, + { + "epoch": 2.2668797443068316, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5674 + }, + { + "epoch": 2.2672792648821414, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5675 + }, + { + "epoch": 2.267678785457451, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5676 + }, + { + "epoch": 2.2680783060327605, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5677 + }, + { + "epoch": 2.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5678 + }, + { + "epoch": 2.26887734718338, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 5679 + }, + { + "epoch": 2.2692768677586894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5680 + }, + { + "epoch": 2.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5681 + }, + { + "epoch": 2.270075908909309, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5682 + }, + { + "epoch": 2.2704754294846183, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5683 + }, + { + "epoch": 2.270874950059928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5684 + }, + { + "epoch": 2.271274470635238, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5685 + }, + { + "epoch": 2.2716739912105472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5686 + }, + { + "epoch": 2.272073511785857, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5687 + }, + { + "epoch": 2.272473032361167, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5688 + }, + { + "epoch": 2.272872552936476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5689 + }, + { + "epoch": 2.273272073511786, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5690 + }, + { + "epoch": 2.2736715940870953, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5691 + }, + { + "epoch": 2.274071114662405, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5692 + }, + { + "epoch": 2.274470635237715, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5693 + }, + { + "epoch": 2.274870155813024, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5694 + }, + { + "epoch": 2.275269676388334, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5695 + }, + { + "epoch": 2.275669196963644, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 5696 + }, + { + "epoch": 2.276068717538953, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5697 + }, + { + "epoch": 2.276468238114263, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5698 + }, + { + "epoch": 2.2768677586895727, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5699 + }, + { + "epoch": 2.277267279264882, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 5700 + }, + { + "epoch": 2.277666799840192, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5701 + }, + { + "epoch": 2.278066320415501, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5702 + }, + { + "epoch": 2.278465840990811, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5703 + }, + { + "epoch": 2.2788653615661207, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5704 + }, + { + "epoch": 2.2792648821414305, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5705 + }, + { + "epoch": 2.27966440271674, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5706 + }, + { + "epoch": 2.2800639232920497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5707 + }, + { + "epoch": 2.280463443867359, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5708 + }, + { + "epoch": 2.280862964442669, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5709 + }, + { + "epoch": 2.2812624850179786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5710 + }, + { + "epoch": 2.281662005593288, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5711 + }, + { + "epoch": 2.2820615261685977, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5712 + }, + { + "epoch": 2.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5713 + }, + { + "epoch": 2.282860567319217, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5714 + }, + { + "epoch": 2.2832600878945266, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5715 + }, + { + "epoch": 2.2836596084698364, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5716 + }, + { + "epoch": 2.2840591290451457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5717 + }, + { + "epoch": 2.2844586496204555, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5718 + }, + { + "epoch": 2.284858170195765, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5719 + }, + { + "epoch": 2.2852576907710747, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5720 + }, + { + "epoch": 2.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5721 + }, + { + "epoch": 2.286056731921694, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5722 + }, + { + "epoch": 2.2864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5723 + }, + { + "epoch": 2.2868557730723134, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5724 + }, + { + "epoch": 2.2872552936476227, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5725 + }, + { + "epoch": 2.2876548142229325, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5726 + }, + { + "epoch": 2.2880543347982423, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5727 + }, + { + "epoch": 2.2884538553735516, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5728 + }, + { + "epoch": 2.2888533759488614, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5729 + }, + { + "epoch": 2.289252896524171, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5730 + }, + { + "epoch": 2.2896524170994805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5731 + }, + { + "epoch": 2.2900519376747903, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5732 + }, + { + "epoch": 2.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5733 + }, + { + "epoch": 2.2908509788254094, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5734 + }, + { + "epoch": 2.2912504994007192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 5735 + }, + { + "epoch": 2.2916500199760286, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 5736 + }, + { + "epoch": 2.2920495405513384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5737 + }, + { + "epoch": 2.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5738 + }, + { + "epoch": 2.2928485817019575, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5739 + }, + { + "epoch": 2.2932481022772673, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5740 + }, + { + "epoch": 2.293647622852577, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5741 + }, + { + "epoch": 2.2940471434278864, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5742 + }, + { + "epoch": 2.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5743 + }, + { + "epoch": 2.294846184578506, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5744 + }, + { + "epoch": 2.2952457051538153, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5745 + }, + { + "epoch": 2.295645225729125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5746 + }, + { + "epoch": 2.2960447463044344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5747 + }, + { + "epoch": 2.2964442668797442, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5748 + }, + { + "epoch": 2.296843787455054, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5749 + }, + { + "epoch": 2.297243308030364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5750 + }, + { + "epoch": 2.297642828605673, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5751 + }, + { + "epoch": 2.298042349180983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5752 + }, + { + "epoch": 2.2984418697562923, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5753 + }, + { + "epoch": 2.298841390331602, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5754 + }, + { + "epoch": 2.299240910906912, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5755 + }, + { + "epoch": 2.299640431482221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5756 + }, + { + "epoch": 2.300039952057531, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5757 + }, + { + "epoch": 2.3004394726328408, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5758 + }, + { + "epoch": 2.30083899320815, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 5759 + }, + { + "epoch": 2.30123851378346, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5760 + }, + { + "epoch": 2.3016380343587697, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5761 + }, + { + "epoch": 2.302037554934079, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5762 + }, + { + "epoch": 2.302437075509389, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5763 + }, + { + "epoch": 2.302836596084698, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5764 + }, + { + "epoch": 2.303236116660008, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5765 + }, + { + "epoch": 2.3036356372353177, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5766 + }, + { + "epoch": 2.304035157810627, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5767 + }, + { + "epoch": 2.304434678385937, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5768 + }, + { + "epoch": 2.3048341989612466, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5769 + }, + { + "epoch": 2.305233719536556, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5770 + }, + { + "epoch": 2.3056332401118658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5771 + }, + { + "epoch": 2.3060327606871756, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5772 + }, + { + "epoch": 2.306432281262485, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5773 + }, + { + "epoch": 2.3068318018377947, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5774 + }, + { + "epoch": 2.3072313224131045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5775 + }, + { + "epoch": 2.307630842988414, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5776 + }, + { + "epoch": 2.3080303635637236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5777 + }, + { + "epoch": 2.3084298841390334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5778 + }, + { + "epoch": 2.3088294047143427, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5779 + }, + { + "epoch": 2.3092289252896525, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5780 + }, + { + "epoch": 2.309628445864962, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5781 + }, + { + "epoch": 2.3100279664402716, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5782 + }, + { + "epoch": 2.3104274870155814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5783 + }, + { + "epoch": 2.3108270075908908, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5784 + }, + { + "epoch": 2.3112265281662006, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5785 + }, + { + "epoch": 2.3116260487415103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5786 + }, + { + "epoch": 2.3120255693168197, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5787 + }, + { + "epoch": 2.3124250898921295, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5788 + }, + { + "epoch": 2.3128246104674393, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5789 + }, + { + "epoch": 2.3132241310427486, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5790 + }, + { + "epoch": 2.3136236516180584, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5791 + }, + { + "epoch": 2.3140231721933677, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5792 + }, + { + "epoch": 2.3144226927686775, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5793 + }, + { + "epoch": 2.3148222133439873, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 5794 + }, + { + "epoch": 2.315221733919297, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5795 + }, + { + "epoch": 2.3156212544946064, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5796 + }, + { + "epoch": 2.316020775069916, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5797 + }, + { + "epoch": 2.3164202956452256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5798 + }, + { + "epoch": 2.3168198162205353, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5799 + }, + { + "epoch": 2.317219336795845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5800 + }, + { + "epoch": 2.3176188573711545, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5801 + }, + { + "epoch": 2.3180183779464643, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5802 + }, + { + "epoch": 2.318417898521774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5803 + }, + { + "epoch": 2.3188174190970834, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5804 + }, + { + "epoch": 2.319216939672393, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5805 + }, + { + "epoch": 2.319616460247703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5806 + }, + { + "epoch": 2.3200159808230123, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5807 + }, + { + "epoch": 2.320415501398322, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5808 + }, + { + "epoch": 2.3208150219736314, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5809 + }, + { + "epoch": 2.321214542548941, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5810 + }, + { + "epoch": 2.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5811 + }, + { + "epoch": 2.3220135836995603, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5812 + }, + { + "epoch": 2.32241310427487, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5813 + }, + { + "epoch": 2.32281262485018, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5814 + }, + { + "epoch": 2.3232121454254893, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5815 + }, + { + "epoch": 2.323611666000799, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5816 + }, + { + "epoch": 2.324011186576109, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5817 + }, + { + "epoch": 2.324410707151418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5818 + }, + { + "epoch": 2.324810227726728, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5819 + }, + { + "epoch": 2.3252097483020377, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5820 + }, + { + "epoch": 2.325609268877347, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5821 + }, + { + "epoch": 2.326008789452657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5822 + }, + { + "epoch": 2.3264083100279667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5823 + }, + { + "epoch": 2.326807830603276, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5824 + }, + { + "epoch": 2.327207351178586, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5825 + }, + { + "epoch": 2.327606871753895, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5826 + }, + { + "epoch": 2.328006392329205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5827 + }, + { + "epoch": 2.3284059129045147, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5828 + }, + { + "epoch": 2.328805433479824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5829 + }, + { + "epoch": 2.329204954055134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5830 + }, + { + "epoch": 2.3296044746304436, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5831 + }, + { + "epoch": 2.330003995205753, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5832 + }, + { + "epoch": 2.3304035157810628, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 5833 + }, + { + "epoch": 2.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5834 + }, + { + "epoch": 2.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5835 + }, + { + "epoch": 2.3316020775069917, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5836 + }, + { + "epoch": 2.332001598082301, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5837 + }, + { + "epoch": 2.332401118657611, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 5838 + }, + { + "epoch": 2.3328006392329206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5839 + }, + { + "epoch": 2.3332001598082304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5840 + }, + { + "epoch": 2.3335996803835397, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5841 + }, + { + "epoch": 2.3339992009588495, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5842 + }, + { + "epoch": 2.334398721534159, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5843 + }, + { + "epoch": 2.3347982421094686, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5844 + }, + { + "epoch": 2.3351977626847784, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5845 + }, + { + "epoch": 2.3355972832600878, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5846 + }, + { + "epoch": 2.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5847 + }, + { + "epoch": 2.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5848 + }, + { + "epoch": 2.3367958449860167, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5849 + }, + { + "epoch": 2.3371953655613265, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5850 + }, + { + "epoch": 2.3375948861366362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5851 + }, + { + "epoch": 2.3379944067119456, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5852 + }, + { + "epoch": 2.3383939272872554, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5853 + }, + { + "epoch": 2.3387934478625647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5854 + }, + { + "epoch": 2.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5855 + }, + { + "epoch": 2.3395924890131843, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5856 + }, + { + "epoch": 2.3399920095884936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5857 + }, + { + "epoch": 2.3403915301638034, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5858 + }, + { + "epoch": 2.340791050739113, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5859 + }, + { + "epoch": 2.3411905713144225, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5860 + }, + { + "epoch": 2.3415900918897323, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5861 + }, + { + "epoch": 2.341989612465042, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5862 + }, + { + "epoch": 2.3423891330403515, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5863 + }, + { + "epoch": 2.3427886536156612, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5864 + }, + { + "epoch": 2.343188174190971, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5865 + }, + { + "epoch": 2.3435876947662804, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5866 + }, + { + "epoch": 2.34398721534159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5867 + }, + { + "epoch": 2.3443867359169, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5868 + }, + { + "epoch": 2.3447862564922093, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5869 + }, + { + "epoch": 2.345185777067519, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5870 + }, + { + "epoch": 2.3455852976428284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5871 + }, + { + "epoch": 2.345984818218138, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5872 + }, + { + "epoch": 2.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5873 + }, + { + "epoch": 2.3467838593687573, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5874 + }, + { + "epoch": 2.347183379944067, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5875 + }, + { + "epoch": 2.347582900519377, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5876 + }, + { + "epoch": 2.3479824210946862, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 5877 + }, + { + "epoch": 2.348381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5878 + }, + { + "epoch": 2.348781462245306, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5879 + }, + { + "epoch": 2.349180982820615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5880 + }, + { + "epoch": 2.349580503395925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 5881 + }, + { + "epoch": 2.3499800239712343, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5882 + }, + { + "epoch": 2.350379544546544, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5883 + }, + { + "epoch": 2.350779065121854, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5884 + }, + { + "epoch": 2.3511785856971636, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5885 + }, + { + "epoch": 2.351578106272473, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5886 + }, + { + "epoch": 2.3519776268477828, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5887 + }, + { + "epoch": 2.352377147423092, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5888 + }, + { + "epoch": 2.352776667998402, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5889 + }, + { + "epoch": 2.3531761885737117, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5890 + }, + { + "epoch": 2.353575709149021, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5891 + }, + { + "epoch": 2.353975229724331, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5892 + }, + { + "epoch": 2.3543747502996406, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5893 + }, + { + "epoch": 2.35477427087495, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5894 + }, + { + "epoch": 2.3551737914502597, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5895 + }, + { + "epoch": 2.3555733120255695, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5896 + }, + { + "epoch": 2.355972832600879, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 5897 + }, + { + "epoch": 2.3563723531761886, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5898 + }, + { + "epoch": 2.356771873751498, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5899 + }, + { + "epoch": 2.357171394326808, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5900 + }, + { + "epoch": 2.3575709149021176, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5901 + }, + { + "epoch": 2.357970435477427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5902 + }, + { + "epoch": 2.3583699560527367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5903 + }, + { + "epoch": 2.3587694766280465, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5904 + }, + { + "epoch": 2.359168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5905 + }, + { + "epoch": 2.3595685177786656, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 5906 + }, + { + "epoch": 2.3599680383539754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5907 + }, + { + "epoch": 2.3603675589292847, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5908 + }, + { + "epoch": 2.3607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5909 + }, + { + "epoch": 2.361166600079904, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5910 + }, + { + "epoch": 2.3615661206552137, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5911 + }, + { + "epoch": 2.3619656412305234, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5912 + }, + { + "epoch": 2.3623651618058332, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5913 + }, + { + "epoch": 2.3627646823811426, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5914 + }, + { + "epoch": 2.3631642029564524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5915 + }, + { + "epoch": 2.3635637235317617, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5916 + }, + { + "epoch": 2.3639632441070715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5917 + }, + { + "epoch": 2.3643627646823813, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5918 + }, + { + "epoch": 2.3647622852576906, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5919 + }, + { + "epoch": 2.3651618058330004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 5920 + }, + { + "epoch": 2.36556132640831, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5921 + }, + { + "epoch": 2.3659608469836195, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5922 + }, + { + "epoch": 2.3663603675589293, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5923 + }, + { + "epoch": 2.366759888134239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5924 + }, + { + "epoch": 2.3671594087095484, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5925 + }, + { + "epoch": 2.3675589292848582, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5926 + }, + { + "epoch": 2.3679584498601676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5927 + }, + { + "epoch": 2.3683579704354774, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5928 + }, + { + "epoch": 2.368757491010787, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5929 + }, + { + "epoch": 2.369157011586097, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5930 + }, + { + "epoch": 2.3695565321614063, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5931 + }, + { + "epoch": 2.369956052736716, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5932 + }, + { + "epoch": 2.3703555733120254, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5933 + }, + { + "epoch": 2.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5934 + }, + { + "epoch": 2.371154614462645, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5935 + }, + { + "epoch": 2.3715541350379543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5936 + }, + { + "epoch": 2.371953655613264, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5937 + }, + { + "epoch": 2.372353176188574, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5938 + }, + { + "epoch": 2.3727526967638832, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5939 + }, + { + "epoch": 2.373152217339193, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5940 + }, + { + "epoch": 2.373551737914503, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5941 + }, + { + "epoch": 2.373951258489812, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5942 + }, + { + "epoch": 2.374350779065122, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 5943 + }, + { + "epoch": 2.3747502996404313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5944 + }, + { + "epoch": 2.375149820215741, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5945 + }, + { + "epoch": 2.375549340791051, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5946 + }, + { + "epoch": 2.37594886136636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5947 + }, + { + "epoch": 2.37634838194167, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5948 + }, + { + "epoch": 2.3767479025169798, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5949 + }, + { + "epoch": 2.377147423092289, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 5950 + }, + { + "epoch": 2.377546943667599, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5951 + }, + { + "epoch": 2.3779464642429087, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5952 + }, + { + "epoch": 2.378345984818218, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 5953 + }, + { + "epoch": 2.378745505393528, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5954 + }, + { + "epoch": 2.379145025968837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5955 + }, + { + "epoch": 2.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5956 + }, + { + "epoch": 2.3799440671194567, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5957 + }, + { + "epoch": 2.3803435876947665, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5958 + }, + { + "epoch": 2.380743108270076, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5959 + }, + { + "epoch": 2.3811426288453856, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5960 + }, + { + "epoch": 2.381542149420695, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5961 + }, + { + "epoch": 2.3819416699960048, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5962 + }, + { + "epoch": 2.3823411905713145, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5963 + }, + { + "epoch": 2.382740711146624, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5964 + }, + { + "epoch": 2.3831402317219337, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5965 + }, + { + "epoch": 2.3835397522972435, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5966 + }, + { + "epoch": 2.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5967 + }, + { + "epoch": 2.3843387934478626, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5968 + }, + { + "epoch": 2.3847383140231724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5969 + }, + { + "epoch": 2.3851378345984817, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5970 + }, + { + "epoch": 2.3855373551737915, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 5971 + }, + { + "epoch": 2.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5972 + }, + { + "epoch": 2.3863363963244106, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 5973 + }, + { + "epoch": 2.3867359168997204, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5974 + }, + { + "epoch": 2.38713543747503, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5975 + }, + { + "epoch": 2.3875349580503396, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5976 + }, + { + "epoch": 2.3879344786256493, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5977 + }, + { + "epoch": 2.3883339992009587, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5978 + }, + { + "epoch": 2.3887335197762685, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5979 + }, + { + "epoch": 2.3891330403515783, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5980 + }, + { + "epoch": 2.3895325609268876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5981 + }, + { + "epoch": 2.3899320815021974, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5982 + }, + { + "epoch": 2.390331602077507, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5983 + }, + { + "epoch": 2.3907311226528165, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5984 + }, + { + "epoch": 2.3911306432281263, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5985 + }, + { + "epoch": 2.391530163803436, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 5986 + }, + { + "epoch": 2.3919296843787454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5987 + }, + { + "epoch": 2.392329204954055, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5988 + }, + { + "epoch": 2.3927287255293646, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5989 + }, + { + "epoch": 2.3931282461046743, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5990 + }, + { + "epoch": 2.393527766679984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5991 + }, + { + "epoch": 2.3939272872552935, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5992 + }, + { + "epoch": 2.3943268078306033, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5993 + }, + { + "epoch": 2.394726328405913, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5994 + }, + { + "epoch": 2.3951258489812224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5995 + }, + { + "epoch": 2.395525369556532, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5996 + }, + { + "epoch": 2.395924890131842, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5997 + }, + { + "epoch": 2.3963244107071513, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5998 + }, + { + "epoch": 2.396723931282461, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5999 + }, + { + "epoch": 2.3971234518577704, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6000 + }, + { + "epoch": 2.39752297243308, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6001 + }, + { + "epoch": 2.39792249300839, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6002 + }, + { + "epoch": 2.3983220135837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6003 + }, + { + "epoch": 2.398721534159009, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6004 + }, + { + "epoch": 2.399121054734319, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6005 + }, + { + "epoch": 2.3995205753096283, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6006 + }, + { + "epoch": 2.399920095884938, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6007 + }, + { + "epoch": 2.400319616460248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6008 + }, + { + "epoch": 2.400719137035557, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6009 + }, + { + "epoch": 2.401118657610867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6010 + }, + { + "epoch": 2.4015181781861767, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6011 + }, + { + "epoch": 2.401917698761486, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6012 + }, + { + "epoch": 2.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6013 + }, + { + "epoch": 2.4027167399121057, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6014 + }, + { + "epoch": 2.403116260487415, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6015 + }, + { + "epoch": 2.403515781062725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6016 + }, + { + "epoch": 2.403915301638034, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6017 + }, + { + "epoch": 2.404314822213344, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6018 + }, + { + "epoch": 2.4047143427886537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6019 + }, + { + "epoch": 2.4051138633639635, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6020 + }, + { + "epoch": 2.405513383939273, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6021 + }, + { + "epoch": 2.4059129045145826, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6022 + }, + { + "epoch": 2.406312425089892, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6023 + }, + { + "epoch": 2.4067119456652017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6024 + }, + { + "epoch": 2.4071114662405115, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6025 + }, + { + "epoch": 2.407510986815821, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6026 + }, + { + "epoch": 2.4079105073911307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6027 + }, + { + "epoch": 2.4083100279664404, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6028 + }, + { + "epoch": 2.40870954854175, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 6029 + }, + { + "epoch": 2.4091090691170596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6030 + }, + { + "epoch": 2.4095085896923694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6031 + }, + { + "epoch": 2.4099081102676787, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6032 + }, + { + "epoch": 2.4103076308429885, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 6033 + }, + { + "epoch": 2.410707151418298, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 6034 + }, + { + "epoch": 2.4111066719936076, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6035 + }, + { + "epoch": 2.4115061925689174, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6036 + }, + { + "epoch": 2.4119057131442267, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6037 + }, + { + "epoch": 2.4123052337195365, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6038 + }, + { + "epoch": 2.4127047542948463, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6039 + }, + { + "epoch": 2.4131042748701557, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6040 + }, + { + "epoch": 2.4135037954454654, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6041 + }, + { + "epoch": 2.4139033160207752, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6042 + }, + { + "epoch": 2.4143028365960846, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 6043 + }, + { + "epoch": 2.4147023571713944, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6044 + }, + { + "epoch": 2.4151018777467037, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6045 + }, + { + "epoch": 2.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6046 + }, + { + "epoch": 2.4159009188973233, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6047 + }, + { + "epoch": 2.416300439472633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 6048 + }, + { + "epoch": 2.4166999600479424, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6049 + }, + { + "epoch": 2.417099480623252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6050 + }, + { + "epoch": 2.4174990011985615, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6051 + }, + { + "epoch": 2.4178985217738713, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6052 + }, + { + "epoch": 2.418298042349181, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6053 + }, + { + "epoch": 2.4186975629244905, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6054 + }, + { + "epoch": 2.4190970834998002, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6055 + }, + { + "epoch": 2.41949660407511, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6056 + }, + { + "epoch": 2.4198961246504194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6057 + }, + { + "epoch": 2.420295645225729, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6058 + }, + { + "epoch": 2.420695165801039, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6059 + }, + { + "epoch": 2.4210946863763483, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6060 + }, + { + "epoch": 2.421494206951658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6061 + }, + { + "epoch": 2.4218937275269674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 6062 + }, + { + "epoch": 2.422293248102277, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6063 + }, + { + "epoch": 2.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6064 + }, + { + "epoch": 2.4230922892528968, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6065 + }, + { + "epoch": 2.423491809828206, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6066 + }, + { + "epoch": 2.423891330403516, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6067 + }, + { + "epoch": 2.4242908509788252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6068 + }, + { + "epoch": 2.424690371554135, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6069 + }, + { + "epoch": 2.425089892129445, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6070 + }, + { + "epoch": 2.425489412704754, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6071 + }, + { + "epoch": 2.425888933280064, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6072 + }, + { + "epoch": 2.4262884538553737, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6073 + }, + { + "epoch": 2.426687974430683, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6074 + }, + { + "epoch": 2.427087495005993, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6075 + }, + { + "epoch": 2.4274870155813026, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6076 + }, + { + "epoch": 2.427886536156612, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6077 + }, + { + "epoch": 2.4282860567319218, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6078 + }, + { + "epoch": 2.428685577307231, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6079 + }, + { + "epoch": 2.429085097882541, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6080 + }, + { + "epoch": 2.4294846184578507, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6081 + }, + { + "epoch": 2.42988413903316, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6082 + }, + { + "epoch": 2.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6083 + }, + { + "epoch": 2.4306831801837796, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6084 + }, + { + "epoch": 2.431082700759089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6085 + }, + { + "epoch": 2.4314822213343987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6086 + }, + { + "epoch": 2.4318817419097085, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6087 + }, + { + "epoch": 2.432281262485018, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6088 + }, + { + "epoch": 2.4326807830603276, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6089 + }, + { + "epoch": 2.433080303635637, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6090 + }, + { + "epoch": 2.4334798242109468, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6091 + }, + { + "epoch": 2.4338793447862566, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 6092 + }, + { + "epoch": 2.4342788653615663, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6093 + }, + { + "epoch": 2.4346783859368757, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6094 + }, + { + "epoch": 2.4350779065121855, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 6095 + }, + { + "epoch": 2.435477427087495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 6096 + }, + { + "epoch": 2.4358769476628046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 6097 + }, + { + "epoch": 2.4362764682381144, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6098 + }, + { + "epoch": 2.4366759888134237, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6099 + }, + { + "epoch": 2.4370755093887335, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6100 + }, + { + "epoch": 2.4374750299640433, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6101 + }, + { + "epoch": 2.4378745505393526, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6102 + }, + { + "epoch": 2.4382740711146624, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6103 + }, + { + "epoch": 2.438673591689972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6104 + }, + { + "epoch": 2.4390731122652816, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6105 + }, + { + "epoch": 2.4394726328405913, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6106 + }, + { + "epoch": 2.4398721534159007, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6107 + }, + { + "epoch": 2.4402716739912105, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6108 + }, + { + "epoch": 2.4406711945665203, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6109 + }, + { + "epoch": 2.44107071514183, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6110 + }, + { + "epoch": 2.4414702357171394, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6111 + }, + { + "epoch": 2.441869756292449, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6112 + }, + { + "epoch": 2.4422692768677585, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6113 + }, + { + "epoch": 2.4426687974430683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6114 + }, + { + "epoch": 2.443068318018378, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6115 + }, + { + "epoch": 2.4434678385936874, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6116 + }, + { + "epoch": 2.443867359168997, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6117 + }, + { + "epoch": 2.444266879744307, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 6118 + }, + { + "epoch": 2.4446664003196164, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6119 + }, + { + "epoch": 2.445065920894926, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6120 + }, + { + "epoch": 2.445465441470236, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6121 + }, + { + "epoch": 2.4458649620455453, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6122 + }, + { + "epoch": 2.446264482620855, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6123 + }, + { + "epoch": 2.4466640031961644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6124 + }, + { + "epoch": 2.447063523771474, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6125 + }, + { + "epoch": 2.447463044346784, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6126 + }, + { + "epoch": 2.4478625649220933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6127 + }, + { + "epoch": 2.448262085497403, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6128 + }, + { + "epoch": 2.448661606072713, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6129 + }, + { + "epoch": 2.4490611266480222, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6130 + }, + { + "epoch": 2.449460647223332, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6131 + }, + { + "epoch": 2.449860167798642, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6132 + }, + { + "epoch": 2.450259688373951, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6133 + }, + { + "epoch": 2.450659208949261, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6134 + }, + { + "epoch": 2.4510587295245703, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 6135 + }, + { + "epoch": 2.45145825009988, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6136 + }, + { + "epoch": 2.45185777067519, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6137 + }, + { + "epoch": 2.4522572912504996, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6138 + }, + { + "epoch": 2.452656811825809, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6139 + }, + { + "epoch": 2.4530563324011188, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6140 + }, + { + "epoch": 2.453455852976428, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6141 + }, + { + "epoch": 2.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6142 + }, + { + "epoch": 2.4542548941270477, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6143 + }, + { + "epoch": 2.454654414702357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6144 + }, + { + "epoch": 2.455053935277667, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6145 + }, + { + "epoch": 2.4554534558529766, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6146 + }, + { + "epoch": 2.455852976428286, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6147 + }, + { + "epoch": 2.4562524970035957, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6148 + }, + { + "epoch": 2.4566520175789055, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6149 + }, + { + "epoch": 2.457051538154215, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6150 + }, + { + "epoch": 2.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6151 + }, + { + "epoch": 2.457850579304834, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6152 + }, + { + "epoch": 2.4582500998801438, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6153 + }, + { + "epoch": 2.4586496204554535, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6154 + }, + { + "epoch": 2.4590491410307633, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6155 + }, + { + "epoch": 2.4594486616060727, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6156 + }, + { + "epoch": 2.4598481821813825, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6157 + }, + { + "epoch": 2.460247702756692, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6158 + }, + { + "epoch": 2.4606472233320016, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6159 + }, + { + "epoch": 2.4610467439073114, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6160 + }, + { + "epoch": 2.4614462644826207, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6161 + }, + { + "epoch": 2.4618457850579305, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6162 + }, + { + "epoch": 2.4622453056332403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6163 + }, + { + "epoch": 2.4626448262085496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6164 + }, + { + "epoch": 2.4630443467838594, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6165 + }, + { + "epoch": 2.463443867359169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6166 + }, + { + "epoch": 2.4638433879344785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6167 + }, + { + "epoch": 2.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6168 + }, + { + "epoch": 2.4646424290850977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 6169 + }, + { + "epoch": 2.4650419496604075, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6170 + }, + { + "epoch": 2.4654414702357172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6171 + }, + { + "epoch": 2.4658409908110266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6172 + }, + { + "epoch": 2.4662405113863364, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6173 + }, + { + "epoch": 2.466640031961646, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6174 + }, + { + "epoch": 2.4670395525369555, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6175 + }, + { + "epoch": 2.4674390731122653, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6176 + }, + { + "epoch": 2.467838593687575, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6177 + }, + { + "epoch": 2.4682381142628844, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6178 + }, + { + "epoch": 2.468637634838194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6179 + }, + { + "epoch": 2.4690371554135035, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6180 + }, + { + "epoch": 2.4694366759888133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6181 + }, + { + "epoch": 2.469836196564123, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6182 + }, + { + "epoch": 2.470235717139433, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6183 + }, + { + "epoch": 2.4706352377147422, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6184 + }, + { + "epoch": 2.471034758290052, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6185 + }, + { + "epoch": 2.4714342788653614, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 6186 + }, + { + "epoch": 2.471833799440671, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6187 + }, + { + "epoch": 2.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6188 + }, + { + "epoch": 2.4726328405912903, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6189 + }, + { + "epoch": 2.4730323611666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6190 + }, + { + "epoch": 2.47343188174191, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6191 + }, + { + "epoch": 2.473831402317219, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6192 + }, + { + "epoch": 2.474230922892529, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6193 + }, + { + "epoch": 2.474630443467839, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6194 + }, + { + "epoch": 2.475029964043148, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6195 + }, + { + "epoch": 2.475429484618458, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6196 + }, + { + "epoch": 2.4758290051937673, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6197 + }, + { + "epoch": 2.476228525769077, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6198 + }, + { + "epoch": 2.476628046344387, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6199 + }, + { + "epoch": 2.4770275669196966, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6200 + }, + { + "epoch": 2.477427087495006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6201 + }, + { + "epoch": 2.4778266080703157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6202 + }, + { + "epoch": 2.478226128645625, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6203 + }, + { + "epoch": 2.478625649220935, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6204 + }, + { + "epoch": 2.4790251697962447, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6205 + }, + { + "epoch": 2.479424690371554, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6206 + }, + { + "epoch": 2.479824210946864, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6207 + }, + { + "epoch": 2.4802237315221736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6208 + }, + { + "epoch": 2.480623252097483, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6209 + }, + { + "epoch": 2.4810227726727927, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6210 + }, + { + "epoch": 2.4814222932481025, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6211 + }, + { + "epoch": 2.481821813823412, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6212 + }, + { + "epoch": 2.4822213343987216, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6213 + }, + { + "epoch": 2.482620854974031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 6214 + }, + { + "epoch": 2.4830203755493407, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6215 + }, + { + "epoch": 2.4834198961246505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6216 + }, + { + "epoch": 2.48381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 6217 + }, + { + "epoch": 2.4842189372752697, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6218 + }, + { + "epoch": 2.4846184578505794, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6219 + }, + { + "epoch": 2.485017978425889, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6220 + }, + { + "epoch": 2.4854174990011986, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6221 + }, + { + "epoch": 2.4858170195765084, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6222 + }, + { + "epoch": 2.4862165401518177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6223 + }, + { + "epoch": 2.4866160607271275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6224 + }, + { + "epoch": 2.487015581302437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6225 + }, + { + "epoch": 2.4874151018777466, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6226 + }, + { + "epoch": 2.4878146224530564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6227 + }, + { + "epoch": 2.488214143028366, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6228 + }, + { + "epoch": 2.4886136636036755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6229 + }, + { + "epoch": 2.4890131841789853, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6230 + }, + { + "epoch": 2.4894127047542947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6231 + }, + { + "epoch": 2.4898122253296044, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6232 + }, + { + "epoch": 2.4902117459049142, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6233 + }, + { + "epoch": 2.4906112664802236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6234 + }, + { + "epoch": 2.4910107870555334, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6235 + }, + { + "epoch": 2.491410307630843, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6236 + }, + { + "epoch": 2.4918098282061525, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6237 + }, + { + "epoch": 2.4922093487814623, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6238 + }, + { + "epoch": 2.492608869356772, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6239 + }, + { + "epoch": 2.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6240 + }, + { + "epoch": 2.493407910507391, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6241 + }, + { + "epoch": 2.4938074310827005, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6242 + }, + { + "epoch": 2.4942069516580103, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6243 + }, + { + "epoch": 2.49460647223332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6244 + }, + { + "epoch": 2.49500599280863, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6245 + }, + { + "epoch": 2.4954055133839392, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 6246 + }, + { + "epoch": 2.495805033959249, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6247 + }, + { + "epoch": 2.4962045545345584, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6248 + }, + { + "epoch": 2.496604075109868, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6249 + }, + { + "epoch": 2.497003595685178, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6250 + }, + { + "epoch": 2.4974031162604873, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6251 + }, + { + "epoch": 2.497802636835797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6252 + }, + { + "epoch": 2.498202157411107, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6253 + }, + { + "epoch": 2.498601677986416, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6254 + }, + { + "epoch": 2.499001198561726, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 6255 + }, + { + "epoch": 2.4994007191370358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6256 + }, + { + "epoch": 2.499800239712345, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6257 + }, + { + "epoch": 2.500199760287655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 6258 + }, + { + "epoch": 2.5005992808629642, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6259 + }, + { + "epoch": 2.500998801438274, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6260 + }, + { + "epoch": 2.501398322013584, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6261 + }, + { + "epoch": 2.5017978425888936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6262 + }, + { + "epoch": 2.502197363164203, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6263 + }, + { + "epoch": 2.5025968837395127, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6264 + }, + { + "epoch": 2.502996404314822, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6265 + }, + { + "epoch": 2.503395924890132, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6266 + }, + { + "epoch": 2.5037954454654416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6267 + }, + { + "epoch": 2.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6268 + }, + { + "epoch": 2.5045944866160608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6269 + }, + { + "epoch": 2.50499400719137, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6270 + }, + { + "epoch": 2.50539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6271 + }, + { + "epoch": 2.5057930483419897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 6272 + }, + { + "epoch": 2.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6273 + }, + { + "epoch": 2.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6274 + }, + { + "epoch": 2.5069916100679186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6275 + }, + { + "epoch": 2.507391130643228, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 6276 + }, + { + "epoch": 2.5077906512185377, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6277 + }, + { + "epoch": 2.5081901717938475, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6278 + }, + { + "epoch": 2.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6279 + }, + { + "epoch": 2.5089892129444666, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6280 + }, + { + "epoch": 2.5093887335197764, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6281 + }, + { + "epoch": 2.5097882540950858, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6282 + }, + { + "epoch": 2.5101877746703956, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6283 + }, + { + "epoch": 2.5105872952457053, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6284 + }, + { + "epoch": 2.5109868158210147, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6285 + }, + { + "epoch": 2.5113863363963245, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6286 + }, + { + "epoch": 2.511785856971634, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6287 + }, + { + "epoch": 2.5121853775469436, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6288 + }, + { + "epoch": 2.5125848981222534, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6289 + }, + { + "epoch": 2.512984418697563, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6290 + }, + { + "epoch": 2.5133839392728725, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6291 + }, + { + "epoch": 2.5137834598481823, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6292 + }, + { + "epoch": 2.5141829804234916, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6293 + }, + { + "epoch": 2.5145825009988014, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6294 + }, + { + "epoch": 2.514982021574111, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6295 + }, + { + "epoch": 2.5153815421494206, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6296 + }, + { + "epoch": 2.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6297 + }, + { + "epoch": 2.5161805833000397, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6298 + }, + { + "epoch": 2.5165801038753495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6299 + }, + { + "epoch": 2.5169796244506593, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6300 + }, + { + "epoch": 2.517379145025969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6301 + }, + { + "epoch": 2.5177786656012784, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6302 + }, + { + "epoch": 2.518178186176588, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6303 + }, + { + "epoch": 2.5185777067518975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6304 + }, + { + "epoch": 2.5189772273272073, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6305 + }, + { + "epoch": 2.519376747902517, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6306 + }, + { + "epoch": 2.519776268477827, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6307 + }, + { + "epoch": 2.520175789053136, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6308 + }, + { + "epoch": 2.520575309628446, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6309 + }, + { + "epoch": 2.5209748302037553, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6310 + }, + { + "epoch": 2.521374350779065, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 6311 + }, + { + "epoch": 2.521773871354375, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6312 + }, + { + "epoch": 2.5221733919296843, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 6313 + }, + { + "epoch": 2.522572912504994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6314 + }, + { + "epoch": 2.5229724330803034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6315 + }, + { + "epoch": 2.523371953655613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6316 + }, + { + "epoch": 2.523771474230923, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6317 + }, + { + "epoch": 2.5241709948062327, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6318 + }, + { + "epoch": 2.524570515381542, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6319 + }, + { + "epoch": 2.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6320 + }, + { + "epoch": 2.525369556532161, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6321 + }, + { + "epoch": 2.525769077107471, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6322 + }, + { + "epoch": 2.526168597682781, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 6323 + }, + { + "epoch": 2.52656811825809, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6324 + }, + { + "epoch": 2.5269676388334, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6325 + }, + { + "epoch": 2.5273671594087097, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6326 + }, + { + "epoch": 2.527766679984019, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 6327 + }, + { + "epoch": 2.528166200559329, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6328 + }, + { + "epoch": 2.5285657211346386, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6329 + }, + { + "epoch": 2.528965241709948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6330 + }, + { + "epoch": 2.5293647622852578, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6331 + }, + { + "epoch": 2.529764282860567, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6332 + }, + { + "epoch": 2.530163803435877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6333 + }, + { + "epoch": 2.5305633240111867, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 6334 + }, + { + "epoch": 2.5309628445864965, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6335 + }, + { + "epoch": 2.531362365161806, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6336 + }, + { + "epoch": 2.5317618857371156, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6337 + }, + { + "epoch": 2.532161406312425, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6338 + }, + { + "epoch": 2.5325609268877347, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6339 + }, + { + "epoch": 2.5329604474630445, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6340 + }, + { + "epoch": 2.533359968038354, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6341 + }, + { + "epoch": 2.5337594886136636, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6342 + }, + { + "epoch": 2.534159009188973, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6343 + }, + { + "epoch": 2.5345585297642828, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6344 + }, + { + "epoch": 2.5349580503395925, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6345 + }, + { + "epoch": 2.5353575709149023, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6346 + }, + { + "epoch": 2.5357570914902117, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 6347 + }, + { + "epoch": 2.5361566120655215, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6348 + }, + { + "epoch": 2.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6349 + }, + { + "epoch": 2.5369556532161406, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 6350 + }, + { + "epoch": 2.5373551737914504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6351 + }, + { + "epoch": 2.53775469436676, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6352 + }, + { + "epoch": 2.5381542149420695, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6353 + }, + { + "epoch": 2.5385537355173793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6354 + }, + { + "epoch": 2.5389532560926886, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6355 + }, + { + "epoch": 2.5393527766679984, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6356 + }, + { + "epoch": 2.539752297243308, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6357 + }, + { + "epoch": 2.5401518178186175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6358 + }, + { + "epoch": 2.5405513383939273, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 6359 + }, + { + "epoch": 2.5409508589692367, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6360 + }, + { + "epoch": 2.5413503795445465, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6361 + }, + { + "epoch": 2.5417499001198562, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 6362 + }, + { + "epoch": 2.542149420695166, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6363 + }, + { + "epoch": 2.5425489412704754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6364 + }, + { + "epoch": 2.542948461845785, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6365 + }, + { + "epoch": 2.5433479824210945, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 6366 + }, + { + "epoch": 2.5437475029964043, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6367 + }, + { + "epoch": 2.544147023571714, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6368 + }, + { + "epoch": 2.5445465441470234, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6369 + }, + { + "epoch": 2.544946064722333, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6370 + }, + { + "epoch": 2.545345585297643, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6371 + }, + { + "epoch": 2.5457451058729523, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6372 + }, + { + "epoch": 2.546144626448262, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6373 + }, + { + "epoch": 2.546544147023572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6374 + }, + { + "epoch": 2.5469436675988812, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6375 + }, + { + "epoch": 2.547343188174191, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6376 + }, + { + "epoch": 2.5477427087495004, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6377 + }, + { + "epoch": 2.54814222932481, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6378 + }, + { + "epoch": 2.54854174990012, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6379 + }, + { + "epoch": 2.5489412704754297, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6380 + }, + { + "epoch": 2.549340791050739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6381 + }, + { + "epoch": 2.549740311626049, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6382 + }, + { + "epoch": 2.550139832201358, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6383 + }, + { + "epoch": 2.550539352776668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6384 + }, + { + "epoch": 2.5509388733519778, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6385 + }, + { + "epoch": 2.551338393927287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 6386 + }, + { + "epoch": 2.551737914502597, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6387 + }, + { + "epoch": 2.5521374350779062, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 6388 + }, + { + "epoch": 2.552536955653216, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6389 + }, + { + "epoch": 2.552936476228526, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6390 + }, + { + "epoch": 2.5533359968038356, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6391 + }, + { + "epoch": 2.553735517379145, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6392 + }, + { + "epoch": 2.5541350379544547, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6393 + }, + { + "epoch": 2.554534558529764, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6394 + }, + { + "epoch": 2.554934079105074, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 6395 + }, + { + "epoch": 2.5553335996803836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6396 + }, + { + "epoch": 2.5557331202556934, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6397 + }, + { + "epoch": 2.556132640831003, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6398 + }, + { + "epoch": 2.5565321614063126, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6399 + }, + { + "epoch": 2.556931681981622, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6400 + }, + { + "epoch": 2.5573312025569317, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6401 + }, + { + "epoch": 2.5577307231322415, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6402 + }, + { + "epoch": 2.558130243707551, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6403 + }, + { + "epoch": 2.5585297642828606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6404 + }, + { + "epoch": 2.55892928485817, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 6405 + }, + { + "epoch": 2.5593288054334797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 6406 + }, + { + "epoch": 2.5597283260087895, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 6407 + }, + { + "epoch": 2.5601278465840993, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6408 + }, + { + "epoch": 2.5605273671594087, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6409 + }, + { + "epoch": 2.5609268877347184, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6410 + }, + { + "epoch": 2.561326408310028, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6411 + }, + { + "epoch": 2.5617259288853376, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 6412 + }, + { + "epoch": 2.5621254494606474, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6413 + }, + { + "epoch": 2.5625249700359567, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6414 + }, + { + "epoch": 2.5629244906112665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6415 + }, + { + "epoch": 2.5633240111865763, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6416 + }, + { + "epoch": 2.5637235317618856, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 6417 + }, + { + "epoch": 2.5641230523371954, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6418 + }, + { + "epoch": 2.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 6419 + }, + { + "epoch": 2.5649220934878145, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6420 + }, + { + "epoch": 2.5653216140631243, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6421 + }, + { + "epoch": 2.5657211346384337, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6422 + }, + { + "epoch": 2.5661206552137434, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6423 + }, + { + "epoch": 2.5665201757890532, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6424 + }, + { + "epoch": 2.566919696364363, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6425 + }, + { + "epoch": 2.5673192169396724, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6426 + }, + { + "epoch": 2.567718737514982, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6427 + }, + { + "epoch": 2.5681182580902915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6428 + }, + { + "epoch": 2.5685177786656013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6429 + }, + { + "epoch": 2.568917299240911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6430 + }, + { + "epoch": 2.5693168198162204, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6431 + }, + { + "epoch": 2.56971634039153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6432 + }, + { + "epoch": 2.5701158609668395, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6433 + }, + { + "epoch": 2.5705153815421493, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 6434 + }, + { + "epoch": 2.570914902117459, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6435 + }, + { + "epoch": 2.571314422692769, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6436 + }, + { + "epoch": 2.5717139432680782, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6437 + }, + { + "epoch": 2.572113463843388, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6438 + }, + { + "epoch": 2.5725129844186974, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6439 + }, + { + "epoch": 2.572912504994007, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6440 + }, + { + "epoch": 2.573312025569317, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6441 + }, + { + "epoch": 2.5737115461446267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6442 + }, + { + "epoch": 2.574111066719936, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6443 + }, + { + "epoch": 2.574510587295246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 6444 + }, + { + "epoch": 2.574910107870555, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6445 + }, + { + "epoch": 2.575309628445865, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6446 + }, + { + "epoch": 2.5757091490211748, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6447 + }, + { + "epoch": 2.576108669596484, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6448 + }, + { + "epoch": 2.576508190171794, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6449 + }, + { + "epoch": 2.5769077107471032, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6450 + }, + { + "epoch": 2.577307231322413, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6451 + }, + { + "epoch": 2.577706751897723, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6452 + }, + { + "epoch": 2.5781062724730326, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6453 + }, + { + "epoch": 2.578505793048342, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6454 + }, + { + "epoch": 2.5789053136236517, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6455 + }, + { + "epoch": 2.579304834198961, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6456 + }, + { + "epoch": 2.579704354774271, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 6457 + }, + { + "epoch": 2.5801038753495806, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6458 + }, + { + "epoch": 2.58050339592489, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6459 + }, + { + "epoch": 2.5809029165001998, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6460 + }, + { + "epoch": 2.5813024370755095, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6461 + }, + { + "epoch": 2.581701957650819, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6462 + }, + { + "epoch": 2.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6463 + }, + { + "epoch": 2.5825009988014385, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6464 + }, + { + "epoch": 2.582900519376748, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6465 + }, + { + "epoch": 2.5833000399520576, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6466 + }, + { + "epoch": 2.583699560527367, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6467 + }, + { + "epoch": 2.5840990811026767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6468 + }, + { + "epoch": 2.5844986016779865, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6469 + }, + { + "epoch": 2.5848981222532963, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6470 + }, + { + "epoch": 2.5852976428286056, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6471 + }, + { + "epoch": 2.5856971634039154, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6472 + }, + { + "epoch": 2.5860966839792248, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6473 + }, + { + "epoch": 2.5864962045545346, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6474 + }, + { + "epoch": 2.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 6475 + }, + { + "epoch": 2.5872952457051537, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6476 + }, + { + "epoch": 2.5876947662804635, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6477 + }, + { + "epoch": 2.588094286855773, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6478 + }, + { + "epoch": 2.5884938074310826, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6479 + }, + { + "epoch": 2.5888933280063924, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6480 + }, + { + "epoch": 2.589292848581702, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6481 + }, + { + "epoch": 2.5896923691570115, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6482 + }, + { + "epoch": 2.5900918897323213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6483 + }, + { + "epoch": 2.5904914103076306, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 6484 + }, + { + "epoch": 2.5908909308829404, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6485 + }, + { + "epoch": 2.59129045145825, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6486 + }, + { + "epoch": 2.59168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6487 + }, + { + "epoch": 2.5920894926088693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6488 + }, + { + "epoch": 2.592489013184179, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6489 + }, + { + "epoch": 2.5928885337594885, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6490 + }, + { + "epoch": 2.5932880543347983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6491 + }, + { + "epoch": 2.593687574910108, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6492 + }, + { + "epoch": 2.5940870954854174, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 6493 + }, + { + "epoch": 2.594486616060727, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6494 + }, + { + "epoch": 2.5948861366360365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6495 + }, + { + "epoch": 2.5952856572113463, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6496 + }, + { + "epoch": 2.595685177786656, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6497 + }, + { + "epoch": 2.596084698361966, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6498 + }, + { + "epoch": 2.596484218937275, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6499 + }, + { + "epoch": 2.596883739512585, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6500 + }, + { + "epoch": 2.5972832600878943, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6501 + }, + { + "epoch": 2.597682780663204, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6502 + }, + { + "epoch": 2.598082301238514, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 6503 + }, + { + "epoch": 2.5984818218138233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6504 + }, + { + "epoch": 2.598881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6505 + }, + { + "epoch": 2.599280862964443, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6506 + }, + { + "epoch": 2.599680383539752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6507 + }, + { + "epoch": 2.600079904115062, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 6508 + }, + { + "epoch": 2.6004794246903717, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6509 + }, + { + "epoch": 2.600878945265681, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6510 + }, + { + "epoch": 2.601278465840991, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6511 + }, + { + "epoch": 2.6016779864163, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6512 + }, + { + "epoch": 2.60207750699161, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6513 + }, + { + "epoch": 2.60247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6514 + }, + { + "epoch": 2.6028765481422296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6515 + }, + { + "epoch": 2.603276068717539, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6516 + }, + { + "epoch": 2.6036755892928487, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6517 + }, + { + "epoch": 2.604075109868158, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6518 + }, + { + "epoch": 2.604474630443468, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6519 + }, + { + "epoch": 2.6048741510187776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6520 + }, + { + "epoch": 2.605273671594087, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6521 + }, + { + "epoch": 2.6056731921693967, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6522 + }, + { + "epoch": 2.606072712744706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 6523 + }, + { + "epoch": 2.606472233320016, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6524 + }, + { + "epoch": 2.6068717538953257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6525 + }, + { + "epoch": 2.6072712744706354, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6526 + }, + { + "epoch": 2.607670795045945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6527 + }, + { + "epoch": 2.6080703156212546, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6528 + }, + { + "epoch": 2.608469836196564, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6529 + }, + { + "epoch": 2.6088693567718737, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 6530 + }, + { + "epoch": 2.6092688773471835, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6531 + }, + { + "epoch": 2.6096683979224933, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6532 + }, + { + "epoch": 2.6100679184978026, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6533 + }, + { + "epoch": 2.6104674390731124, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6534 + }, + { + "epoch": 2.6108669596484217, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6535 + }, + { + "epoch": 2.6112664802237315, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6536 + }, + { + "epoch": 2.6116660007990413, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6537 + }, + { + "epoch": 2.6120655213743507, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6538 + }, + { + "epoch": 2.6124650419496604, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6539 + }, + { + "epoch": 2.61286456252497, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6540 + }, + { + "epoch": 2.6132640831002796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 6541 + }, + { + "epoch": 2.6136636036755894, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6542 + }, + { + "epoch": 2.614063124250899, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6543 + }, + { + "epoch": 2.6144626448262085, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6544 + }, + { + "epoch": 2.6148621654015183, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6545 + }, + { + "epoch": 2.6152616859768276, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6546 + }, + { + "epoch": 2.6156612065521374, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6547 + }, + { + "epoch": 2.616060727127447, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 6548 + }, + { + "epoch": 2.6164602477027565, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6549 + }, + { + "epoch": 2.6168597682780663, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6550 + }, + { + "epoch": 2.617259288853376, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6551 + }, + { + "epoch": 2.6176588094286855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6552 + }, + { + "epoch": 2.6180583300039952, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6553 + }, + { + "epoch": 2.618457850579305, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6554 + }, + { + "epoch": 2.6188573711546144, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6555 + }, + { + "epoch": 2.619256891729924, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6556 + }, + { + "epoch": 2.6196564123052335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6557 + }, + { + "epoch": 2.6200559328805433, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6558 + }, + { + "epoch": 2.620455453455853, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 6559 + }, + { + "epoch": 2.620854974031163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6560 + }, + { + "epoch": 2.621254494606472, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6561 + }, + { + "epoch": 2.621654015181782, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6562 + }, + { + "epoch": 2.6220535357570913, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6563 + }, + { + "epoch": 2.622453056332401, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 6564 + }, + { + "epoch": 2.622852576907711, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6565 + }, + { + "epoch": 2.6232520974830202, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6566 + }, + { + "epoch": 2.62365161805833, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6567 + }, + { + "epoch": 2.6240511386336394, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6568 + }, + { + "epoch": 2.624450659208949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6569 + }, + { + "epoch": 2.624850179784259, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6570 + }, + { + "epoch": 2.6252497003595687, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6571 + }, + { + "epoch": 2.625649220934878, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6572 + }, + { + "epoch": 2.626048741510188, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6573 + }, + { + "epoch": 2.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6574 + }, + { + "epoch": 2.626847782660807, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6575 + }, + { + "epoch": 2.6272473032361168, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6576 + }, + { + "epoch": 2.6276468238114266, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 6577 + }, + { + "epoch": 2.628046344386736, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6578 + }, + { + "epoch": 2.6284458649620457, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6579 + }, + { + "epoch": 2.628845385537355, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6580 + }, + { + "epoch": 2.629244906112665, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6581 + }, + { + "epoch": 2.6296444266879746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6582 + }, + { + "epoch": 2.630043947263284, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6583 + }, + { + "epoch": 2.6304434678385937, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6584 + }, + { + "epoch": 2.630842988413903, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 6585 + }, + { + "epoch": 2.631242508989213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6586 + }, + { + "epoch": 2.6316420295645226, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6587 + }, + { + "epoch": 2.6320415501398324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6588 + }, + { + "epoch": 2.6324410707151418, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6589 + }, + { + "epoch": 2.6328405912904516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6590 + }, + { + "epoch": 2.633240111865761, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6591 + }, + { + "epoch": 2.6336396324410707, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6592 + }, + { + "epoch": 2.6340391530163805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6593 + }, + { + "epoch": 2.63443867359169, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6594 + }, + { + "epoch": 2.6348381941669996, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6595 + }, + { + "epoch": 2.6352377147423094, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6596 + }, + { + "epoch": 2.6356372353176187, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6597 + }, + { + "epoch": 2.6360367558929285, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6598 + }, + { + "epoch": 2.6364362764682383, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6599 + }, + { + "epoch": 2.6368357970435476, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6600 + }, + { + "epoch": 2.6372353176188574, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6601 + }, + { + "epoch": 2.6376348381941668, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6602 + }, + { + "epoch": 2.6380343587694766, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6603 + }, + { + "epoch": 2.6384338793447863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6604 + }, + { + "epoch": 2.638833399920096, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6605 + }, + { + "epoch": 2.6392329204954055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6606 + }, + { + "epoch": 2.6396324410707153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6607 + }, + { + "epoch": 2.6400319616460246, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6608 + }, + { + "epoch": 2.6404314822213344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6609 + }, + { + "epoch": 2.640831002796644, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6610 + }, + { + "epoch": 2.6412305233719535, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6611 + }, + { + "epoch": 2.6416300439472633, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6612 + }, + { + "epoch": 2.6420295645225726, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6613 + }, + { + "epoch": 2.6424290850978824, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6614 + }, + { + "epoch": 2.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6615 + }, + { + "epoch": 2.643228126248502, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6616 + }, + { + "epoch": 2.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6617 + }, + { + "epoch": 2.644027167399121, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6618 + }, + { + "epoch": 2.6444266879744305, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6619 + }, + { + "epoch": 2.6448262085497403, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6620 + }, + { + "epoch": 2.64522572912505, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6621 + }, + { + "epoch": 2.64562524970036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6622 + }, + { + "epoch": 2.646024770275669, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6623 + }, + { + "epoch": 2.646424290850979, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6624 + }, + { + "epoch": 2.6468238114262883, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6625 + }, + { + "epoch": 2.647223332001598, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6626 + }, + { + "epoch": 2.647622852576908, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6627 + }, + { + "epoch": 2.6480223731522172, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6628 + }, + { + "epoch": 2.648421893727527, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6629 + }, + { + "epoch": 2.6488214143028364, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6630 + }, + { + "epoch": 2.649220934878146, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6631 + }, + { + "epoch": 2.649620455453456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 6632 + }, + { + "epoch": 2.6500199760287657, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6633 + }, + { + "epoch": 2.650419496604075, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6634 + }, + { + "epoch": 2.650819017179385, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6635 + }, + { + "epoch": 2.651218537754694, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6636 + }, + { + "epoch": 2.651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6637 + }, + { + "epoch": 2.6520175789053138, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6638 + }, + { + "epoch": 2.652417099480623, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 6639 + }, + { + "epoch": 2.652816620055933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6640 + }, + { + "epoch": 2.6532161406312427, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6641 + }, + { + "epoch": 2.653615661206552, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6642 + }, + { + "epoch": 2.654015181781862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6643 + }, + { + "epoch": 2.6544147023571716, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6644 + }, + { + "epoch": 2.654814222932481, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6645 + }, + { + "epoch": 2.6552137435077907, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6646 + }, + { + "epoch": 2.6556132640831, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 6647 + }, + { + "epoch": 2.65601278465841, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6648 + }, + { + "epoch": 2.6564123052337196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6649 + }, + { + "epoch": 2.6568118258090294, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6650 + }, + { + "epoch": 2.6572113463843388, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6651 + }, + { + "epoch": 2.6576108669596485, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6652 + }, + { + "epoch": 2.658010387534958, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6653 + }, + { + "epoch": 2.6584099081102677, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6654 + }, + { + "epoch": 2.6588094286855775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6655 + }, + { + "epoch": 2.659208949260887, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6656 + }, + { + "epoch": 2.6596084698361966, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6657 + }, + { + "epoch": 2.660007990411506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6658 + }, + { + "epoch": 2.6604075109868157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6659 + }, + { + "epoch": 2.6608070315621255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6660 + }, + { + "epoch": 2.6612065521374353, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6661 + }, + { + "epoch": 2.6616060727127446, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6662 + }, + { + "epoch": 2.6620055932880544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6663 + }, + { + "epoch": 2.6624051138633638, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6664 + }, + { + "epoch": 2.6628046344386735, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6665 + }, + { + "epoch": 2.6632041550139833, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 6666 + }, + { + "epoch": 2.663603675589293, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6667 + }, + { + "epoch": 2.6640031961646025, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6668 + }, + { + "epoch": 2.6644027167399122, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6669 + }, + { + "epoch": 2.6648022373152216, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6670 + }, + { + "epoch": 2.6652017578905314, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6671 + }, + { + "epoch": 2.665601278465841, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6672 + }, + { + "epoch": 2.6660007990411505, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6673 + }, + { + "epoch": 2.6664003196164603, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6674 + }, + { + "epoch": 2.6667998401917696, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6675 + }, + { + "epoch": 2.6671993607670794, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6676 + }, + { + "epoch": 2.667598881342389, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6677 + }, + { + "epoch": 2.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6678 + }, + { + "epoch": 2.6683979224930083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6679 + }, + { + "epoch": 2.668797443068318, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6680 + }, + { + "epoch": 2.6691969636436275, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6681 + }, + { + "epoch": 2.6695964842189372, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6682 + }, + { + "epoch": 2.669996004794247, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6683 + }, + { + "epoch": 2.6703955253695564, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6684 + }, + { + "epoch": 2.670795045944866, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6685 + }, + { + "epoch": 2.671194566520176, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6686 + }, + { + "epoch": 2.6715940870954853, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 6687 + }, + { + "epoch": 2.671993607670795, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6688 + }, + { + "epoch": 2.672393128246105, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6689 + }, + { + "epoch": 2.672792648821414, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6690 + }, + { + "epoch": 2.673192169396724, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6691 + }, + { + "epoch": 2.6735916899720333, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6692 + }, + { + "epoch": 2.673991210547343, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6693 + }, + { + "epoch": 2.674390731122653, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6694 + }, + { + "epoch": 2.6747902516979627, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6695 + }, + { + "epoch": 2.675189772273272, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6696 + }, + { + "epoch": 2.675589292848582, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6697 + }, + { + "epoch": 2.675988813423891, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6698 + }, + { + "epoch": 2.676388333999201, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6699 + }, + { + "epoch": 2.6767878545745107, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6700 + }, + { + "epoch": 2.67718737514982, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6701 + }, + { + "epoch": 2.67758689572513, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6702 + }, + { + "epoch": 2.677986416300439, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 6703 + }, + { + "epoch": 2.678385936875749, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6704 + }, + { + "epoch": 2.678785457451059, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6705 + }, + { + "epoch": 2.6791849780263686, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6706 + }, + { + "epoch": 2.679584498601678, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6707 + }, + { + "epoch": 2.6799840191769877, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6708 + }, + { + "epoch": 2.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6709 + }, + { + "epoch": 2.680783060327607, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6710 + }, + { + "epoch": 2.6811825809029166, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6711 + }, + { + "epoch": 2.6815821014782264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6712 + }, + { + "epoch": 2.6819816220535357, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 6713 + }, + { + "epoch": 2.6823811426288455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6714 + }, + { + "epoch": 2.682780663204155, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6715 + }, + { + "epoch": 2.6831801837794647, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6716 + }, + { + "epoch": 2.6835797043547744, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6717 + }, + { + "epoch": 2.683979224930084, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6718 + }, + { + "epoch": 2.6843787455053936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6719 + }, + { + "epoch": 2.684778266080703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 6720 + }, + { + "epoch": 2.6851777866560127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6721 + }, + { + "epoch": 2.6855773072313225, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6722 + }, + { + "epoch": 2.6859768278066323, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6723 + }, + { + "epoch": 2.6863763483819416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6724 + }, + { + "epoch": 2.6867758689572514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6725 + }, + { + "epoch": 2.6871753895325607, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6726 + }, + { + "epoch": 2.6875749101078705, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6727 + }, + { + "epoch": 2.6879744306831803, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 6728 + }, + { + "epoch": 2.6883739512584897, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6729 + }, + { + "epoch": 2.6887734718337994, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6730 + }, + { + "epoch": 2.6891729924091092, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6731 + }, + { + "epoch": 2.6895725129844186, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6732 + }, + { + "epoch": 2.6899720335597284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6733 + }, + { + "epoch": 2.690371554135038, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6734 + }, + { + "epoch": 2.6907710747103475, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6735 + }, + { + "epoch": 2.6911705952856573, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6736 + }, + { + "epoch": 2.6915701158609666, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6737 + }, + { + "epoch": 2.6919696364362764, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6738 + }, + { + "epoch": 2.692369157011586, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6739 + }, + { + "epoch": 2.692768677586896, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 6740 + }, + { + "epoch": 2.6931681981622053, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6741 + }, + { + "epoch": 2.693567718737515, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6742 + }, + { + "epoch": 2.6939672393128244, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6743 + }, + { + "epoch": 2.6943667598881342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6744 + }, + { + "epoch": 2.694766280463444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6745 + }, + { + "epoch": 2.6951658010387534, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 6746 + }, + { + "epoch": 2.695565321614063, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6747 + }, + { + "epoch": 2.6959648421893725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6748 + }, + { + "epoch": 2.6963643627646823, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6749 + }, + { + "epoch": 2.696763883339992, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 6750 + }, + { + "epoch": 2.697163403915302, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6751 + }, + { + "epoch": 2.697562924490611, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6752 + }, + { + "epoch": 2.697962445065921, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6753 + }, + { + "epoch": 2.6983619656412303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6754 + }, + { + "epoch": 2.69876148621654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6755 + }, + { + "epoch": 2.69916100679185, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6756 + }, + { + "epoch": 2.6995605273671597, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6757 + }, + { + "epoch": 2.699960047942469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6758 + }, + { + "epoch": 2.700359568517779, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6759 + }, + { + "epoch": 2.700759089093088, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6760 + }, + { + "epoch": 2.701158609668398, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6761 + }, + { + "epoch": 2.7015581302437077, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6762 + }, + { + "epoch": 2.701957650819017, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6763 + }, + { + "epoch": 2.702357171394327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6764 + }, + { + "epoch": 2.702756691969636, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6765 + }, + { + "epoch": 2.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6766 + }, + { + "epoch": 2.7035557331202558, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6767 + }, + { + "epoch": 2.7039552536955656, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6768 + }, + { + "epoch": 2.704354774270875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6769 + }, + { + "epoch": 2.7047542948461847, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 6770 + }, + { + "epoch": 2.705153815421494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6771 + }, + { + "epoch": 2.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6772 + }, + { + "epoch": 2.7059528565721136, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6773 + }, + { + "epoch": 2.706352377147423, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6774 + }, + { + "epoch": 2.7067518977227327, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6775 + }, + { + "epoch": 2.7071514182980425, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6776 + }, + { + "epoch": 2.707550938873352, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6777 + }, + { + "epoch": 2.7079504594486616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6778 + }, + { + "epoch": 2.7083499800239714, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6779 + }, + { + "epoch": 2.7087495005992808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6780 + }, + { + "epoch": 2.7091490211745906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 6781 + }, + { + "epoch": 2.7095485417499, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6782 + }, + { + "epoch": 2.7099480623252097, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6783 + }, + { + "epoch": 2.7103475829005195, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 6784 + }, + { + "epoch": 2.7107471034758293, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6785 + }, + { + "epoch": 2.7111466240511386, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6786 + }, + { + "epoch": 2.7115461446264484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 6787 + }, + { + "epoch": 2.7119456652017577, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6788 + }, + { + "epoch": 2.7123451857770675, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6789 + }, + { + "epoch": 2.7127447063523773, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6790 + }, + { + "epoch": 2.7131442269276866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 6791 + }, + { + "epoch": 2.7135437475029964, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6792 + }, + { + "epoch": 2.7139432680783058, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6793 + }, + { + "epoch": 2.7143427886536156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6794 + }, + { + "epoch": 2.7147423092289253, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6795 + }, + { + "epoch": 2.715141829804235, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6796 + }, + { + "epoch": 2.7155413503795445, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6797 + }, + { + "epoch": 2.7159408709548543, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 6798 + }, + { + "epoch": 2.7163403915301636, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6799 + }, + { + "epoch": 2.7167399121054734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6800 + }, + { + "epoch": 2.717139432680783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6801 + }, + { + "epoch": 2.717538953256093, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6802 + }, + { + "epoch": 2.7179384738314023, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6803 + }, + { + "epoch": 2.718337994406712, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6804 + }, + { + "epoch": 2.7187375149820214, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6805 + }, + { + "epoch": 2.719137035557331, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6806 + }, + { + "epoch": 2.719536556132641, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6807 + }, + { + "epoch": 2.7199360767079503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6808 + }, + { + "epoch": 2.72033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6809 + }, + { + "epoch": 2.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6810 + }, + { + "epoch": 2.7211346384338793, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6811 + }, + { + "epoch": 2.721534159009189, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6812 + }, + { + "epoch": 2.721933679584499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6813 + }, + { + "epoch": 2.722333200159808, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6814 + }, + { + "epoch": 2.722732720735118, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6815 + }, + { + "epoch": 2.7231322413104273, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6816 + }, + { + "epoch": 2.723531761885737, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6817 + }, + { + "epoch": 2.723931282461047, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6818 + }, + { + "epoch": 2.724330803036356, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6819 + }, + { + "epoch": 2.724730323611666, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6820 + }, + { + "epoch": 2.725129844186976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6821 + }, + { + "epoch": 2.725529364762285, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6822 + }, + { + "epoch": 2.725928885337595, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6823 + }, + { + "epoch": 2.7263284059129047, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6824 + }, + { + "epoch": 2.726727926488214, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6825 + }, + { + "epoch": 2.727127447063524, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6826 + }, + { + "epoch": 2.727526967638833, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6827 + }, + { + "epoch": 2.727926488214143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6828 + }, + { + "epoch": 2.7283260087894528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6829 + }, + { + "epoch": 2.7287255293647625, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6830 + }, + { + "epoch": 2.729125049940072, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6831 + }, + { + "epoch": 2.7295245705153817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6832 + }, + { + "epoch": 2.729924091090691, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6833 + }, + { + "epoch": 2.730323611666001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6834 + }, + { + "epoch": 2.7307231322413106, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6835 + }, + { + "epoch": 2.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 6836 + }, + { + "epoch": 2.7315221733919297, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6837 + }, + { + "epoch": 2.731921693967239, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6838 + }, + { + "epoch": 2.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6839 + }, + { + "epoch": 2.7327207351178586, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6840 + }, + { + "epoch": 2.7331202556931684, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6841 + }, + { + "epoch": 2.7335197762684778, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6842 + }, + { + "epoch": 2.7339192968437875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6843 + }, + { + "epoch": 2.734318817419097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6844 + }, + { + "epoch": 2.7347183379944067, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6845 + }, + { + "epoch": 2.7351178585697165, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6846 + }, + { + "epoch": 2.7355173791450262, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6847 + }, + { + "epoch": 2.7359168997203356, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6848 + }, + { + "epoch": 2.7363164202956454, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6849 + }, + { + "epoch": 2.7367159408709547, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6850 + }, + { + "epoch": 2.7371154614462645, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6851 + }, + { + "epoch": 2.7375149820215743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6852 + }, + { + "epoch": 2.7379145025968836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6853 + }, + { + "epoch": 2.7383140231721934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6854 + }, + { + "epoch": 2.7387135437475028, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6855 + }, + { + "epoch": 2.7391130643228125, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6856 + }, + { + "epoch": 2.7395125848981223, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6857 + }, + { + "epoch": 2.739912105473432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6858 + }, + { + "epoch": 2.7403116260487415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6859 + }, + { + "epoch": 2.7407111466240512, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6860 + }, + { + "epoch": 2.7411106671993606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6861 + }, + { + "epoch": 2.7415101877746704, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6862 + }, + { + "epoch": 2.74190970834998, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6863 + }, + { + "epoch": 2.7423092289252895, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6864 + }, + { + "epoch": 2.7427087495005993, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6865 + }, + { + "epoch": 2.743108270075909, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6866 + }, + { + "epoch": 2.7435077906512184, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6867 + }, + { + "epoch": 2.743907311226528, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6868 + }, + { + "epoch": 2.744306831801838, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6869 + }, + { + "epoch": 2.7447063523771473, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6870 + }, + { + "epoch": 2.745105872952457, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6871 + }, + { + "epoch": 2.7455053935277665, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6872 + }, + { + "epoch": 2.7459049141030762, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6873 + }, + { + "epoch": 2.746304434678386, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6874 + }, + { + "epoch": 2.746703955253696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6875 + }, + { + "epoch": 2.747103475829005, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6876 + }, + { + "epoch": 2.747502996404315, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 6877 + }, + { + "epoch": 2.7479025169796243, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6878 + }, + { + "epoch": 2.748302037554934, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6879 + }, + { + "epoch": 2.748701558130244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6880 + }, + { + "epoch": 2.749101078705553, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6881 + }, + { + "epoch": 2.749500599280863, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6882 + }, + { + "epoch": 2.7499001198561723, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6883 + }, + { + "epoch": 2.750299640431482, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6884 + }, + { + "epoch": 2.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6885 + }, + { + "epoch": 2.7510986815821017, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6886 + }, + { + "epoch": 2.751498202157411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6887 + }, + { + "epoch": 2.751897722732721, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6888 + }, + { + "epoch": 2.75229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6889 + }, + { + "epoch": 2.75269676388334, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6890 + }, + { + "epoch": 2.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6891 + }, + { + "epoch": 2.7534958050339595, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6892 + }, + { + "epoch": 2.753895325609269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6893 + }, + { + "epoch": 2.7542948461845786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6894 + }, + { + "epoch": 2.754694366759888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6895 + }, + { + "epoch": 2.7550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6896 + }, + { + "epoch": 2.7554934079105076, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6897 + }, + { + "epoch": 2.755892928485817, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6898 + }, + { + "epoch": 2.7562924490611267, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6899 + }, + { + "epoch": 2.756691969636436, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6900 + }, + { + "epoch": 2.757091490211746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6901 + }, + { + "epoch": 2.7574910107870556, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6902 + }, + { + "epoch": 2.7578905313623654, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6903 + }, + { + "epoch": 2.7582900519376747, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6904 + }, + { + "epoch": 2.7586895725129845, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6905 + }, + { + "epoch": 2.759089093088294, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6906 + }, + { + "epoch": 2.7594886136636037, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6907 + }, + { + "epoch": 2.7598881342389134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6908 + }, + { + "epoch": 2.760287654814223, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6909 + }, + { + "epoch": 2.7606871753895326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6910 + }, + { + "epoch": 2.7610866959648424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6911 + }, + { + "epoch": 2.7614862165401517, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6912 + }, + { + "epoch": 2.7618857371154615, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6913 + }, + { + "epoch": 2.7622852576907713, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6914 + }, + { + "epoch": 2.7626847782660806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6915 + }, + { + "epoch": 2.7630842988413904, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6916 + }, + { + "epoch": 2.7634838194166997, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6917 + }, + { + "epoch": 2.7638833399920095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6918 + }, + { + "epoch": 2.7642828605673193, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6919 + }, + { + "epoch": 2.764682381142629, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6920 + }, + { + "epoch": 2.7650819017179384, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6921 + }, + { + "epoch": 2.7654814222932482, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6922 + }, + { + "epoch": 2.7658809428685576, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6923 + }, + { + "epoch": 2.7662804634438674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6924 + }, + { + "epoch": 2.766679984019177, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6925 + }, + { + "epoch": 2.7670795045944865, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6926 + }, + { + "epoch": 2.7674790251697963, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6927 + }, + { + "epoch": 2.7678785457451056, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6928 + }, + { + "epoch": 2.7682780663204154, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6929 + }, + { + "epoch": 2.768677586895725, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6930 + }, + { + "epoch": 2.769077107471035, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6931 + }, + { + "epoch": 2.7694766280463443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6932 + }, + { + "epoch": 2.769876148621654, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6933 + }, + { + "epoch": 2.7702756691969634, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6934 + }, + { + "epoch": 2.7706751897722732, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6935 + }, + { + "epoch": 2.771074710347583, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6936 + }, + { + "epoch": 2.771474230922893, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6937 + }, + { + "epoch": 2.771873751498202, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6938 + }, + { + "epoch": 2.772273272073512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6939 + }, + { + "epoch": 2.7726727926488213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6940 + }, + { + "epoch": 2.773072313224131, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6941 + }, + { + "epoch": 2.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6942 + }, + { + "epoch": 2.77387135437475, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6943 + }, + { + "epoch": 2.77427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6944 + }, + { + "epoch": 2.7746703955253693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 6945 + }, + { + "epoch": 2.775069916100679, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6946 + }, + { + "epoch": 2.775469436675989, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6947 + }, + { + "epoch": 2.7758689572512987, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 6948 + }, + { + "epoch": 2.776268477826608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6949 + }, + { + "epoch": 2.776667998401918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6950 + }, + { + "epoch": 2.777067518977227, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6951 + }, + { + "epoch": 2.777467039552537, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6952 + }, + { + "epoch": 2.7778665601278467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6953 + }, + { + "epoch": 2.778266080703156, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6954 + }, + { + "epoch": 2.778665601278466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6955 + }, + { + "epoch": 2.7790651218537756, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6956 + }, + { + "epoch": 2.779464642429085, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6957 + }, + { + "epoch": 2.7798641630043948, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6958 + }, + { + "epoch": 2.7802636835797045, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6959 + }, + { + "epoch": 2.780663204155014, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6960 + }, + { + "epoch": 2.7810627247303237, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6961 + }, + { + "epoch": 2.781462245305633, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6962 + }, + { + "epoch": 2.781861765880943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6963 + }, + { + "epoch": 2.7822612864562526, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6964 + }, + { + "epoch": 2.7826608070315624, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6965 + }, + { + "epoch": 2.7830603276068717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6966 + }, + { + "epoch": 2.7834598481821815, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6967 + }, + { + "epoch": 2.783859368757491, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6968 + }, + { + "epoch": 2.7842588893328006, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6969 + }, + { + "epoch": 2.7846584099081104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6970 + }, + { + "epoch": 2.7850579304834198, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 6971 + }, + { + "epoch": 2.7854574510587295, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6972 + }, + { + "epoch": 2.785856971634039, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6973 + }, + { + "epoch": 2.7862564922093487, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6974 + }, + { + "epoch": 2.7866560127846585, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6975 + }, + { + "epoch": 2.7870555333599683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6976 + }, + { + "epoch": 2.7874550539352776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6977 + }, + { + "epoch": 2.7878545745105874, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 6978 + }, + { + "epoch": 2.7882540950858967, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6979 + }, + { + "epoch": 2.7886536156612065, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6980 + }, + { + "epoch": 2.7890531362365163, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6981 + }, + { + "epoch": 2.789452656811826, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6982 + }, + { + "epoch": 2.7898521773871354, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6983 + }, + { + "epoch": 2.790251697962445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 6984 + }, + { + "epoch": 2.7906512185377546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6985 + }, + { + "epoch": 2.7910507391130643, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6986 + }, + { + "epoch": 2.791450259688374, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6987 + }, + { + "epoch": 2.7918497802636835, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6988 + }, + { + "epoch": 2.7922493008389933, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6989 + }, + { + "epoch": 2.7926488214143026, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6990 + }, + { + "epoch": 2.7930483419896124, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6991 + }, + { + "epoch": 2.793447862564922, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6992 + }, + { + "epoch": 2.793847383140232, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6993 + }, + { + "epoch": 2.7942469037155413, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6994 + }, + { + "epoch": 2.794646424290851, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6995 + }, + { + "epoch": 2.7950459448661604, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 6996 + }, + { + "epoch": 2.79544546544147, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6997 + }, + { + "epoch": 2.79584498601678, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6998 + }, + { + "epoch": 2.7962445065920893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6999 + }, + { + "epoch": 2.796644027167399, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7000 + }, + { + "epoch": 2.797043547742709, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7001 + }, + { + "epoch": 2.7974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7002 + }, + { + "epoch": 2.797842588893328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7003 + }, + { + "epoch": 2.798242109468638, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7004 + }, + { + "epoch": 2.798641630043947, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7005 + }, + { + "epoch": 2.799041150619257, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7006 + }, + { + "epoch": 2.7994406711945663, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7007 + }, + { + "epoch": 2.799840191769876, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7008 + }, + { + "epoch": 2.800239712345186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 7009 + }, + { + "epoch": 2.8006392329204957, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7010 + }, + { + "epoch": 2.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7011 + }, + { + "epoch": 2.801438274071115, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7012 + }, + { + "epoch": 2.801837794646424, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7013 + }, + { + "epoch": 2.802237315221734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7014 + }, + { + "epoch": 2.8026368357970437, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7015 + }, + { + "epoch": 2.803036356372353, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7016 + }, + { + "epoch": 2.803435876947663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7017 + }, + { + "epoch": 2.803835397522972, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7018 + }, + { + "epoch": 2.804234918098282, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7019 + }, + { + "epoch": 2.8046344386735917, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7020 + }, + { + "epoch": 2.8050339592489015, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7021 + }, + { + "epoch": 2.805433479824211, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7022 + }, + { + "epoch": 2.8058330003995207, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7023 + }, + { + "epoch": 2.80623252097483, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 7024 + }, + { + "epoch": 2.80663204155014, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7025 + }, + { + "epoch": 2.8070315621254496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7026 + }, + { + "epoch": 2.8074310827007594, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7027 + }, + { + "epoch": 2.8078306032760687, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7028 + }, + { + "epoch": 2.8082301238513785, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7029 + }, + { + "epoch": 2.808629644426688, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7030 + }, + { + "epoch": 2.8090291650019976, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7031 + }, + { + "epoch": 2.8094286855773074, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7032 + }, + { + "epoch": 2.8098282061526167, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7033 + }, + { + "epoch": 2.8102277267279265, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7034 + }, + { + "epoch": 2.810627247303236, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7035 + }, + { + "epoch": 2.8110267678785457, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7036 + }, + { + "epoch": 2.8114262884538554, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7037 + }, + { + "epoch": 2.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7038 + }, + { + "epoch": 2.8122253296044746, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7039 + }, + { + "epoch": 2.8126248501797844, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7040 + }, + { + "epoch": 2.8130243707550937, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7041 + }, + { + "epoch": 2.8134238913304035, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7042 + }, + { + "epoch": 2.8138234119057133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7043 + }, + { + "epoch": 2.8142229324810226, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7044 + }, + { + "epoch": 2.8146224530563324, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7045 + }, + { + "epoch": 2.815021973631642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 7046 + }, + { + "epoch": 2.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7047 + }, + { + "epoch": 2.8158210147822613, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 7048 + }, + { + "epoch": 2.816220535357571, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7049 + }, + { + "epoch": 2.8166200559328805, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7050 + }, + { + "epoch": 2.8170195765081902, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7051 + }, + { + "epoch": 2.8174190970834996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7052 + }, + { + "epoch": 2.8178186176588094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 7053 + }, + { + "epoch": 2.818218138234119, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7054 + }, + { + "epoch": 2.818617658809429, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7055 + }, + { + "epoch": 2.8190171793847383, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7056 + }, + { + "epoch": 2.819416699960048, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7057 + }, + { + "epoch": 2.8198162205353574, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7058 + }, + { + "epoch": 2.820215741110667, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7059 + }, + { + "epoch": 2.820615261685977, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7060 + }, + { + "epoch": 2.8210147822612863, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7061 + }, + { + "epoch": 2.821414302836596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7062 + }, + { + "epoch": 2.8218138234119055, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7063 + }, + { + "epoch": 2.8222133439872152, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7064 + }, + { + "epoch": 2.822612864562525, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7065 + }, + { + "epoch": 2.823012385137835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7066 + }, + { + "epoch": 2.823411905713144, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7067 + }, + { + "epoch": 2.823811426288454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7068 + }, + { + "epoch": 2.8242109468637633, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7069 + }, + { + "epoch": 2.824610467439073, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7070 + }, + { + "epoch": 2.825009988014383, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7071 + }, + { + "epoch": 2.8254095085896926, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7072 + }, + { + "epoch": 2.825809029165002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7073 + }, + { + "epoch": 2.8262085497403118, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 7074 + }, + { + "epoch": 2.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7075 + }, + { + "epoch": 2.827007590890931, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 7076 + }, + { + "epoch": 2.8274071114662407, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7077 + }, + { + "epoch": 2.82780663204155, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 7078 + }, + { + "epoch": 2.82820615261686, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7079 + }, + { + "epoch": 2.828605673192169, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7080 + }, + { + "epoch": 2.829005193767479, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7081 + }, + { + "epoch": 2.8294047143427887, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7082 + }, + { + "epoch": 2.8298042349180985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7083 + }, + { + "epoch": 2.830203755493408, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7084 + }, + { + "epoch": 2.8306032760687176, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 7085 + }, + { + "epoch": 2.831002796644027, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7086 + }, + { + "epoch": 2.8314023172193368, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7087 + }, + { + "epoch": 2.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7088 + }, + { + "epoch": 2.832201358369956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7089 + }, + { + "epoch": 2.8326008789452657, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7090 + }, + { + "epoch": 2.8330003995205755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 7091 + }, + { + "epoch": 2.833399920095885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7092 + }, + { + "epoch": 2.8337994406711946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7093 + }, + { + "epoch": 2.8341989612465044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7094 + }, + { + "epoch": 2.8345984818218137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 7095 + }, + { + "epoch": 2.8349980023971235, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7096 + }, + { + "epoch": 2.835397522972433, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7097 + }, + { + "epoch": 2.8357970435477426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7098 + }, + { + "epoch": 2.8361965641230524, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7099 + }, + { + "epoch": 2.836596084698362, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7100 + }, + { + "epoch": 2.8369956052736716, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7101 + }, + { + "epoch": 2.8373951258489813, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7102 + }, + { + "epoch": 2.8377946464242907, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7103 + }, + { + "epoch": 2.8381941669996005, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7104 + }, + { + "epoch": 2.8385936875749103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7105 + }, + { + "epoch": 2.8389932081502196, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7106 + }, + { + "epoch": 2.8393927287255294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7107 + }, + { + "epoch": 2.8397922493008387, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7108 + }, + { + "epoch": 2.8401917698761485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7109 + }, + { + "epoch": 2.8405912904514583, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7110 + }, + { + "epoch": 2.840990811026768, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7111 + }, + { + "epoch": 2.8413903316020774, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7112 + }, + { + "epoch": 2.841789852177387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7113 + }, + { + "epoch": 2.8421893727526966, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7114 + }, + { + "epoch": 2.8425888933280063, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7115 + }, + { + "epoch": 2.842988413903316, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 7116 + }, + { + "epoch": 2.843387934478626, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7117 + }, + { + "epoch": 2.8437874550539353, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7118 + }, + { + "epoch": 2.844186975629245, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7119 + }, + { + "epoch": 2.8445864962045544, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7120 + }, + { + "epoch": 2.844986016779864, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7121 + }, + { + "epoch": 2.845385537355174, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7122 + }, + { + "epoch": 2.8457850579304833, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7123 + }, + { + "epoch": 2.846184578505793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 7124 + }, + { + "epoch": 2.8465840990811024, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7125 + }, + { + "epoch": 2.8469836196564122, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7126 + }, + { + "epoch": 2.847383140231722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7127 + }, + { + "epoch": 2.847782660807032, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7128 + }, + { + "epoch": 2.848182181382341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7129 + }, + { + "epoch": 2.848581701957651, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7130 + }, + { + "epoch": 2.8489812225329603, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7131 + }, + { + "epoch": 2.84938074310827, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7132 + }, + { + "epoch": 2.84978026368358, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7133 + }, + { + "epoch": 2.850179784258889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7134 + }, + { + "epoch": 2.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7135 + }, + { + "epoch": 2.8509788254095088, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7136 + }, + { + "epoch": 2.851378345984818, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 7137 + }, + { + "epoch": 2.851777866560128, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7138 + }, + { + "epoch": 2.8521773871354377, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7139 + }, + { + "epoch": 2.852576907710747, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7140 + }, + { + "epoch": 2.852976428286057, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7141 + }, + { + "epoch": 2.853375948861366, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7142 + }, + { + "epoch": 2.853775469436676, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 7143 + }, + { + "epoch": 2.8541749900119857, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7144 + }, + { + "epoch": 2.8545745105872955, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7145 + }, + { + "epoch": 2.854974031162605, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7146 + }, + { + "epoch": 2.8553735517379146, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7147 + }, + { + "epoch": 2.855773072313224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7148 + }, + { + "epoch": 2.8561725928885338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7149 + }, + { + "epoch": 2.8565721134638435, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7150 + }, + { + "epoch": 2.856971634039153, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7151 + }, + { + "epoch": 2.8573711546144627, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7152 + }, + { + "epoch": 2.857770675189772, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7153 + }, + { + "epoch": 2.858170195765082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7154 + }, + { + "epoch": 2.8585697163403916, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7155 + }, + { + "epoch": 2.8589692369157014, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7156 + }, + { + "epoch": 2.8593687574910107, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7157 + }, + { + "epoch": 2.8597682780663205, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7158 + }, + { + "epoch": 2.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7159 + }, + { + "epoch": 2.8605673192169396, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7160 + }, + { + "epoch": 2.8609668397922494, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7161 + }, + { + "epoch": 2.861366360367559, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7162 + }, + { + "epoch": 2.8617658809428685, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7163 + }, + { + "epoch": 2.8621654015181783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7164 + }, + { + "epoch": 2.8625649220934877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 7165 + }, + { + "epoch": 2.8629644426687975, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7166 + }, + { + "epoch": 2.8633639632441072, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7167 + }, + { + "epoch": 2.8637634838194166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7168 + }, + { + "epoch": 2.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7169 + }, + { + "epoch": 2.8645625249700357, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7170 + }, + { + "epoch": 2.8649620455453455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7171 + }, + { + "epoch": 2.8653615661206553, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 7172 + }, + { + "epoch": 2.865761086695965, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7173 + }, + { + "epoch": 2.8661606072712744, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7174 + }, + { + "epoch": 2.866560127846584, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7175 + }, + { + "epoch": 2.8669596484218935, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7176 + }, + { + "epoch": 2.8673591689972033, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7177 + }, + { + "epoch": 2.867758689572513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7178 + }, + { + "epoch": 2.8681582101478225, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7179 + }, + { + "epoch": 2.8685577307231322, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7180 + }, + { + "epoch": 2.868957251298442, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7181 + }, + { + "epoch": 2.8693567718737514, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7182 + }, + { + "epoch": 2.869756292449061, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7183 + }, + { + "epoch": 2.870155813024371, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7184 + }, + { + "epoch": 2.8705553335996803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7185 + }, + { + "epoch": 2.87095485417499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7186 + }, + { + "epoch": 2.8713543747502994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7187 + }, + { + "epoch": 2.871753895325609, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7188 + }, + { + "epoch": 2.872153415900919, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7189 + }, + { + "epoch": 2.872552936476229, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7190 + }, + { + "epoch": 2.872952457051538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7191 + }, + { + "epoch": 2.873351977626848, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7192 + }, + { + "epoch": 2.8737514982021573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7193 + }, + { + "epoch": 2.874151018777467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7194 + }, + { + "epoch": 2.874550539352777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7195 + }, + { + "epoch": 2.874950059928086, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 7196 + }, + { + "epoch": 2.875349580503396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7197 + }, + { + "epoch": 2.8757491010787053, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7198 + }, + { + "epoch": 2.876148621654015, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7199 + }, + { + "epoch": 2.876548142229325, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 7200 + }, + { + "epoch": 2.8769476628046347, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7201 + }, + { + "epoch": 2.877347183379944, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7202 + }, + { + "epoch": 2.877746703955254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7203 + }, + { + "epoch": 2.878146224530563, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7204 + }, + { + "epoch": 2.878545745105873, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7205 + }, + { + "epoch": 2.8789452656811827, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 7206 + }, + { + "epoch": 2.8793447862564925, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7207 + }, + { + "epoch": 2.879744306831802, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7208 + }, + { + "epoch": 2.8801438274071116, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7209 + }, + { + "epoch": 2.880543347982421, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 7210 + }, + { + "epoch": 2.8809428685577307, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7211 + }, + { + "epoch": 2.8813423891330405, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7212 + }, + { + "epoch": 2.88174190970835, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7213 + }, + { + "epoch": 2.8821414302836597, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7214 + }, + { + "epoch": 2.882540950858969, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7215 + }, + { + "epoch": 2.882940471434279, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7216 + }, + { + "epoch": 2.8833399920095886, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7217 + }, + { + "epoch": 2.8837395125848984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 7218 + }, + { + "epoch": 2.8841390331602077, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7219 + }, + { + "epoch": 2.8845385537355175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7220 + }, + { + "epoch": 2.884938074310827, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7221 + }, + { + "epoch": 2.8853375948861366, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7222 + }, + { + "epoch": 2.8857371154614464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7223 + }, + { + "epoch": 2.8861366360367557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7224 + }, + { + "epoch": 2.8865361566120655, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7225 + }, + { + "epoch": 2.8869356771873753, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7226 + }, + { + "epoch": 2.8873351977626847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7227 + }, + { + "epoch": 2.8877347183379944, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7228 + }, + { + "epoch": 2.8881342389133042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7229 + }, + { + "epoch": 2.8885337594886136, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7230 + }, + { + "epoch": 2.8889332800639234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7231 + }, + { + "epoch": 2.8893328006392327, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7232 + }, + { + "epoch": 2.8897323212145425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 7233 + }, + { + "epoch": 2.8901318417898523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7234 + }, + { + "epoch": 2.890531362365162, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7235 + }, + { + "epoch": 2.8909308829404714, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7236 + }, + { + "epoch": 2.891330403515781, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7237 + }, + { + "epoch": 2.8917299240910905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7238 + }, + { + "epoch": 2.8921294446664003, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7239 + }, + { + "epoch": 2.89252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7240 + }, + { + "epoch": 2.8929284858170194, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7241 + }, + { + "epoch": 2.8933280063923292, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7242 + }, + { + "epoch": 2.8937275269676386, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7243 + }, + { + "epoch": 2.8941270475429484, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7244 + }, + { + "epoch": 2.894526568118258, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 7245 + }, + { + "epoch": 2.894926088693568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7246 + }, + { + "epoch": 2.8953256092688773, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7247 + }, + { + "epoch": 2.895725129844187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7248 + }, + { + "epoch": 2.8961246504194964, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7249 + }, + { + "epoch": 2.896524170994806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 7250 + }, + { + "epoch": 2.896923691570116, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 7251 + }, + { + "epoch": 2.8973232121454258, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7252 + }, + { + "epoch": 2.897722732720735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7253 + }, + { + "epoch": 2.898122253296045, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7254 + }, + { + "epoch": 2.8985217738713542, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7255 + }, + { + "epoch": 2.898921294446664, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 7256 + }, + { + "epoch": 2.899320815021974, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7257 + }, + { + "epoch": 2.899720335597283, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7258 + }, + { + "epoch": 2.900119856172593, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7259 + }, + { + "epoch": 2.9005193767479023, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7260 + }, + { + "epoch": 2.900918897323212, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7261 + }, + { + "epoch": 2.901318417898522, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7262 + }, + { + "epoch": 2.9017179384738316, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7263 + }, + { + "epoch": 2.902117459049141, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7264 + }, + { + "epoch": 2.9025169796244508, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7265 + }, + { + "epoch": 2.90291650019976, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 7266 + }, + { + "epoch": 2.90331602077507, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7267 + }, + { + "epoch": 2.9037155413503797, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7268 + }, + { + "epoch": 2.904115061925689, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7269 + }, + { + "epoch": 2.904514582500999, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7270 + }, + { + "epoch": 2.9049141030763086, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7271 + }, + { + "epoch": 2.905313623651618, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7272 + }, + { + "epoch": 2.9057131442269277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7273 + }, + { + "epoch": 2.9061126648022375, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7274 + }, + { + "epoch": 2.906512185377547, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7275 + }, + { + "epoch": 2.9069117059528566, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7276 + }, + { + "epoch": 2.907311226528166, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7277 + }, + { + "epoch": 2.9077107471034758, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7278 + }, + { + "epoch": 2.9081102676787856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7279 + }, + { + "epoch": 2.9085097882540953, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7280 + }, + { + "epoch": 2.9089093088294047, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 7281 + }, + { + "epoch": 2.9093088294047145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7282 + }, + { + "epoch": 2.909708349980024, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7283 + }, + { + "epoch": 2.9101078705553336, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7284 + }, + { + "epoch": 2.9105073911306434, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7285 + }, + { + "epoch": 2.9109069117059527, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7286 + }, + { + "epoch": 2.9113064322812625, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7287 + }, + { + "epoch": 2.911705952856572, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 7288 + }, + { + "epoch": 2.9121054734318816, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7289 + }, + { + "epoch": 2.9125049940071914, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7290 + }, + { + "epoch": 2.912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7291 + }, + { + "epoch": 2.9133040351578106, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7292 + }, + { + "epoch": 2.9137035557331203, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7293 + }, + { + "epoch": 2.9141030763084297, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7294 + }, + { + "epoch": 2.9145025968837395, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7295 + }, + { + "epoch": 2.9149021174590493, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7296 + }, + { + "epoch": 2.915301638034359, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7297 + }, + { + "epoch": 2.9157011586096684, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7298 + }, + { + "epoch": 2.916100679184978, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7299 + }, + { + "epoch": 2.9165001997602875, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7300 + }, + { + "epoch": 2.9168997203355973, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7301 + }, + { + "epoch": 2.917299240910907, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7302 + }, + { + "epoch": 2.9176987614862164, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7303 + }, + { + "epoch": 2.918098282061526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7304 + }, + { + "epoch": 2.9184978026368356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7305 + }, + { + "epoch": 2.9188973232121453, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7306 + }, + { + "epoch": 2.919296843787455, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7307 + }, + { + "epoch": 2.919696364362765, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7308 + }, + { + "epoch": 2.9200958849380743, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7309 + }, + { + "epoch": 2.920495405513384, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7310 + }, + { + "epoch": 2.9208949260886934, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 7311 + }, + { + "epoch": 2.921294446664003, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7312 + }, + { + "epoch": 2.921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7313 + }, + { + "epoch": 2.9220934878146223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 7314 + }, + { + "epoch": 2.922493008389932, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7315 + }, + { + "epoch": 2.922892528965242, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7316 + }, + { + "epoch": 2.923292049540551, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7317 + }, + { + "epoch": 2.923691570115861, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7318 + }, + { + "epoch": 2.924091090691171, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7319 + }, + { + "epoch": 2.92449061126648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7320 + }, + { + "epoch": 2.92489013184179, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7321 + }, + { + "epoch": 2.9252896524170993, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7322 + }, + { + "epoch": 2.925689172992409, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7323 + }, + { + "epoch": 2.926088693567719, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7324 + }, + { + "epoch": 2.9264882141430286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7325 + }, + { + "epoch": 2.926887734718338, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7326 + }, + { + "epoch": 2.9272872552936477, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7327 + }, + { + "epoch": 2.927686775868957, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7328 + }, + { + "epoch": 2.928086296444267, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7329 + }, + { + "epoch": 2.9284858170195767, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7330 + }, + { + "epoch": 2.928885337594886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 7331 + }, + { + "epoch": 2.929284858170196, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7332 + }, + { + "epoch": 2.929684378745505, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7333 + }, + { + "epoch": 2.930083899320815, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7334 + }, + { + "epoch": 2.9304834198961247, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7335 + }, + { + "epoch": 2.9308829404714345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7336 + }, + { + "epoch": 2.931282461046744, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 7337 + }, + { + "epoch": 2.9316819816220536, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7338 + }, + { + "epoch": 2.932081502197363, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7339 + }, + { + "epoch": 2.9324810227726728, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7340 + }, + { + "epoch": 2.9328805433479825, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7341 + }, + { + "epoch": 2.9332800639232923, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7342 + }, + { + "epoch": 2.9336795844986017, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7343 + }, + { + "epoch": 2.9340791050739115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7344 + }, + { + "epoch": 2.934478625649221, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7345 + }, + { + "epoch": 2.9348781462245306, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7346 + }, + { + "epoch": 2.9352776667998404, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7347 + }, + { + "epoch": 2.9356771873751497, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7348 + }, + { + "epoch": 2.9360767079504595, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7349 + }, + { + "epoch": 2.936476228525769, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7350 + }, + { + "epoch": 2.9368757491010786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7351 + }, + { + "epoch": 2.9372752696763884, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7352 + }, + { + "epoch": 2.937674790251698, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7353 + }, + { + "epoch": 2.9380743108270075, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7354 + }, + { + "epoch": 2.9384738314023173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7355 + }, + { + "epoch": 2.9388733519776267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7356 + }, + { + "epoch": 2.9392728725529365, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7357 + }, + { + "epoch": 2.9396723931282462, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7358 + }, + { + "epoch": 2.9400719137035556, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7359 + }, + { + "epoch": 2.9404714342788654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7360 + }, + { + "epoch": 2.9408709548541747, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7361 + }, + { + "epoch": 2.9412704754294845, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7362 + }, + { + "epoch": 2.9416699960047943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7363 + }, + { + "epoch": 2.942069516580104, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7364 + }, + { + "epoch": 2.9424690371554134, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7365 + }, + { + "epoch": 2.942868557730723, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7366 + }, + { + "epoch": 2.9432680783060325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7367 + }, + { + "epoch": 2.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7368 + }, + { + "epoch": 2.944067119456652, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7369 + }, + { + "epoch": 2.944466640031962, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7370 + }, + { + "epoch": 2.9448661606072712, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7371 + }, + { + "epoch": 2.945265681182581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7372 + }, + { + "epoch": 2.9456652017578904, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7373 + }, + { + "epoch": 2.9460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7374 + }, + { + "epoch": 2.94646424290851, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7375 + }, + { + "epoch": 2.9468637634838193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7376 + }, + { + "epoch": 2.947263284059129, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7377 + }, + { + "epoch": 2.9476628046344384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7378 + }, + { + "epoch": 2.948062325209748, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7379 + }, + { + "epoch": 2.948461845785058, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7380 + }, + { + "epoch": 2.9488613663603678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7381 + }, + { + "epoch": 2.949260886935677, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7382 + }, + { + "epoch": 2.949660407510987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 7383 + }, + { + "epoch": 2.9500599280862962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7384 + }, + { + "epoch": 2.950459448661606, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7385 + }, + { + "epoch": 2.950858969236916, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7386 + }, + { + "epoch": 2.9512584898122256, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7387 + }, + { + "epoch": 2.951658010387535, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7388 + }, + { + "epoch": 2.9520575309628447, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7389 + }, + { + "epoch": 2.952457051538154, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7390 + }, + { + "epoch": 2.952856572113464, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7391 + }, + { + "epoch": 2.9532560926887736, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7392 + }, + { + "epoch": 2.953655613264083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7393 + }, + { + "epoch": 2.9540551338393928, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7394 + }, + { + "epoch": 2.954454654414702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7395 + }, + { + "epoch": 2.954854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7396 + }, + { + "epoch": 2.9552536955653217, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7397 + }, + { + "epoch": 2.9556532161406315, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7398 + }, + { + "epoch": 2.956052736715941, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7399 + }, + { + "epoch": 2.9564522572912506, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7400 + }, + { + "epoch": 2.95685177786656, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7401 + }, + { + "epoch": 2.9572512984418697, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7402 + }, + { + "epoch": 2.9576508190171795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7403 + }, + { + "epoch": 2.958050339592489, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7404 + }, + { + "epoch": 2.9584498601677987, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7405 + }, + { + "epoch": 2.958849380743108, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7406 + }, + { + "epoch": 2.959248901318418, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 7407 + }, + { + "epoch": 2.9596484218937276, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7408 + }, + { + "epoch": 2.9600479424690374, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7409 + }, + { + "epoch": 2.9604474630443467, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7410 + }, + { + "epoch": 2.9608469836196565, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 7411 + }, + { + "epoch": 2.961246504194966, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7412 + }, + { + "epoch": 2.9616460247702756, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7413 + }, + { + "epoch": 2.9620455453455854, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7414 + }, + { + "epoch": 2.962445065920895, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7415 + }, + { + "epoch": 2.9628445864962045, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7416 + }, + { + "epoch": 2.9632441070715143, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7417 + }, + { + "epoch": 2.9636436276468237, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7418 + }, + { + "epoch": 2.9640431482221334, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7419 + }, + { + "epoch": 2.9644426687974432, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7420 + }, + { + "epoch": 2.9648421893727526, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7421 + }, + { + "epoch": 2.9652417099480624, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7422 + }, + { + "epoch": 2.9656412305233717, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7423 + }, + { + "epoch": 2.9660407510986815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7424 + }, + { + "epoch": 2.9664402716739913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7425 + }, + { + "epoch": 2.966839792249301, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7426 + }, + { + "epoch": 2.9672393128246104, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7427 + }, + { + "epoch": 2.96763883339992, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7428 + }, + { + "epoch": 2.9680383539752295, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7429 + }, + { + "epoch": 2.9684378745505393, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7430 + }, + { + "epoch": 2.968837395125849, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7431 + }, + { + "epoch": 2.969236915701159, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7432 + }, + { + "epoch": 2.9696364362764682, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7433 + }, + { + "epoch": 2.970035956851778, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7434 + }, + { + "epoch": 2.9704354774270874, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7435 + }, + { + "epoch": 2.970834998002397, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7436 + }, + { + "epoch": 2.971234518577707, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 7437 + }, + { + "epoch": 2.9716340391530163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7438 + }, + { + "epoch": 2.972033559728326, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7439 + }, + { + "epoch": 2.9724330803036354, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7440 + }, + { + "epoch": 2.972832600878945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7441 + }, + { + "epoch": 2.973232121454255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7442 + }, + { + "epoch": 2.9736316420295648, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7443 + }, + { + "epoch": 2.974031162604874, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7444 + }, + { + "epoch": 2.974430683180184, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7445 + }, + { + "epoch": 2.9748302037554932, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7446 + }, + { + "epoch": 2.975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7447 + }, + { + "epoch": 2.975629244906113, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7448 + }, + { + "epoch": 2.976028765481422, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7449 + }, + { + "epoch": 2.976428286056732, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7450 + }, + { + "epoch": 2.9768278066320413, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7451 + }, + { + "epoch": 2.977227327207351, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7452 + }, + { + "epoch": 2.977626847782661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7453 + }, + { + "epoch": 2.9780263683579706, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7454 + }, + { + "epoch": 2.97842588893328, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7455 + }, + { + "epoch": 2.9788254095085898, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7456 + }, + { + "epoch": 2.979224930083899, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7457 + }, + { + "epoch": 2.979624450659209, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7458 + }, + { + "epoch": 2.9800239712345187, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7459 + }, + { + "epoch": 2.9804234918098285, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7460 + }, + { + "epoch": 2.980823012385138, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7461 + }, + { + "epoch": 2.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7462 + }, + { + "epoch": 2.981622053535757, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7463 + }, + { + "epoch": 2.9820215741110667, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7464 + }, + { + "epoch": 2.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7465 + }, + { + "epoch": 2.982820615261686, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7466 + }, + { + "epoch": 2.9832201358369956, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7467 + }, + { + "epoch": 2.983619656412305, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7468 + }, + { + "epoch": 2.9840191769876148, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7469 + }, + { + "epoch": 2.9844186975629245, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7470 + }, + { + "epoch": 2.9848182181382343, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7471 + }, + { + "epoch": 2.9852177387135437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7472 + }, + { + "epoch": 2.9856172592888535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7473 + }, + { + "epoch": 2.986016779864163, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7474 + }, + { + "epoch": 2.9864163004394726, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7475 + }, + { + "epoch": 2.9868158210147824, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7476 + }, + { + "epoch": 2.987215341590092, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7477 + }, + { + "epoch": 2.9876148621654015, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7478 + }, + { + "epoch": 2.9880143827407113, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7479 + }, + { + "epoch": 2.9884139033160206, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 7480 + }, + { + "epoch": 2.9888134238913304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7481 + }, + { + "epoch": 2.98921294446664, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7482 + }, + { + "epoch": 2.9896124650419496, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7483 + }, + { + "epoch": 2.9900119856172593, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7484 + }, + { + "epoch": 2.9904115061925687, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7485 + }, + { + "epoch": 2.9908110267678785, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 7486 + }, + { + "epoch": 2.9912105473431883, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7487 + }, + { + "epoch": 2.991610067918498, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7488 + }, + { + "epoch": 2.9920095884938074, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7489 + }, + { + "epoch": 2.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7490 + }, + { + "epoch": 2.9928086296444265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7491 + }, + { + "epoch": 2.9932081502197363, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7492 + }, + { + "epoch": 2.993607670795046, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7493 + }, + { + "epoch": 2.9940071913703554, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7494 + }, + { + "epoch": 2.994406711945665, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7495 + }, + { + "epoch": 2.9948062325209746, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7496 + }, + { + "epoch": 2.9952057530962843, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7497 + }, + { + "epoch": 2.995605273671594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7498 + }, + { + "epoch": 2.996004794246904, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7499 + }, + { + "epoch": 2.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7500 + }, + { + "epoch": 2.996803835397523, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7501 + }, + { + "epoch": 2.9972033559728324, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7502 + }, + { + "epoch": 2.997602876548142, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7503 + }, + { + "epoch": 2.998002397123452, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7504 + }, + { + "epoch": 2.9984019176987617, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7505 + }, + { + "epoch": 2.998801438274071, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7506 + }, + { + "epoch": 2.999200958849381, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7507 + }, + { + "epoch": 2.99960047942469, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7508 + }, + { + "epoch": 3.0, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7509 + }, + { + "epoch": 3.00039952057531, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7510 + }, + { + "epoch": 3.000799041150619, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7511 + }, + { + "epoch": 3.001198561725929, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7512 + }, + { + "epoch": 3.0015980823012387, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7513 + }, + { + "epoch": 3.001997602876548, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7514 + }, + { + "epoch": 3.002397123451858, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7515 + }, + { + "epoch": 3.0027966440271676, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7516 + }, + { + "epoch": 3.003196164602477, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7517 + }, + { + "epoch": 3.0035956851777867, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7518 + }, + { + "epoch": 3.003995205753096, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7519 + }, + { + "epoch": 3.004394726328406, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7520 + }, + { + "epoch": 3.0047942469037157, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7521 + }, + { + "epoch": 3.005193767479025, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7522 + }, + { + "epoch": 3.005593288054335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7523 + }, + { + "epoch": 3.0059928086296446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7524 + }, + { + "epoch": 3.006392329204954, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 7525 + }, + { + "epoch": 3.0067918497802637, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7526 + }, + { + "epoch": 3.0071913703555735, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7527 + }, + { + "epoch": 3.007590890930883, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7528 + }, + { + "epoch": 3.0079904115061926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7529 + }, + { + "epoch": 3.0083899320815024, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7530 + }, + { + "epoch": 3.0087894526568117, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7531 + }, + { + "epoch": 3.0091889732321215, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7532 + }, + { + "epoch": 3.009588493807431, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7533 + }, + { + "epoch": 3.0099880143827407, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7534 + }, + { + "epoch": 3.0103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7535 + }, + { + "epoch": 3.01078705553336, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7536 + }, + { + "epoch": 3.0111865761086696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7537 + }, + { + "epoch": 3.0115860966839794, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7538 + }, + { + "epoch": 3.0119856172592887, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7539 + }, + { + "epoch": 3.0123851378345985, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7540 + }, + { + "epoch": 3.0127846584099083, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7541 + }, + { + "epoch": 3.0131841789852176, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7542 + }, + { + "epoch": 3.0135836995605274, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7543 + }, + { + "epoch": 3.013983220135837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7544 + }, + { + "epoch": 3.0143827407111465, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7545 + }, + { + "epoch": 3.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 7546 + }, + { + "epoch": 3.0151817818617657, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 7547 + }, + { + "epoch": 3.0155813024370755, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7548 + }, + { + "epoch": 3.0159808230123852, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7549 + }, + { + "epoch": 3.0163803435876946, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7550 + }, + { + "epoch": 3.0167798641630044, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7551 + }, + { + "epoch": 3.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7552 + }, + { + "epoch": 3.0175789053136235, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7553 + }, + { + "epoch": 3.0179784258889333, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7554 + }, + { + "epoch": 3.018377946464243, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7555 + }, + { + "epoch": 3.0187774670395524, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7556 + }, + { + "epoch": 3.019176987614862, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7557 + }, + { + "epoch": 3.019576508190172, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7558 + }, + { + "epoch": 3.0199760287654813, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7559 + }, + { + "epoch": 3.020375549340791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7560 + }, + { + "epoch": 3.020775069916101, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7561 + }, + { + "epoch": 3.0211745904914102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7562 + }, + { + "epoch": 3.02157411106672, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7563 + }, + { + "epoch": 3.0219736316420294, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7564 + }, + { + "epoch": 3.022373152217339, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7565 + }, + { + "epoch": 3.022772672792649, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7566 + }, + { + "epoch": 3.0231721933679583, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7567 + }, + { + "epoch": 3.023571713943268, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7568 + }, + { + "epoch": 3.023971234518578, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7569 + }, + { + "epoch": 3.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7570 + }, + { + "epoch": 3.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7571 + }, + { + "epoch": 3.0251697962445068, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7572 + }, + { + "epoch": 3.025569316819816, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7573 + }, + { + "epoch": 3.025968837395126, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7574 + }, + { + "epoch": 3.0263683579704357, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7575 + }, + { + "epoch": 3.026767878545745, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7576 + }, + { + "epoch": 3.027167399121055, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7577 + }, + { + "epoch": 3.027566919696364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7578 + }, + { + "epoch": 3.027966440271674, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7579 + }, + { + "epoch": 3.0283659608469837, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7580 + }, + { + "epoch": 3.028765481422293, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7581 + }, + { + "epoch": 3.029165001997603, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7582 + }, + { + "epoch": 3.0295645225729126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 7583 + }, + { + "epoch": 3.029964043148222, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7584 + }, + { + "epoch": 3.0303635637235318, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7585 + }, + { + "epoch": 3.0307630842988416, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7586 + }, + { + "epoch": 3.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7587 + }, + { + "epoch": 3.0315621254494607, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7588 + }, + { + "epoch": 3.0319616460247705, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7589 + }, + { + "epoch": 3.03236116660008, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7590 + }, + { + "epoch": 3.0327606871753896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 7591 + }, + { + "epoch": 3.033160207750699, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 7592 + }, + { + "epoch": 3.0335597283260087, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7593 + }, + { + "epoch": 3.0339592489013185, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7594 + }, + { + "epoch": 3.034358769476628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7595 + }, + { + "epoch": 3.0347582900519376, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7596 + }, + { + "epoch": 3.0351578106272474, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7597 + }, + { + "epoch": 3.0355573312025568, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 7598 + }, + { + "epoch": 3.0359568517778666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7599 + }, + { + "epoch": 3.0363563723531763, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7600 + }, + { + "epoch": 3.0367558929284857, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 7601 + }, + { + "epoch": 3.0371554135037955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7602 + }, + { + "epoch": 3.0375549340791053, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7603 + }, + { + "epoch": 3.0379544546544146, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7604 + }, + { + "epoch": 3.0383539752297244, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7605 + }, + { + "epoch": 3.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7606 + }, + { + "epoch": 3.0391530163803435, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7607 + }, + { + "epoch": 3.0395525369556533, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7608 + }, + { + "epoch": 3.0399520575309626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7609 + }, + { + "epoch": 3.0403515781062724, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7610 + }, + { + "epoch": 3.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7611 + }, + { + "epoch": 3.0411506192568916, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7612 + }, + { + "epoch": 3.0415501398322013, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7613 + }, + { + "epoch": 3.041949660407511, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7614 + }, + { + "epoch": 3.0423491809828205, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7615 + }, + { + "epoch": 3.0427487015581303, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7616 + }, + { + "epoch": 3.04314822213344, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7617 + }, + { + "epoch": 3.0435477427087494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7618 + }, + { + "epoch": 3.043947263284059, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7619 + }, + { + "epoch": 3.044346783859369, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7620 + }, + { + "epoch": 3.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7621 + }, + { + "epoch": 3.045145825009988, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7622 + }, + { + "epoch": 3.0455453455852974, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7623 + }, + { + "epoch": 3.0459448661606072, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7624 + }, + { + "epoch": 3.046344386735917, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 7625 + }, + { + "epoch": 3.0467439073112264, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7626 + }, + { + "epoch": 3.047143427886536, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7627 + }, + { + "epoch": 3.047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7628 + }, + { + "epoch": 3.0479424690371553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7629 + }, + { + "epoch": 3.048341989612465, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 7630 + }, + { + "epoch": 3.048741510187775, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7631 + }, + { + "epoch": 3.049141030763084, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7632 + }, + { + "epoch": 3.049540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7633 + }, + { + "epoch": 3.0499400719137038, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 7634 + }, + { + "epoch": 3.050339592489013, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7635 + }, + { + "epoch": 3.050739113064323, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7636 + }, + { + "epoch": 3.0511386336396322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7637 + }, + { + "epoch": 3.051538154214942, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7638 + }, + { + "epoch": 3.051937674790252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7639 + }, + { + "epoch": 3.052337195365561, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7640 + }, + { + "epoch": 3.052736715940871, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7641 + }, + { + "epoch": 3.0531362365161807, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7642 + }, + { + "epoch": 3.05353575709149, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7643 + }, + { + "epoch": 3.0539352776668, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7644 + }, + { + "epoch": 3.0543347982421096, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7645 + }, + { + "epoch": 3.054734318817419, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7646 + }, + { + "epoch": 3.0551338393927288, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7647 + }, + { + "epoch": 3.0555333599680385, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7648 + }, + { + "epoch": 3.055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7649 + }, + { + "epoch": 3.0563324011186577, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 7650 + }, + { + "epoch": 3.0567319216939675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7651 + }, + { + "epoch": 3.057131442269277, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7652 + }, + { + "epoch": 3.0575309628445866, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7653 + }, + { + "epoch": 3.057930483419896, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7654 + }, + { + "epoch": 3.0583300039952057, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7655 + }, + { + "epoch": 3.0587295245705155, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7656 + }, + { + "epoch": 3.059129045145825, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7657 + }, + { + "epoch": 3.0595285657211346, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7658 + }, + { + "epoch": 3.0599280862964444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7659 + }, + { + "epoch": 3.0603276068717538, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7660 + }, + { + "epoch": 3.0607271274470635, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7661 + }, + { + "epoch": 3.0611266480223733, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7662 + }, + { + "epoch": 3.0615261685976827, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7663 + }, + { + "epoch": 3.0619256891729925, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7664 + }, + { + "epoch": 3.0623252097483022, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7665 + }, + { + "epoch": 3.0627247303236116, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7666 + }, + { + "epoch": 3.0631242508989214, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7667 + }, + { + "epoch": 3.0635237714742307, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7668 + }, + { + "epoch": 3.0639232920495405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7669 + }, + { + "epoch": 3.0643228126248503, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7670 + }, + { + "epoch": 3.0647223332001596, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 7671 + }, + { + "epoch": 3.0651218537754694, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 7672 + }, + { + "epoch": 3.065521374350779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7673 + }, + { + "epoch": 3.0659208949260885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7674 + }, + { + "epoch": 3.0663204155013983, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7675 + }, + { + "epoch": 3.066719936076708, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7676 + }, + { + "epoch": 3.0671194566520175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7677 + }, + { + "epoch": 3.0675189772273272, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7678 + }, + { + "epoch": 3.067918497802637, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7679 + }, + { + "epoch": 3.0683180183779464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7680 + }, + { + "epoch": 3.068717538953256, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7681 + }, + { + "epoch": 3.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7682 + }, + { + "epoch": 3.0695165801038753, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7683 + }, + { + "epoch": 3.069916100679185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7684 + }, + { + "epoch": 3.0703156212544944, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7685 + }, + { + "epoch": 3.070715141829804, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7686 + }, + { + "epoch": 3.071114662405114, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7687 + }, + { + "epoch": 3.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7688 + }, + { + "epoch": 3.071913703555733, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 7689 + }, + { + "epoch": 3.072313224131043, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7690 + }, + { + "epoch": 3.0727127447063523, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7691 + }, + { + "epoch": 3.073112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7692 + }, + { + "epoch": 3.073511785856972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7693 + }, + { + "epoch": 3.073911306432281, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7694 + }, + { + "epoch": 3.074310827007591, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 7695 + }, + { + "epoch": 3.0747103475829007, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7696 + }, + { + "epoch": 3.07510986815821, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7697 + }, + { + "epoch": 3.07550938873352, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7698 + }, + { + "epoch": 3.075908909308829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7699 + }, + { + "epoch": 3.076308429884139, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7700 + }, + { + "epoch": 3.076707950459449, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7701 + }, + { + "epoch": 3.077107471034758, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7702 + }, + { + "epoch": 3.077506991610068, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7703 + }, + { + "epoch": 3.0779065121853777, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7704 + }, + { + "epoch": 3.078306032760687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7705 + }, + { + "epoch": 3.078705553335997, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7706 + }, + { + "epoch": 3.0791050739113066, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7707 + }, + { + "epoch": 3.079504594486616, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7708 + }, + { + "epoch": 3.0799041150619257, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7709 + }, + { + "epoch": 3.0803036356372355, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7710 + }, + { + "epoch": 3.080703156212545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7711 + }, + { + "epoch": 3.0811026767878547, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7712 + }, + { + "epoch": 3.081502197363164, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7713 + }, + { + "epoch": 3.081901717938474, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7714 + }, + { + "epoch": 3.0823012385137836, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7715 + }, + { + "epoch": 3.082700759089093, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7716 + }, + { + "epoch": 3.0831002796644027, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7717 + }, + { + "epoch": 3.0834998002397125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7718 + }, + { + "epoch": 3.083899320815022, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7719 + }, + { + "epoch": 3.0842988413903316, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7720 + }, + { + "epoch": 3.0846983619656414, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7721 + }, + { + "epoch": 3.0850978825409507, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7722 + }, + { + "epoch": 3.0854974031162605, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7723 + }, + { + "epoch": 3.0858969236915703, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 7724 + }, + { + "epoch": 3.0862964442668797, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7725 + }, + { + "epoch": 3.0866959648421894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7726 + }, + { + "epoch": 3.087095485417499, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7727 + }, + { + "epoch": 3.0874950059928086, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7728 + }, + { + "epoch": 3.0878945265681184, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7729 + }, + { + "epoch": 3.0882940471434277, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7730 + }, + { + "epoch": 3.0886935677187375, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7731 + }, + { + "epoch": 3.0890930882940473, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7732 + }, + { + "epoch": 3.0894926088693566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7733 + }, + { + "epoch": 3.0898921294446664, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7734 + }, + { + "epoch": 3.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7735 + }, + { + "epoch": 3.0906911705952855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7736 + }, + { + "epoch": 3.0910906911705953, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7737 + }, + { + "epoch": 3.091490211745905, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7738 + }, + { + "epoch": 3.0918897323212144, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7739 + }, + { + "epoch": 3.0922892528965242, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7740 + }, + { + "epoch": 3.092688773471834, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7741 + }, + { + "epoch": 3.0930882940471434, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7742 + }, + { + "epoch": 3.093487814622453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7743 + }, + { + "epoch": 3.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7744 + }, + { + "epoch": 3.0942868557730723, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7745 + }, + { + "epoch": 3.094686376348382, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7746 + }, + { + "epoch": 3.0950858969236914, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7747 + }, + { + "epoch": 3.095485417499001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 7748 + }, + { + "epoch": 3.095884938074311, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7749 + }, + { + "epoch": 3.0962844586496203, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7750 + }, + { + "epoch": 3.09668397922493, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7751 + }, + { + "epoch": 3.09708349980024, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7752 + }, + { + "epoch": 3.0974830203755492, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7753 + }, + { + "epoch": 3.097882540950859, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 7754 + }, + { + "epoch": 3.098282061526169, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7755 + }, + { + "epoch": 3.098681582101478, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7756 + }, + { + "epoch": 3.099081102676788, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7757 + }, + { + "epoch": 3.0994806232520973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7758 + }, + { + "epoch": 3.099880143827407, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7759 + }, + { + "epoch": 3.100279664402717, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7760 + }, + { + "epoch": 3.100679184978026, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7761 + }, + { + "epoch": 3.101078705553336, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7762 + }, + { + "epoch": 3.1014782261286458, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7763 + }, + { + "epoch": 3.101877746703955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7764 + }, + { + "epoch": 3.102277267279265, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7765 + }, + { + "epoch": 3.1026767878545747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 7766 + }, + { + "epoch": 3.103076308429884, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7767 + }, + { + "epoch": 3.103475829005194, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7768 + }, + { + "epoch": 3.1038753495805036, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7769 + }, + { + "epoch": 3.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7770 + }, + { + "epoch": 3.1046743907311227, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7771 + }, + { + "epoch": 3.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7772 + }, + { + "epoch": 3.105473431881742, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7773 + }, + { + "epoch": 3.1058729524570516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7774 + }, + { + "epoch": 3.106272473032361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7775 + }, + { + "epoch": 3.1066719936076708, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7776 + }, + { + "epoch": 3.1070715141829806, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7777 + }, + { + "epoch": 3.10747103475829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7778 + }, + { + "epoch": 3.1078705553335997, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7779 + }, + { + "epoch": 3.1082700759089095, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7780 + }, + { + "epoch": 3.108669596484219, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7781 + }, + { + "epoch": 3.1090691170595286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 7782 + }, + { + "epoch": 3.1094686376348384, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7783 + }, + { + "epoch": 3.1098681582101477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7784 + }, + { + "epoch": 3.1102676787854575, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 7785 + }, + { + "epoch": 3.1106671993607673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7786 + }, + { + "epoch": 3.1110667199360766, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7787 + }, + { + "epoch": 3.1114662405113864, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7788 + }, + { + "epoch": 3.1118657610866958, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7789 + }, + { + "epoch": 3.1122652816620056, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7790 + }, + { + "epoch": 3.1126648022373153, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7791 + }, + { + "epoch": 3.1130643228126247, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7792 + }, + { + "epoch": 3.1134638433879345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 7793 + }, + { + "epoch": 3.1138633639632443, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7794 + }, + { + "epoch": 3.1142628845385536, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7795 + }, + { + "epoch": 3.1146624051138634, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7796 + }, + { + "epoch": 3.115061925689173, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7797 + }, + { + "epoch": 3.1154614462644825, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7798 + }, + { + "epoch": 3.1158609668397923, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7799 + }, + { + "epoch": 3.116260487415102, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7800 + }, + { + "epoch": 3.1166600079904114, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7801 + }, + { + "epoch": 3.117059528565721, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 7802 + }, + { + "epoch": 3.1174590491410306, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7803 + }, + { + "epoch": 3.1178585697163403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7804 + }, + { + "epoch": 3.11825809029165, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7805 + }, + { + "epoch": 3.1186576108669595, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 7806 + }, + { + "epoch": 3.1190571314422693, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7807 + }, + { + "epoch": 3.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7808 + }, + { + "epoch": 3.1198561725928884, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7809 + }, + { + "epoch": 3.120255693168198, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7810 + }, + { + "epoch": 3.120655213743508, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7811 + }, + { + "epoch": 3.1210547343188173, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7812 + }, + { + "epoch": 3.121454254894127, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7813 + }, + { + "epoch": 3.121853775469437, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7814 + }, + { + "epoch": 3.122253296044746, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7815 + }, + { + "epoch": 3.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7816 + }, + { + "epoch": 3.1230523371953653, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7817 + }, + { + "epoch": 3.123451857770675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7818 + }, + { + "epoch": 3.123851378345985, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7819 + }, + { + "epoch": 3.1242508989212943, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7820 + }, + { + "epoch": 3.124650419496604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7821 + }, + { + "epoch": 3.125049940071914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7822 + }, + { + "epoch": 3.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7823 + }, + { + "epoch": 3.125848981222533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7824 + }, + { + "epoch": 3.1262485017978427, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7825 + }, + { + "epoch": 3.126648022373152, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7826 + }, + { + "epoch": 3.127047542948462, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7827 + }, + { + "epoch": 3.1274470635237717, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7828 + }, + { + "epoch": 3.127846584099081, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7829 + }, + { + "epoch": 3.128246104674391, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7830 + }, + { + "epoch": 3.1286456252497006, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7831 + }, + { + "epoch": 3.12904514582501, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7832 + }, + { + "epoch": 3.1294446664003197, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7833 + }, + { + "epoch": 3.129844186975629, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7834 + }, + { + "epoch": 3.130243707550939, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7835 + }, + { + "epoch": 3.1306432281262486, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7836 + }, + { + "epoch": 3.131042748701558, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7837 + }, + { + "epoch": 3.1314422692768678, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7838 + }, + { + "epoch": 3.1318417898521775, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7839 + }, + { + "epoch": 3.132241310427487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7840 + }, + { + "epoch": 3.1326408310027967, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7841 + }, + { + "epoch": 3.1330403515781065, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7842 + }, + { + "epoch": 3.133439872153416, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7843 + }, + { + "epoch": 3.1338393927287256, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7844 + }, + { + "epoch": 3.134238913304035, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7845 + }, + { + "epoch": 3.1346384338793447, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7846 + }, + { + "epoch": 3.1350379544546545, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7847 + }, + { + "epoch": 3.135437475029964, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7848 + }, + { + "epoch": 3.1358369956052736, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7849 + }, + { + "epoch": 3.1362365161805834, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7850 + }, + { + "epoch": 3.1366360367558928, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7851 + }, + { + "epoch": 3.1370355573312025, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7852 + }, + { + "epoch": 3.1374350779065123, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7853 + }, + { + "epoch": 3.1378345984818217, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7854 + }, + { + "epoch": 3.1382341190571315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7855 + }, + { + "epoch": 3.1386336396324412, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7856 + }, + { + "epoch": 3.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7857 + }, + { + "epoch": 3.1394326807830604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7858 + }, + { + "epoch": 3.13983220135837, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7859 + }, + { + "epoch": 3.1402317219336795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7860 + }, + { + "epoch": 3.1406312425089893, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7861 + }, + { + "epoch": 3.1410307630842986, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7862 + }, + { + "epoch": 3.1414302836596084, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7863 + }, + { + "epoch": 3.141829804234918, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7864 + }, + { + "epoch": 3.1422293248102275, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7865 + }, + { + "epoch": 3.1426288453855373, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7866 + }, + { + "epoch": 3.143028365960847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7867 + }, + { + "epoch": 3.1434278865361565, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7868 + }, + { + "epoch": 3.1438274071114662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7869 + }, + { + "epoch": 3.144226927686776, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7870 + }, + { + "epoch": 3.1446264482620854, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7871 + }, + { + "epoch": 3.145025968837395, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7872 + }, + { + "epoch": 3.145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7873 + }, + { + "epoch": 3.1458250099880143, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 7874 + }, + { + "epoch": 3.146224530563324, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7875 + }, + { + "epoch": 3.146624051138634, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7876 + }, + { + "epoch": 3.147023571713943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7877 + }, + { + "epoch": 3.147423092289253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7878 + }, + { + "epoch": 3.1478226128645623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7879 + }, + { + "epoch": 3.148222133439872, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7880 + }, + { + "epoch": 3.148621654015182, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7881 + }, + { + "epoch": 3.1490211745904912, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7882 + }, + { + "epoch": 3.149420695165801, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7883 + }, + { + "epoch": 3.149820215741111, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7884 + }, + { + "epoch": 3.15021973631642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7885 + }, + { + "epoch": 3.15061925689173, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7886 + }, + { + "epoch": 3.1510187774670397, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 7887 + }, + { + "epoch": 3.151418298042349, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7888 + }, + { + "epoch": 3.151817818617659, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7889 + }, + { + "epoch": 3.152217339192968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7890 + }, + { + "epoch": 3.152616859768278, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7891 + }, + { + "epoch": 3.1530163803435878, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7892 + }, + { + "epoch": 3.153415900918897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7893 + }, + { + "epoch": 3.153815421494207, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7894 + }, + { + "epoch": 3.1542149420695167, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7895 + }, + { + "epoch": 3.154614462644826, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7896 + }, + { + "epoch": 3.155013983220136, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7897 + }, + { + "epoch": 3.1554135037954456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7898 + }, + { + "epoch": 3.155813024370755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7899 + }, + { + "epoch": 3.1562125449460647, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7900 + }, + { + "epoch": 3.1566120655213745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7901 + }, + { + "epoch": 3.157011586096684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7902 + }, + { + "epoch": 3.1574111066719937, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7903 + }, + { + "epoch": 3.1578106272473034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7904 + }, + { + "epoch": 3.158210147822613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7905 + }, + { + "epoch": 3.1586096683979226, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7906 + }, + { + "epoch": 3.159009188973232, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7907 + }, + { + "epoch": 3.1594087095485417, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7908 + }, + { + "epoch": 3.1598082301238515, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 7909 + }, + { + "epoch": 3.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7910 + }, + { + "epoch": 3.1606072712744706, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7911 + }, + { + "epoch": 3.1610067918497804, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7912 + }, + { + "epoch": 3.1614063124250897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7913 + }, + { + "epoch": 3.1618058330003995, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 7914 + }, + { + "epoch": 3.1622053535757093, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7915 + }, + { + "epoch": 3.1626048741510187, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7916 + }, + { + "epoch": 3.1630043947263284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 7917 + }, + { + "epoch": 3.1634039153016382, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7918 + }, + { + "epoch": 3.1638034358769476, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7919 + }, + { + "epoch": 3.1642029564522574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7920 + }, + { + "epoch": 3.164602477027567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7921 + }, + { + "epoch": 3.1650019976028765, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7922 + }, + { + "epoch": 3.1654015181781863, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7923 + }, + { + "epoch": 3.1658010387534956, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7924 + }, + { + "epoch": 3.1662005593288054, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 7925 + }, + { + "epoch": 3.166600079904115, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7926 + }, + { + "epoch": 3.1669996004794245, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7927 + }, + { + "epoch": 3.1673991210547343, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7928 + }, + { + "epoch": 3.167798641630044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7929 + }, + { + "epoch": 3.1681981622053534, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 7930 + }, + { + "epoch": 3.1685976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7931 + }, + { + "epoch": 3.168997203355973, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7932 + }, + { + "epoch": 3.1693967239312824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7933 + }, + { + "epoch": 3.169796244506592, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7934 + }, + { + "epoch": 3.1701957650819015, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7935 + }, + { + "epoch": 3.1705952856572113, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7936 + }, + { + "epoch": 3.170994806232521, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7937 + }, + { + "epoch": 3.1713943268078304, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7938 + }, + { + "epoch": 3.17179384738314, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7939 + }, + { + "epoch": 3.17219336795845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 7940 + }, + { + "epoch": 3.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7941 + }, + { + "epoch": 3.172992409109069, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7942 + }, + { + "epoch": 3.173391929684379, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7943 + }, + { + "epoch": 3.1737914502596882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7944 + }, + { + "epoch": 3.174190970834998, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7945 + }, + { + "epoch": 3.174590491410308, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7946 + }, + { + "epoch": 3.174990011985617, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 7947 + }, + { + "epoch": 3.175389532560927, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7948 + }, + { + "epoch": 3.1757890531362367, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7949 + }, + { + "epoch": 3.176188573711546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 7950 + }, + { + "epoch": 3.176588094286856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7951 + }, + { + "epoch": 3.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7952 + }, + { + "epoch": 3.177387135437475, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7953 + }, + { + "epoch": 3.1777866560127848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7954 + }, + { + "epoch": 3.178186176588094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 7955 + }, + { + "epoch": 3.178585697163404, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7956 + }, + { + "epoch": 3.1789852177387137, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7957 + }, + { + "epoch": 3.179384738314023, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7958 + }, + { + "epoch": 3.179784258889333, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7959 + }, + { + "epoch": 3.1801837794646426, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 7960 + }, + { + "epoch": 3.180583300039952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 7961 + }, + { + "epoch": 3.1809828206152617, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7962 + }, + { + "epoch": 3.1813823411905715, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7963 + }, + { + "epoch": 3.181781861765881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7964 + }, + { + "epoch": 3.1821813823411906, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7965 + }, + { + "epoch": 3.1825809029165004, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7966 + }, + { + "epoch": 3.1829804234918098, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7967 + }, + { + "epoch": 3.1833799440671195, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7968 + }, + { + "epoch": 3.183779464642429, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7969 + }, + { + "epoch": 3.1841789852177387, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7970 + }, + { + "epoch": 3.1845785057930485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7971 + }, + { + "epoch": 3.184978026368358, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7972 + }, + { + "epoch": 3.1853775469436676, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7973 + }, + { + "epoch": 3.1857770675189774, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 7974 + }, + { + "epoch": 3.1861765880942867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7975 + }, + { + "epoch": 3.1865761086695965, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7976 + }, + { + "epoch": 3.1869756292449063, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7977 + }, + { + "epoch": 3.1873751498202156, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7978 + }, + { + "epoch": 3.1877746703955254, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7979 + }, + { + "epoch": 3.1881741909708348, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7980 + }, + { + "epoch": 3.1885737115461446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7981 + }, + { + "epoch": 3.1889732321214543, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7982 + }, + { + "epoch": 3.1893727526967637, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7983 + }, + { + "epoch": 3.1897722732720735, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7984 + }, + { + "epoch": 3.1901717938473833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7985 + }, + { + "epoch": 3.1905713144226926, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7986 + }, + { + "epoch": 3.1909708349980024, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7987 + }, + { + "epoch": 3.191370355573312, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7988 + }, + { + "epoch": 3.1917698761486215, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7989 + }, + { + "epoch": 3.1921693967239313, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7990 + }, + { + "epoch": 3.192568917299241, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7991 + }, + { + "epoch": 3.1929684378745504, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 7992 + }, + { + "epoch": 3.19336795844986, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7993 + }, + { + "epoch": 3.19376747902517, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7994 + }, + { + "epoch": 3.1941669996004793, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 7995 + }, + { + "epoch": 3.194566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7996 + }, + { + "epoch": 3.1949660407510985, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7997 + }, + { + "epoch": 3.1953655613264083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7998 + }, + { + "epoch": 3.195765081901718, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7999 + }, + { + "epoch": 3.1961646024770274, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8000 + }, + { + "epoch": 3.196564123052337, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8001 + }, + { + "epoch": 3.196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8002 + }, + { + "epoch": 3.1973631642029563, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8003 + }, + { + "epoch": 3.197762684778266, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8004 + }, + { + "epoch": 3.198162205353576, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8005 + }, + { + "epoch": 3.198561725928885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 8006 + }, + { + "epoch": 3.198961246504195, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8007 + }, + { + "epoch": 3.199360767079505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8008 + }, + { + "epoch": 3.199760287654814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8009 + }, + { + "epoch": 3.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 8010 + }, + { + "epoch": 3.2005593288054337, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8011 + }, + { + "epoch": 3.200958849380743, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8012 + }, + { + "epoch": 3.201358369956053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8013 + }, + { + "epoch": 3.201757890531362, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8014 + }, + { + "epoch": 3.202157411106672, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8015 + }, + { + "epoch": 3.2025569316819817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8016 + }, + { + "epoch": 3.202956452257291, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8017 + }, + { + "epoch": 3.203355972832601, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8018 + }, + { + "epoch": 3.2037554934079107, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8019 + }, + { + "epoch": 3.20415501398322, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 8020 + }, + { + "epoch": 3.20455453455853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8021 + }, + { + "epoch": 3.2049540551338396, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 8022 + }, + { + "epoch": 3.205353575709149, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8023 + }, + { + "epoch": 3.2057530962844587, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8024 + }, + { + "epoch": 3.206152616859768, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8025 + }, + { + "epoch": 3.206552137435078, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8026 + }, + { + "epoch": 3.2069516580103876, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8027 + }, + { + "epoch": 3.207351178585697, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8028 + }, + { + "epoch": 3.2077506991610067, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8029 + }, + { + "epoch": 3.2081502197363165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 8030 + }, + { + "epoch": 3.208549740311626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8031 + }, + { + "epoch": 3.2089492608869357, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8032 + }, + { + "epoch": 3.2093487814622454, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8033 + }, + { + "epoch": 3.209748302037555, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8034 + }, + { + "epoch": 3.2101478226128646, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8035 + }, + { + "epoch": 3.2105473431881744, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8036 + }, + { + "epoch": 3.2109468637634837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 8037 + }, + { + "epoch": 3.2113463843387935, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8038 + }, + { + "epoch": 3.2117459049141033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8039 + }, + { + "epoch": 3.2121454254894126, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8040 + }, + { + "epoch": 3.2125449460647224, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8041 + }, + { + "epoch": 3.2129444666400317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8042 + }, + { + "epoch": 3.2133439872153415, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8043 + }, + { + "epoch": 3.2137435077906513, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8044 + }, + { + "epoch": 3.2141430283659607, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8045 + }, + { + "epoch": 3.2145425489412705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8046 + }, + { + "epoch": 3.2149420695165802, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8047 + }, + { + "epoch": 3.2153415900918896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8048 + }, + { + "epoch": 3.2157411106671994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8049 + }, + { + "epoch": 3.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8050 + }, + { + "epoch": 3.2165401518178185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8051 + }, + { + "epoch": 3.2169396723931283, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8052 + }, + { + "epoch": 3.217339192968438, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8053 + }, + { + "epoch": 3.2177387135437474, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8054 + }, + { + "epoch": 3.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8055 + }, + { + "epoch": 3.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8056 + }, + { + "epoch": 3.2189372752696763, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8057 + }, + { + "epoch": 3.219336795844986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8058 + }, + { + "epoch": 3.2197363164202955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8059 + }, + { + "epoch": 3.2201358369956052, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 8060 + }, + { + "epoch": 3.220535357570915, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8061 + }, + { + "epoch": 3.2209348781462244, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8062 + }, + { + "epoch": 3.221334398721534, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8063 + }, + { + "epoch": 3.221733919296844, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8064 + }, + { + "epoch": 3.2221334398721533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8065 + }, + { + "epoch": 3.222532960447463, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 8066 + }, + { + "epoch": 3.222932481022773, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8067 + }, + { + "epoch": 3.223332001598082, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 8068 + }, + { + "epoch": 3.223731522173392, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8069 + }, + { + "epoch": 3.2241310427487013, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8070 + }, + { + "epoch": 3.224530563324011, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8071 + }, + { + "epoch": 3.224930083899321, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8072 + }, + { + "epoch": 3.2253296044746302, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8073 + }, + { + "epoch": 3.22572912504994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8074 + }, + { + "epoch": 3.22612864562525, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8075 + }, + { + "epoch": 3.226528166200559, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8076 + }, + { + "epoch": 3.226927686775869, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8077 + }, + { + "epoch": 3.2273272073511787, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8078 + }, + { + "epoch": 3.227726727926488, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8079 + }, + { + "epoch": 3.228126248501798, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8080 + }, + { + "epoch": 3.2285257690771076, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8081 + }, + { + "epoch": 3.228925289652417, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8082 + }, + { + "epoch": 3.2293248102277268, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8083 + }, + { + "epoch": 3.2297243308030366, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8084 + }, + { + "epoch": 3.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8085 + }, + { + "epoch": 3.2305233719536557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8086 + }, + { + "epoch": 3.230922892528965, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8087 + }, + { + "epoch": 3.231322413104275, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8088 + }, + { + "epoch": 3.2317219336795846, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8089 + }, + { + "epoch": 3.232121454254894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8090 + }, + { + "epoch": 3.2325209748302037, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8091 + }, + { + "epoch": 3.2329204954055135, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8092 + }, + { + "epoch": 3.233320015980823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8093 + }, + { + "epoch": 3.2337195365561326, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8094 + }, + { + "epoch": 3.2341190571314424, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8095 + }, + { + "epoch": 3.2345185777067518, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8096 + }, + { + "epoch": 3.2349180982820616, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 8097 + }, + { + "epoch": 3.2353176188573713, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8098 + }, + { + "epoch": 3.2357171394326807, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8099 + }, + { + "epoch": 3.2361166600079905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 8100 + }, + { + "epoch": 3.2365161805833003, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8101 + }, + { + "epoch": 3.2369157011586096, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 8102 + }, + { + "epoch": 3.2373152217339194, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8103 + }, + { + "epoch": 3.2377147423092287, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8104 + }, + { + "epoch": 3.2381142628845385, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8105 + }, + { + "epoch": 3.2385137834598483, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8106 + }, + { + "epoch": 3.2389133040351576, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 8107 + }, + { + "epoch": 3.2393128246104674, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8108 + }, + { + "epoch": 3.239712345185777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8109 + }, + { + "epoch": 3.2401118657610866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8110 + }, + { + "epoch": 3.2405113863363963, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 8111 + }, + { + "epoch": 3.240910906911706, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8112 + }, + { + "epoch": 3.2413104274870155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8113 + }, + { + "epoch": 3.2417099480623253, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8114 + }, + { + "epoch": 3.2421094686376346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8115 + }, + { + "epoch": 3.2425089892129444, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8116 + }, + { + "epoch": 3.242908509788254, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8117 + }, + { + "epoch": 3.2433080303635635, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8118 + }, + { + "epoch": 3.2437075509388733, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 8119 + }, + { + "epoch": 3.244107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8120 + }, + { + "epoch": 3.2445065920894924, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8121 + }, + { + "epoch": 3.2449061126648022, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8122 + }, + { + "epoch": 3.245305633240112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8123 + }, + { + "epoch": 3.2457051538154214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8124 + }, + { + "epoch": 3.246104674390731, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8125 + }, + { + "epoch": 3.246504194966041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8126 + }, + { + "epoch": 3.2469037155413503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8127 + }, + { + "epoch": 3.24730323611666, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8128 + }, + { + "epoch": 3.24770275669197, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8129 + }, + { + "epoch": 3.248102277267279, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8130 + }, + { + "epoch": 3.248501797842589, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8131 + }, + { + "epoch": 3.2489013184178983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8132 + }, + { + "epoch": 3.249300838993208, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8133 + }, + { + "epoch": 3.249700359568518, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8134 + }, + { + "epoch": 3.2500998801438272, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8135 + }, + { + "epoch": 3.250499400719137, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8136 + }, + { + "epoch": 3.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8137 + }, + { + "epoch": 3.251298441869756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8138 + }, + { + "epoch": 3.251697962445066, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 8139 + }, + { + "epoch": 3.2520974830203757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8140 + }, + { + "epoch": 3.252497003595685, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8141 + }, + { + "epoch": 3.252896524170995, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 8142 + }, + { + "epoch": 3.2532960447463046, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8143 + }, + { + "epoch": 3.253695565321614, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8144 + }, + { + "epoch": 3.2540950858969238, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8145 + }, + { + "epoch": 3.2544946064722335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8146 + }, + { + "epoch": 3.254894127047543, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8147 + }, + { + "epoch": 3.2552936476228527, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8148 + }, + { + "epoch": 3.255693168198162, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8149 + }, + { + "epoch": 3.256092688773472, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8150 + }, + { + "epoch": 3.2564922093487816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8151 + }, + { + "epoch": 3.256891729924091, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8152 + }, + { + "epoch": 3.2572912504994007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8153 + }, + { + "epoch": 3.2576907710747105, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8154 + }, + { + "epoch": 3.25809029165002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8155 + }, + { + "epoch": 3.2584898122253296, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8156 + }, + { + "epoch": 3.2588893328006394, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8157 + }, + { + "epoch": 3.2592888533759488, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8158 + }, + { + "epoch": 3.2596883739512585, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8159 + }, + { + "epoch": 3.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8160 + }, + { + "epoch": 3.2604874151018777, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8161 + }, + { + "epoch": 3.2608869356771875, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8162 + }, + { + "epoch": 3.2612864562524972, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8163 + }, + { + "epoch": 3.2616859768278066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8164 + }, + { + "epoch": 3.2620854974031164, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 8165 + }, + { + "epoch": 3.2624850179784257, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8166 + }, + { + "epoch": 3.2628845385537355, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8167 + }, + { + "epoch": 3.2632840591290453, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8168 + }, + { + "epoch": 3.2636835797043546, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8169 + }, + { + "epoch": 3.2640831002796644, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8170 + }, + { + "epoch": 3.264482620854974, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8171 + }, + { + "epoch": 3.2648821414302835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8172 + }, + { + "epoch": 3.2652816620055933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8173 + }, + { + "epoch": 3.265681182580903, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8174 + }, + { + "epoch": 3.2660807031562125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8175 + }, + { + "epoch": 3.2664802237315222, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8176 + }, + { + "epoch": 3.2668797443068316, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8177 + }, + { + "epoch": 3.2672792648821414, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8178 + }, + { + "epoch": 3.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8179 + }, + { + "epoch": 3.2680783060327605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8180 + }, + { + "epoch": 3.2684778266080703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8181 + }, + { + "epoch": 3.26887734718338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8182 + }, + { + "epoch": 3.2692768677586894, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8183 + }, + { + "epoch": 3.269676388333999, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8184 + }, + { + "epoch": 3.270075908909309, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8185 + }, + { + "epoch": 3.2704754294846183, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8186 + }, + { + "epoch": 3.270874950059928, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 8187 + }, + { + "epoch": 3.271274470635238, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8188 + }, + { + "epoch": 3.2716739912105472, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8189 + }, + { + "epoch": 3.272073511785857, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8190 + }, + { + "epoch": 3.272473032361167, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8191 + }, + { + "epoch": 3.272872552936476, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 8192 + }, + { + "epoch": 3.273272073511786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8193 + }, + { + "epoch": 3.2736715940870953, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8194 + }, + { + "epoch": 3.274071114662405, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8195 + }, + { + "epoch": 3.274470635237715, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8196 + }, + { + "epoch": 3.274870155813024, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 8197 + }, + { + "epoch": 3.275269676388334, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8198 + }, + { + "epoch": 3.275669196963644, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8199 + }, + { + "epoch": 3.276068717538953, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8200 + }, + { + "epoch": 3.276468238114263, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8201 + }, + { + "epoch": 3.2768677586895727, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8202 + }, + { + "epoch": 3.277267279264882, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8203 + }, + { + "epoch": 3.277666799840192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8204 + }, + { + "epoch": 3.278066320415501, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8205 + }, + { + "epoch": 3.278465840990811, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8206 + }, + { + "epoch": 3.2788653615661207, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8207 + }, + { + "epoch": 3.2792648821414305, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8208 + }, + { + "epoch": 3.27966440271674, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8209 + }, + { + "epoch": 3.2800639232920497, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8210 + }, + { + "epoch": 3.280463443867359, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 8211 + }, + { + "epoch": 3.280862964442669, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8212 + }, + { + "epoch": 3.2812624850179786, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8213 + }, + { + "epoch": 3.281662005593288, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 8214 + }, + { + "epoch": 3.2820615261685977, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8215 + }, + { + "epoch": 3.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8216 + }, + { + "epoch": 3.282860567319217, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8217 + }, + { + "epoch": 3.2832600878945266, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8218 + }, + { + "epoch": 3.2836596084698364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8219 + }, + { + "epoch": 3.2840591290451457, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8220 + }, + { + "epoch": 3.2844586496204555, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8221 + }, + { + "epoch": 3.284858170195765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8222 + }, + { + "epoch": 3.2852576907710747, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8223 + }, + { + "epoch": 3.2856572113463844, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8224 + }, + { + "epoch": 3.286056731921694, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8225 + }, + { + "epoch": 3.2864562524970036, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8226 + }, + { + "epoch": 3.2868557730723134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8227 + }, + { + "epoch": 3.2872552936476227, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 8228 + }, + { + "epoch": 3.2876548142229325, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8229 + }, + { + "epoch": 3.2880543347982423, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8230 + }, + { + "epoch": 3.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8231 + }, + { + "epoch": 3.2888533759488614, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8232 + }, + { + "epoch": 3.289252896524171, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8233 + }, + { + "epoch": 3.2896524170994805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8234 + }, + { + "epoch": 3.2900519376747903, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8235 + }, + { + "epoch": 3.2904514582501, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8236 + }, + { + "epoch": 3.2908509788254094, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8237 + }, + { + "epoch": 3.2912504994007192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8238 + }, + { + "epoch": 3.2916500199760286, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8239 + }, + { + "epoch": 3.2920495405513384, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8240 + }, + { + "epoch": 3.292449061126648, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8241 + }, + { + "epoch": 3.2928485817019575, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8242 + }, + { + "epoch": 3.2932481022772673, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8243 + }, + { + "epoch": 3.293647622852577, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8244 + }, + { + "epoch": 3.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8245 + }, + { + "epoch": 3.294446664003196, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8246 + }, + { + "epoch": 3.294846184578506, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8247 + }, + { + "epoch": 3.2952457051538153, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8248 + }, + { + "epoch": 3.295645225729125, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8249 + }, + { + "epoch": 3.2960447463044344, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8250 + }, + { + "epoch": 3.2964442668797442, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8251 + }, + { + "epoch": 3.296843787455054, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8252 + }, + { + "epoch": 3.297243308030364, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8253 + }, + { + "epoch": 3.297642828605673, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8254 + }, + { + "epoch": 3.298042349180983, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8255 + }, + { + "epoch": 3.2984418697562923, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8256 + }, + { + "epoch": 3.298841390331602, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8257 + }, + { + "epoch": 3.299240910906912, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8258 + }, + { + "epoch": 3.299640431482221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8259 + }, + { + "epoch": 3.300039952057531, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8260 + }, + { + "epoch": 3.3004394726328408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8261 + }, + { + "epoch": 3.30083899320815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8262 + }, + { + "epoch": 3.30123851378346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8263 + }, + { + "epoch": 3.3016380343587697, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8264 + }, + { + "epoch": 3.302037554934079, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8265 + }, + { + "epoch": 3.302437075509389, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8266 + }, + { + "epoch": 3.302836596084698, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8267 + }, + { + "epoch": 3.303236116660008, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8268 + }, + { + "epoch": 3.3036356372353177, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8269 + }, + { + "epoch": 3.304035157810627, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8270 + }, + { + "epoch": 3.304434678385937, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8271 + }, + { + "epoch": 3.3048341989612466, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8272 + }, + { + "epoch": 3.305233719536556, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8273 + }, + { + "epoch": 3.3056332401118658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8274 + }, + { + "epoch": 3.3060327606871756, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8275 + }, + { + "epoch": 3.306432281262485, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8276 + }, + { + "epoch": 3.3068318018377947, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8277 + }, + { + "epoch": 3.3072313224131045, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8278 + }, + { + "epoch": 3.307630842988414, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 8279 + }, + { + "epoch": 3.3080303635637236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8280 + }, + { + "epoch": 3.3084298841390334, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8281 + }, + { + "epoch": 3.3088294047143427, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8282 + }, + { + "epoch": 3.3092289252896525, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8283 + }, + { + "epoch": 3.309628445864962, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8284 + }, + { + "epoch": 3.3100279664402716, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8285 + }, + { + "epoch": 3.3104274870155814, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8286 + }, + { + "epoch": 3.3108270075908908, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8287 + }, + { + "epoch": 3.3112265281662006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8288 + }, + { + "epoch": 3.3116260487415103, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8289 + }, + { + "epoch": 3.3120255693168197, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8290 + }, + { + "epoch": 3.3124250898921295, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8291 + }, + { + "epoch": 3.3128246104674393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8292 + }, + { + "epoch": 3.3132241310427486, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8293 + }, + { + "epoch": 3.3136236516180584, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8294 + }, + { + "epoch": 3.3140231721933677, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8295 + }, + { + "epoch": 3.3144226927686775, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8296 + }, + { + "epoch": 3.3148222133439873, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8297 + }, + { + "epoch": 3.315221733919297, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8298 + }, + { + "epoch": 3.3156212544946064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8299 + }, + { + "epoch": 3.316020775069916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 8300 + }, + { + "epoch": 3.3164202956452256, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8301 + }, + { + "epoch": 3.3168198162205353, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8302 + }, + { + "epoch": 3.317219336795845, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8303 + }, + { + "epoch": 3.3176188573711545, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8304 + }, + { + "epoch": 3.3180183779464643, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8305 + }, + { + "epoch": 3.318417898521774, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8306 + }, + { + "epoch": 3.3188174190970834, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8307 + }, + { + "epoch": 3.319216939672393, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8308 + }, + { + "epoch": 3.319616460247703, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8309 + }, + { + "epoch": 3.3200159808230123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8310 + }, + { + "epoch": 3.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8311 + }, + { + "epoch": 3.3208150219736314, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8312 + }, + { + "epoch": 3.321214542548941, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8313 + }, + { + "epoch": 3.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8314 + }, + { + "epoch": 3.3220135836995603, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8315 + }, + { + "epoch": 3.32241310427487, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8316 + }, + { + "epoch": 3.32281262485018, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8317 + }, + { + "epoch": 3.3232121454254893, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8318 + }, + { + "epoch": 3.323611666000799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8319 + }, + { + "epoch": 3.324011186576109, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8320 + }, + { + "epoch": 3.324410707151418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8321 + }, + { + "epoch": 3.324810227726728, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8322 + }, + { + "epoch": 3.3252097483020377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8323 + }, + { + "epoch": 3.325609268877347, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8324 + }, + { + "epoch": 3.326008789452657, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8325 + }, + { + "epoch": 3.3264083100279667, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 8326 + }, + { + "epoch": 3.326807830603276, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8327 + }, + { + "epoch": 3.327207351178586, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8328 + }, + { + "epoch": 3.327606871753895, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8329 + }, + { + "epoch": 3.328006392329205, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8330 + }, + { + "epoch": 3.3284059129045147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8331 + }, + { + "epoch": 3.328805433479824, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8332 + }, + { + "epoch": 3.329204954055134, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8333 + }, + { + "epoch": 3.3296044746304436, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8334 + }, + { + "epoch": 3.330003995205753, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8335 + }, + { + "epoch": 3.3304035157810628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8336 + }, + { + "epoch": 3.3308030363563725, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8337 + }, + { + "epoch": 3.331202556931682, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8338 + }, + { + "epoch": 3.3316020775069917, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8339 + }, + { + "epoch": 3.332001598082301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8340 + }, + { + "epoch": 3.332401118657611, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 8341 + }, + { + "epoch": 3.3328006392329206, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8342 + }, + { + "epoch": 3.3332001598082304, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8343 + }, + { + "epoch": 3.3335996803835397, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8344 + }, + { + "epoch": 3.3339992009588495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8345 + }, + { + "epoch": 3.334398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8346 + }, + { + "epoch": 3.3347982421094686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8347 + }, + { + "epoch": 3.3351977626847784, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8348 + }, + { + "epoch": 3.3355972832600878, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8349 + }, + { + "epoch": 3.3359968038353975, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8350 + }, + { + "epoch": 3.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8351 + }, + { + "epoch": 3.3367958449860167, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8352 + }, + { + "epoch": 3.3371953655613265, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8353 + }, + { + "epoch": 3.3375948861366362, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8354 + }, + { + "epoch": 3.3379944067119456, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8355 + }, + { + "epoch": 3.3383939272872554, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8356 + }, + { + "epoch": 3.3387934478625647, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8357 + }, + { + "epoch": 3.3391929684378745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8358 + }, + { + "epoch": 3.3395924890131843, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 8359 + }, + { + "epoch": 3.3399920095884936, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8360 + }, + { + "epoch": 3.3403915301638034, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8361 + }, + { + "epoch": 3.340791050739113, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8362 + }, + { + "epoch": 3.3411905713144225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 8363 + }, + { + "epoch": 3.3415900918897323, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8364 + }, + { + "epoch": 3.341989612465042, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8365 + }, + { + "epoch": 3.3423891330403515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8366 + }, + { + "epoch": 3.3427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8367 + }, + { + "epoch": 3.343188174190971, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8368 + }, + { + "epoch": 3.3435876947662804, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8369 + }, + { + "epoch": 3.34398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8370 + }, + { + "epoch": 3.3443867359169, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8371 + }, + { + "epoch": 3.3447862564922093, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8372 + }, + { + "epoch": 3.345185777067519, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8373 + }, + { + "epoch": 3.3455852976428284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8374 + }, + { + "epoch": 3.345984818218138, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8375 + }, + { + "epoch": 3.346384338793448, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8376 + }, + { + "epoch": 3.3467838593687573, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8377 + }, + { + "epoch": 3.347183379944067, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8378 + }, + { + "epoch": 3.347582900519377, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8379 + }, + { + "epoch": 3.3479824210946862, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8380 + }, + { + "epoch": 3.348381941669996, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8381 + }, + { + "epoch": 3.348781462245306, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8382 + }, + { + "epoch": 3.349180982820615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8383 + }, + { + "epoch": 3.349580503395925, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8384 + }, + { + "epoch": 3.3499800239712343, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8385 + }, + { + "epoch": 3.350379544546544, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8386 + }, + { + "epoch": 3.350779065121854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8387 + }, + { + "epoch": 3.3511785856971636, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8388 + }, + { + "epoch": 3.351578106272473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8389 + }, + { + "epoch": 3.3519776268477828, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8390 + }, + { + "epoch": 3.352377147423092, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8391 + }, + { + "epoch": 3.352776667998402, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8392 + }, + { + "epoch": 3.3531761885737117, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8393 + }, + { + "epoch": 3.353575709149021, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8394 + }, + { + "epoch": 3.353975229724331, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8395 + }, + { + "epoch": 3.3543747502996406, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8396 + }, + { + "epoch": 3.35477427087495, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8397 + }, + { + "epoch": 3.3551737914502597, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8398 + }, + { + "epoch": 3.3555733120255695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8399 + }, + { + "epoch": 3.355972832600879, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 8400 + }, + { + "epoch": 3.3563723531761886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8401 + }, + { + "epoch": 3.356771873751498, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8402 + }, + { + "epoch": 3.357171394326808, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8403 + }, + { + "epoch": 3.3575709149021176, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8404 + }, + { + "epoch": 3.357970435477427, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8405 + }, + { + "epoch": 3.3583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 8406 + }, + { + "epoch": 3.3587694766280465, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8407 + }, + { + "epoch": 3.359168997203356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 8408 + }, + { + "epoch": 3.3595685177786656, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8409 + }, + { + "epoch": 3.3599680383539754, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8410 + }, + { + "epoch": 3.3603675589292847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8411 + }, + { + "epoch": 3.3607670795045945, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8412 + }, + { + "epoch": 3.361166600079904, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 8413 + }, + { + "epoch": 3.3615661206552137, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8414 + }, + { + "epoch": 3.3619656412305234, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8415 + }, + { + "epoch": 3.3623651618058332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 8416 + }, + { + "epoch": 3.3627646823811426, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8417 + }, + { + "epoch": 3.3631642029564524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8418 + }, + { + "epoch": 3.3635637235317617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8419 + }, + { + "epoch": 3.3639632441070715, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 8420 + }, + { + "epoch": 3.3643627646823813, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8421 + }, + { + "epoch": 3.3647622852576906, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8422 + }, + { + "epoch": 3.3651618058330004, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8423 + }, + { + "epoch": 3.36556132640831, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8424 + }, + { + "epoch": 3.3659608469836195, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8425 + }, + { + "epoch": 3.3663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8426 + }, + { + "epoch": 3.366759888134239, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8427 + }, + { + "epoch": 3.3671594087095484, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8428 + }, + { + "epoch": 3.3675589292848582, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8429 + }, + { + "epoch": 3.3679584498601676, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8430 + }, + { + "epoch": 3.3683579704354774, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8431 + }, + { + "epoch": 3.368757491010787, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8432 + }, + { + "epoch": 3.369157011586097, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8433 + }, + { + "epoch": 3.3695565321614063, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8434 + }, + { + "epoch": 3.369956052736716, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8435 + }, + { + "epoch": 3.3703555733120254, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8436 + }, + { + "epoch": 3.370755093887335, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8437 + }, + { + "epoch": 3.371154614462645, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8438 + }, + { + "epoch": 3.3715541350379543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8439 + }, + { + "epoch": 3.371953655613264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8440 + }, + { + "epoch": 3.372353176188574, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8441 + }, + { + "epoch": 3.3727526967638832, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8442 + }, + { + "epoch": 3.373152217339193, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8443 + }, + { + "epoch": 3.373551737914503, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8444 + }, + { + "epoch": 3.373951258489812, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8445 + }, + { + "epoch": 3.374350779065122, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8446 + }, + { + "epoch": 3.3747502996404313, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8447 + }, + { + "epoch": 3.375149820215741, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8448 + }, + { + "epoch": 3.375549340791051, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8449 + }, + { + "epoch": 3.37594886136636, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8450 + }, + { + "epoch": 3.37634838194167, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8451 + }, + { + "epoch": 3.3767479025169798, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8452 + }, + { + "epoch": 3.377147423092289, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 8453 + }, + { + "epoch": 3.377546943667599, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8454 + }, + { + "epoch": 3.3779464642429087, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8455 + }, + { + "epoch": 3.378345984818218, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8456 + }, + { + "epoch": 3.378745505393528, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8457 + }, + { + "epoch": 3.379145025968837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8458 + }, + { + "epoch": 3.379544546544147, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8459 + }, + { + "epoch": 3.3799440671194567, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8460 + }, + { + "epoch": 3.3803435876947665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8461 + }, + { + "epoch": 3.380743108270076, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8462 + }, + { + "epoch": 3.3811426288453856, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 8463 + }, + { + "epoch": 3.381542149420695, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8464 + }, + { + "epoch": 3.3819416699960048, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8465 + }, + { + "epoch": 3.3823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 8466 + }, + { + "epoch": 3.382740711146624, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 8467 + }, + { + "epoch": 3.3831402317219337, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8468 + }, + { + "epoch": 3.3835397522972435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8469 + }, + { + "epoch": 3.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8470 + }, + { + "epoch": 3.3843387934478626, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8471 + }, + { + "epoch": 3.3847383140231724, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8472 + }, + { + "epoch": 3.3851378345984817, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8473 + }, + { + "epoch": 3.3855373551737915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 8474 + }, + { + "epoch": 3.385936875749101, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8475 + }, + { + "epoch": 3.3863363963244106, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8476 + }, + { + "epoch": 3.3867359168997204, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8477 + }, + { + "epoch": 3.38713543747503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8478 + }, + { + "epoch": 3.3875349580503396, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8479 + }, + { + "epoch": 3.3879344786256493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8480 + }, + { + "epoch": 3.3883339992009587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8481 + }, + { + "epoch": 3.3887335197762685, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8482 + }, + { + "epoch": 3.3891330403515783, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8483 + }, + { + "epoch": 3.3895325609268876, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8484 + }, + { + "epoch": 3.3899320815021974, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8485 + }, + { + "epoch": 3.390331602077507, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8486 + }, + { + "epoch": 3.3907311226528165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8487 + }, + { + "epoch": 3.3911306432281263, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8488 + }, + { + "epoch": 3.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8489 + }, + { + "epoch": 3.3919296843787454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8490 + }, + { + "epoch": 3.392329204954055, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8491 + }, + { + "epoch": 3.3927287255293646, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8492 + }, + { + "epoch": 3.3931282461046743, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8493 + }, + { + "epoch": 3.393527766679984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8494 + }, + { + "epoch": 3.3939272872552935, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8495 + }, + { + "epoch": 3.3943268078306033, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8496 + }, + { + "epoch": 3.394726328405913, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8497 + }, + { + "epoch": 3.3951258489812224, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8498 + }, + { + "epoch": 3.395525369556532, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8499 + }, + { + "epoch": 3.395924890131842, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8500 + }, + { + "epoch": 3.3963244107071513, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8501 + }, + { + "epoch": 3.396723931282461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 8502 + }, + { + "epoch": 3.3971234518577704, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8503 + }, + { + "epoch": 3.39752297243308, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8504 + }, + { + "epoch": 3.39792249300839, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8505 + }, + { + "epoch": 3.3983220135837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8506 + }, + { + "epoch": 3.398721534159009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8507 + }, + { + "epoch": 3.399121054734319, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8508 + }, + { + "epoch": 3.3995205753096283, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8509 + }, + { + "epoch": 3.399920095884938, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8510 + }, + { + "epoch": 3.400319616460248, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8511 + }, + { + "epoch": 3.400719137035557, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8512 + }, + { + "epoch": 3.401118657610867, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8513 + }, + { + "epoch": 3.4015181781861767, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8514 + }, + { + "epoch": 3.401917698761486, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8515 + }, + { + "epoch": 3.402317219336796, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8516 + }, + { + "epoch": 3.4027167399121057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8517 + }, + { + "epoch": 3.403116260487415, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8518 + }, + { + "epoch": 3.403515781062725, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 8519 + }, + { + "epoch": 3.403915301638034, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8520 + }, + { + "epoch": 3.404314822213344, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8521 + }, + { + "epoch": 3.4047143427886537, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8522 + }, + { + "epoch": 3.4051138633639635, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8523 + }, + { + "epoch": 3.405513383939273, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8524 + }, + { + "epoch": 3.4059129045145826, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8525 + }, + { + "epoch": 3.406312425089892, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8526 + }, + { + "epoch": 3.4067119456652017, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8527 + }, + { + "epoch": 3.4071114662405115, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8528 + }, + { + "epoch": 3.407510986815821, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8529 + }, + { + "epoch": 3.4079105073911307, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8530 + }, + { + "epoch": 3.4083100279664404, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8531 + }, + { + "epoch": 3.40870954854175, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8532 + }, + { + "epoch": 3.4091090691170596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8533 + }, + { + "epoch": 3.4095085896923694, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8534 + }, + { + "epoch": 3.4099081102676787, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8535 + }, + { + "epoch": 3.4103076308429885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8536 + }, + { + "epoch": 3.410707151418298, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8537 + }, + { + "epoch": 3.4111066719936076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 8538 + }, + { + "epoch": 3.4115061925689174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8539 + }, + { + "epoch": 3.4119057131442267, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8540 + }, + { + "epoch": 3.4123052337195365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8541 + }, + { + "epoch": 3.4127047542948463, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8542 + }, + { + "epoch": 3.4131042748701557, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8543 + }, + { + "epoch": 3.4135037954454654, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8544 + }, + { + "epoch": 3.4139033160207752, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8545 + }, + { + "epoch": 3.4143028365960846, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8546 + }, + { + "epoch": 3.4147023571713944, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8547 + }, + { + "epoch": 3.4151018777467037, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8548 + }, + { + "epoch": 3.4155013983220135, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8549 + }, + { + "epoch": 3.4159009188973233, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8550 + }, + { + "epoch": 3.416300439472633, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8551 + }, + { + "epoch": 3.4166999600479424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8552 + }, + { + "epoch": 3.417099480623252, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8553 + }, + { + "epoch": 3.4174990011985615, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8554 + }, + { + "epoch": 3.4178985217738713, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8555 + }, + { + "epoch": 3.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8556 + }, + { + "epoch": 3.4186975629244905, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8557 + }, + { + "epoch": 3.4190970834998002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8558 + }, + { + "epoch": 3.41949660407511, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8559 + }, + { + "epoch": 3.4198961246504194, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8560 + }, + { + "epoch": 3.420295645225729, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8561 + }, + { + "epoch": 3.420695165801039, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8562 + }, + { + "epoch": 3.4210946863763483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8563 + }, + { + "epoch": 3.421494206951658, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8564 + }, + { + "epoch": 3.4218937275269674, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8565 + }, + { + "epoch": 3.422293248102277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8566 + }, + { + "epoch": 3.422692768677587, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8567 + }, + { + "epoch": 3.4230922892528968, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8568 + }, + { + "epoch": 3.423491809828206, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8569 + }, + { + "epoch": 3.423891330403516, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8570 + }, + { + "epoch": 3.4242908509788252, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8571 + }, + { + "epoch": 3.424690371554135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8572 + }, + { + "epoch": 3.425089892129445, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 8573 + }, + { + "epoch": 3.425489412704754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8574 + }, + { + "epoch": 3.425888933280064, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 8575 + }, + { + "epoch": 3.4262884538553737, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8576 + }, + { + "epoch": 3.426687974430683, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 8577 + }, + { + "epoch": 3.427087495005993, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8578 + }, + { + "epoch": 3.4274870155813026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8579 + }, + { + "epoch": 3.427886536156612, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8580 + }, + { + "epoch": 3.4282860567319218, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8581 + }, + { + "epoch": 3.428685577307231, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8582 + }, + { + "epoch": 3.429085097882541, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8583 + }, + { + "epoch": 3.4294846184578507, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8584 + }, + { + "epoch": 3.42988413903316, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8585 + }, + { + "epoch": 3.43028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8586 + }, + { + "epoch": 3.4306831801837796, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8587 + }, + { + "epoch": 3.431082700759089, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8588 + }, + { + "epoch": 3.4314822213343987, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8589 + }, + { + "epoch": 3.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8590 + }, + { + "epoch": 3.432281262485018, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8591 + }, + { + "epoch": 3.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 8592 + }, + { + "epoch": 3.433080303635637, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8593 + }, + { + "epoch": 3.4334798242109468, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 8594 + }, + { + "epoch": 3.4338793447862566, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8595 + }, + { + "epoch": 3.4342788653615663, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8596 + }, + { + "epoch": 3.4346783859368757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8597 + }, + { + "epoch": 3.4350779065121855, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 8598 + }, + { + "epoch": 3.435477427087495, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8599 + }, + { + "epoch": 3.4358769476628046, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8600 + }, + { + "epoch": 3.4362764682381144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8601 + }, + { + "epoch": 3.4366759888134237, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8602 + }, + { + "epoch": 3.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8603 + }, + { + "epoch": 3.4374750299640433, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8604 + }, + { + "epoch": 3.4378745505393526, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8605 + }, + { + "epoch": 3.4382740711146624, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8606 + }, + { + "epoch": 3.438673591689972, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8607 + }, + { + "epoch": 3.4390731122652816, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8608 + }, + { + "epoch": 3.4394726328405913, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8609 + }, + { + "epoch": 3.4398721534159007, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8610 + }, + { + "epoch": 3.4402716739912105, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 8611 + }, + { + "epoch": 3.4406711945665203, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8612 + }, + { + "epoch": 3.44107071514183, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8613 + }, + { + "epoch": 3.4414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8614 + }, + { + "epoch": 3.441869756292449, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8615 + }, + { + "epoch": 3.4422692768677585, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8616 + }, + { + "epoch": 3.4426687974430683, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8617 + }, + { + "epoch": 3.443068318018378, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8618 + }, + { + "epoch": 3.4434678385936874, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8619 + }, + { + "epoch": 3.443867359168997, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8620 + }, + { + "epoch": 3.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8621 + }, + { + "epoch": 3.4446664003196164, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8622 + }, + { + "epoch": 3.445065920894926, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8623 + }, + { + "epoch": 3.445465441470236, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8624 + }, + { + "epoch": 3.4458649620455453, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8625 + }, + { + "epoch": 3.446264482620855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8626 + }, + { + "epoch": 3.4466640031961644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8627 + }, + { + "epoch": 3.447063523771474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 8628 + }, + { + "epoch": 3.447463044346784, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8629 + }, + { + "epoch": 3.4478625649220933, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8630 + }, + { + "epoch": 3.448262085497403, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8631 + }, + { + "epoch": 3.448661606072713, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8632 + }, + { + "epoch": 3.4490611266480222, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8633 + }, + { + "epoch": 3.449460647223332, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8634 + }, + { + "epoch": 3.449860167798642, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8635 + }, + { + "epoch": 3.450259688373951, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8636 + }, + { + "epoch": 3.450659208949261, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8637 + }, + { + "epoch": 3.4510587295245703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8638 + }, + { + "epoch": 3.45145825009988, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8639 + }, + { + "epoch": 3.45185777067519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 8640 + }, + { + "epoch": 3.4522572912504996, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8641 + }, + { + "epoch": 3.452656811825809, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8642 + }, + { + "epoch": 3.4530563324011188, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8643 + }, + { + "epoch": 3.453455852976428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8644 + }, + { + "epoch": 3.453855373551738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8645 + }, + { + "epoch": 3.4542548941270477, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8646 + }, + { + "epoch": 3.454654414702357, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8647 + }, + { + "epoch": 3.455053935277667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8648 + }, + { + "epoch": 3.4554534558529766, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8649 + }, + { + "epoch": 3.455852976428286, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8650 + }, + { + "epoch": 3.4562524970035957, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 8651 + }, + { + "epoch": 3.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8652 + }, + { + "epoch": 3.457051538154215, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8653 + }, + { + "epoch": 3.4574510587295246, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8654 + }, + { + "epoch": 3.457850579304834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8655 + }, + { + "epoch": 3.4582500998801438, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8656 + }, + { + "epoch": 3.4586496204554535, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8657 + }, + { + "epoch": 3.4590491410307633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8658 + }, + { + "epoch": 3.4594486616060727, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8659 + }, + { + "epoch": 3.4598481821813825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 8660 + }, + { + "epoch": 3.460247702756692, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8661 + }, + { + "epoch": 3.4606472233320016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8662 + }, + { + "epoch": 3.4610467439073114, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8663 + }, + { + "epoch": 3.4614462644826207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8664 + }, + { + "epoch": 3.4618457850579305, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8665 + }, + { + "epoch": 3.4622453056332403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8666 + }, + { + "epoch": 3.4626448262085496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8667 + }, + { + "epoch": 3.4630443467838594, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8668 + }, + { + "epoch": 3.463443867359169, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8669 + }, + { + "epoch": 3.4638433879344785, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8670 + }, + { + "epoch": 3.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8671 + }, + { + "epoch": 3.4646424290850977, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8672 + }, + { + "epoch": 3.4650419496604075, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 8673 + }, + { + "epoch": 3.4654414702357172, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8674 + }, + { + "epoch": 3.4658409908110266, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 8675 + }, + { + "epoch": 3.4662405113863364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8676 + }, + { + "epoch": 3.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8677 + }, + { + "epoch": 3.4670395525369555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8678 + }, + { + "epoch": 3.4674390731122653, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8679 + }, + { + "epoch": 3.467838593687575, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8680 + }, + { + "epoch": 3.4682381142628844, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8681 + }, + { + "epoch": 3.468637634838194, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8682 + }, + { + "epoch": 3.4690371554135035, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8683 + }, + { + "epoch": 3.4694366759888133, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8684 + }, + { + "epoch": 3.469836196564123, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 8685 + }, + { + "epoch": 3.470235717139433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8686 + }, + { + "epoch": 3.4706352377147422, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8687 + }, + { + "epoch": 3.471034758290052, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8688 + }, + { + "epoch": 3.4714342788653614, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8689 + }, + { + "epoch": 3.471833799440671, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8690 + }, + { + "epoch": 3.472233320015981, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8691 + }, + { + "epoch": 3.4726328405912903, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8692 + }, + { + "epoch": 3.4730323611666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8693 + }, + { + "epoch": 3.47343188174191, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8694 + }, + { + "epoch": 3.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8695 + }, + { + "epoch": 3.474230922892529, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8696 + }, + { + "epoch": 3.474630443467839, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8697 + }, + { + "epoch": 3.475029964043148, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8698 + }, + { + "epoch": 3.475429484618458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 8699 + }, + { + "epoch": 3.4758290051937673, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8700 + }, + { + "epoch": 3.476228525769077, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8701 + }, + { + "epoch": 3.476628046344387, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8702 + }, + { + "epoch": 3.4770275669196966, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8703 + }, + { + "epoch": 3.477427087495006, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8704 + }, + { + "epoch": 3.4778266080703157, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8705 + }, + { + "epoch": 3.478226128645625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8706 + }, + { + "epoch": 3.478625649220935, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8707 + }, + { + "epoch": 3.4790251697962447, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8708 + }, + { + "epoch": 3.479424690371554, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8709 + }, + { + "epoch": 3.479824210946864, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8710 + }, + { + "epoch": 3.4802237315221736, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8711 + }, + { + "epoch": 3.480623252097483, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8712 + }, + { + "epoch": 3.4810227726727927, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8713 + }, + { + "epoch": 3.4814222932481025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8714 + }, + { + "epoch": 3.481821813823412, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8715 + }, + { + "epoch": 3.4822213343987216, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8716 + }, + { + "epoch": 3.482620854974031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8717 + }, + { + "epoch": 3.4830203755493407, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 8718 + }, + { + "epoch": 3.4834198961246505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8719 + }, + { + "epoch": 3.48381941669996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8720 + }, + { + "epoch": 3.4842189372752697, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8721 + }, + { + "epoch": 3.4846184578505794, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8722 + }, + { + "epoch": 3.485017978425889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8723 + }, + { + "epoch": 3.4854174990011986, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8724 + }, + { + "epoch": 3.4858170195765084, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8725 + }, + { + "epoch": 3.4862165401518177, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8726 + }, + { + "epoch": 3.4866160607271275, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 8727 + }, + { + "epoch": 3.487015581302437, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8728 + }, + { + "epoch": 3.4874151018777466, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8729 + }, + { + "epoch": 3.4878146224530564, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8730 + }, + { + "epoch": 3.488214143028366, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8731 + }, + { + "epoch": 3.4886136636036755, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8732 + }, + { + "epoch": 3.4890131841789853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8733 + }, + { + "epoch": 3.4894127047542947, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8734 + }, + { + "epoch": 3.4898122253296044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8735 + }, + { + "epoch": 3.4902117459049142, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8736 + }, + { + "epoch": 3.4906112664802236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8737 + }, + { + "epoch": 3.4910107870555334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8738 + }, + { + "epoch": 3.491410307630843, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8739 + }, + { + "epoch": 3.4918098282061525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8740 + }, + { + "epoch": 3.4922093487814623, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8741 + }, + { + "epoch": 3.492608869356772, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8742 + }, + { + "epoch": 3.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8743 + }, + { + "epoch": 3.493407910507391, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8744 + }, + { + "epoch": 3.4938074310827005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8745 + }, + { + "epoch": 3.4942069516580103, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8746 + }, + { + "epoch": 3.49460647223332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 8747 + }, + { + "epoch": 3.49500599280863, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8748 + }, + { + "epoch": 3.4954055133839392, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8749 + }, + { + "epoch": 3.495805033959249, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8750 + }, + { + "epoch": 3.4962045545345584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8751 + }, + { + "epoch": 3.496604075109868, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8752 + }, + { + "epoch": 3.497003595685178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8753 + }, + { + "epoch": 3.4974031162604873, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 8754 + }, + { + "epoch": 3.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8755 + }, + { + "epoch": 3.498202157411107, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8756 + }, + { + "epoch": 3.498601677986416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8757 + }, + { + "epoch": 3.499001198561726, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8758 + }, + { + "epoch": 3.4994007191370358, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8759 + }, + { + "epoch": 3.499800239712345, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8760 + }, + { + "epoch": 3.500199760287655, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8761 + }, + { + "epoch": 3.5005992808629642, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8762 + }, + { + "epoch": 3.500998801438274, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8763 + }, + { + "epoch": 3.501398322013584, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8764 + }, + { + "epoch": 3.5017978425888936, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8765 + }, + { + "epoch": 3.502197363164203, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8766 + }, + { + "epoch": 3.5025968837395127, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8767 + }, + { + "epoch": 3.502996404314822, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8768 + }, + { + "epoch": 3.503395924890132, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8769 + }, + { + "epoch": 3.5037954454654416, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 8770 + }, + { + "epoch": 3.504194966040751, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8771 + }, + { + "epoch": 3.5045944866160608, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8772 + }, + { + "epoch": 3.50499400719137, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8773 + }, + { + "epoch": 3.50539352776668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8774 + }, + { + "epoch": 3.5057930483419897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8775 + }, + { + "epoch": 3.5061925689172995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8776 + }, + { + "epoch": 3.506592089492609, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8777 + }, + { + "epoch": 3.5069916100679186, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8778 + }, + { + "epoch": 3.507391130643228, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 8779 + }, + { + "epoch": 3.5077906512185377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8780 + }, + { + "epoch": 3.5081901717938475, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8781 + }, + { + "epoch": 3.508589692369157, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8782 + }, + { + "epoch": 3.5089892129444666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8783 + }, + { + "epoch": 3.5093887335197764, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8784 + }, + { + "epoch": 3.5097882540950858, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8785 + }, + { + "epoch": 3.5101877746703956, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8786 + }, + { + "epoch": 3.5105872952457053, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8787 + }, + { + "epoch": 3.5109868158210147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8788 + }, + { + "epoch": 3.5113863363963245, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8789 + }, + { + "epoch": 3.511785856971634, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8790 + }, + { + "epoch": 3.5121853775469436, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8791 + }, + { + "epoch": 3.5125848981222534, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8792 + }, + { + "epoch": 3.512984418697563, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8793 + }, + { + "epoch": 3.5133839392728725, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8794 + }, + { + "epoch": 3.5137834598481823, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8795 + }, + { + "epoch": 3.5141829804234916, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8796 + }, + { + "epoch": 3.5145825009988014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8797 + }, + { + "epoch": 3.514982021574111, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 8798 + }, + { + "epoch": 3.5153815421494206, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8799 + }, + { + "epoch": 3.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8800 + }, + { + "epoch": 3.5161805833000397, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8801 + }, + { + "epoch": 3.5165801038753495, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8802 + }, + { + "epoch": 3.5169796244506593, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8803 + }, + { + "epoch": 3.517379145025969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8804 + }, + { + "epoch": 3.5177786656012784, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8805 + }, + { + "epoch": 3.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8806 + }, + { + "epoch": 3.5185777067518975, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8807 + }, + { + "epoch": 3.5189772273272073, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8808 + }, + { + "epoch": 3.519376747902517, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8809 + }, + { + "epoch": 3.519776268477827, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8810 + }, + { + "epoch": 3.520175789053136, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8811 + }, + { + "epoch": 3.520575309628446, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8812 + }, + { + "epoch": 3.5209748302037553, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 8813 + }, + { + "epoch": 3.521374350779065, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8814 + }, + { + "epoch": 3.521773871354375, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8815 + }, + { + "epoch": 3.5221733919296843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8816 + }, + { + "epoch": 3.522572912504994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8817 + }, + { + "epoch": 3.5229724330803034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8818 + }, + { + "epoch": 3.523371953655613, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8819 + }, + { + "epoch": 3.523771474230923, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8820 + }, + { + "epoch": 3.5241709948062327, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8821 + }, + { + "epoch": 3.524570515381542, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8822 + }, + { + "epoch": 3.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8823 + }, + { + "epoch": 3.525369556532161, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8824 + }, + { + "epoch": 3.525769077107471, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8825 + }, + { + "epoch": 3.526168597682781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8826 + }, + { + "epoch": 3.52656811825809, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8827 + }, + { + "epoch": 3.5269676388334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8828 + }, + { + "epoch": 3.5273671594087097, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8829 + }, + { + "epoch": 3.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8830 + }, + { + "epoch": 3.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8831 + }, + { + "epoch": 3.5285657211346386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8832 + }, + { + "epoch": 3.528965241709948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8833 + }, + { + "epoch": 3.5293647622852578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8834 + }, + { + "epoch": 3.529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8835 + }, + { + "epoch": 3.530163803435877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8836 + }, + { + "epoch": 3.5305633240111867, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8837 + }, + { + "epoch": 3.5309628445864965, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8838 + }, + { + "epoch": 3.531362365161806, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8839 + }, + { + "epoch": 3.5317618857371156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8840 + }, + { + "epoch": 3.532161406312425, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8841 + }, + { + "epoch": 3.5325609268877347, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8842 + }, + { + "epoch": 3.5329604474630445, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8843 + }, + { + "epoch": 3.533359968038354, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8844 + }, + { + "epoch": 3.5337594886136636, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8845 + }, + { + "epoch": 3.534159009188973, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8846 + }, + { + "epoch": 3.5345585297642828, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8847 + }, + { + "epoch": 3.5349580503395925, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8848 + }, + { + "epoch": 3.5353575709149023, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8849 + }, + { + "epoch": 3.5357570914902117, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8850 + }, + { + "epoch": 3.5361566120655215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8851 + }, + { + "epoch": 3.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8852 + }, + { + "epoch": 3.5369556532161406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8853 + }, + { + "epoch": 3.5373551737914504, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8854 + }, + { + "epoch": 3.53775469436676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8855 + }, + { + "epoch": 3.5381542149420695, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8856 + }, + { + "epoch": 3.5385537355173793, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8857 + }, + { + "epoch": 3.5389532560926886, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8858 + }, + { + "epoch": 3.5393527766679984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8859 + }, + { + "epoch": 3.539752297243308, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8860 + }, + { + "epoch": 3.5401518178186175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8861 + }, + { + "epoch": 3.5405513383939273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8862 + }, + { + "epoch": 3.5409508589692367, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8863 + }, + { + "epoch": 3.5413503795445465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8864 + }, + { + "epoch": 3.5417499001198562, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 8865 + }, + { + "epoch": 3.542149420695166, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8866 + }, + { + "epoch": 3.5425489412704754, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8867 + }, + { + "epoch": 3.542948461845785, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8868 + }, + { + "epoch": 3.5433479824210945, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 8869 + }, + { + "epoch": 3.5437475029964043, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8870 + }, + { + "epoch": 3.544147023571714, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8871 + }, + { + "epoch": 3.5445465441470234, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8872 + }, + { + "epoch": 3.544946064722333, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8873 + }, + { + "epoch": 3.545345585297643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8874 + }, + { + "epoch": 3.5457451058729523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8875 + }, + { + "epoch": 3.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8876 + }, + { + "epoch": 3.546544147023572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8877 + }, + { + "epoch": 3.5469436675988812, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8878 + }, + { + "epoch": 3.547343188174191, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8879 + }, + { + "epoch": 3.5477427087495004, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8880 + }, + { + "epoch": 3.54814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8881 + }, + { + "epoch": 3.54854174990012, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8882 + }, + { + "epoch": 3.5489412704754297, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8883 + }, + { + "epoch": 3.549340791050739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8884 + }, + { + "epoch": 3.549740311626049, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8885 + }, + { + "epoch": 3.550139832201358, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8886 + }, + { + "epoch": 3.550539352776668, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8887 + }, + { + "epoch": 3.5509388733519778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8888 + }, + { + "epoch": 3.551338393927287, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8889 + }, + { + "epoch": 3.551737914502597, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8890 + }, + { + "epoch": 3.5521374350779062, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8891 + }, + { + "epoch": 3.552536955653216, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8892 + }, + { + "epoch": 3.552936476228526, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8893 + }, + { + "epoch": 3.5533359968038356, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8894 + }, + { + "epoch": 3.553735517379145, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8895 + }, + { + "epoch": 3.5541350379544547, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8896 + }, + { + "epoch": 3.554534558529764, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8897 + }, + { + "epoch": 3.554934079105074, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8898 + }, + { + "epoch": 3.5553335996803836, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8899 + }, + { + "epoch": 3.5557331202556934, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8900 + }, + { + "epoch": 3.556132640831003, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8901 + }, + { + "epoch": 3.5565321614063126, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8902 + }, + { + "epoch": 3.556931681981622, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8903 + }, + { + "epoch": 3.5573312025569317, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8904 + }, + { + "epoch": 3.5577307231322415, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8905 + }, + { + "epoch": 3.558130243707551, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8906 + }, + { + "epoch": 3.5585297642828606, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8907 + }, + { + "epoch": 3.55892928485817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8908 + }, + { + "epoch": 3.5593288054334797, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8909 + }, + { + "epoch": 3.5597283260087895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8910 + }, + { + "epoch": 3.5601278465840993, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8911 + }, + { + "epoch": 3.5605273671594087, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8912 + }, + { + "epoch": 3.5609268877347184, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8913 + }, + { + "epoch": 3.561326408310028, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8914 + }, + { + "epoch": 3.5617259288853376, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8915 + }, + { + "epoch": 3.5621254494606474, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8916 + }, + { + "epoch": 3.5625249700359567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8917 + }, + { + "epoch": 3.5629244906112665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8918 + }, + { + "epoch": 3.5633240111865763, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8919 + }, + { + "epoch": 3.5637235317618856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8920 + }, + { + "epoch": 3.5641230523371954, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8921 + }, + { + "epoch": 3.564522572912505, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8922 + }, + { + "epoch": 3.5649220934878145, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8923 + }, + { + "epoch": 3.5653216140631243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8924 + }, + { + "epoch": 3.5657211346384337, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 8925 + }, + { + "epoch": 3.5661206552137434, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8926 + }, + { + "epoch": 3.5665201757890532, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8927 + }, + { + "epoch": 3.566919696364363, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8928 + }, + { + "epoch": 3.5673192169396724, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8929 + }, + { + "epoch": 3.567718737514982, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8930 + }, + { + "epoch": 3.5681182580902915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8931 + }, + { + "epoch": 3.5685177786656013, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8932 + }, + { + "epoch": 3.568917299240911, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 8933 + }, + { + "epoch": 3.5693168198162204, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8934 + }, + { + "epoch": 3.56971634039153, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8935 + }, + { + "epoch": 3.5701158609668395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8936 + }, + { + "epoch": 3.5705153815421493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8937 + }, + { + "epoch": 3.570914902117459, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8938 + }, + { + "epoch": 3.571314422692769, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8939 + }, + { + "epoch": 3.5717139432680782, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8940 + }, + { + "epoch": 3.572113463843388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8941 + }, + { + "epoch": 3.5725129844186974, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8942 + }, + { + "epoch": 3.572912504994007, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 8943 + }, + { + "epoch": 3.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8944 + }, + { + "epoch": 3.5737115461446267, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8945 + }, + { + "epoch": 3.574111066719936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8946 + }, + { + "epoch": 3.574510587295246, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8947 + }, + { + "epoch": 3.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8948 + }, + { + "epoch": 3.575309628445865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 8949 + }, + { + "epoch": 3.5757091490211748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8950 + }, + { + "epoch": 3.576108669596484, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8951 + }, + { + "epoch": 3.576508190171794, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8952 + }, + { + "epoch": 3.5769077107471032, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8953 + }, + { + "epoch": 3.577307231322413, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8954 + }, + { + "epoch": 3.577706751897723, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8955 + }, + { + "epoch": 3.5781062724730326, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8956 + }, + { + "epoch": 3.578505793048342, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8957 + }, + { + "epoch": 3.5789053136236517, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8958 + }, + { + "epoch": 3.579304834198961, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8959 + }, + { + "epoch": 3.579704354774271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8960 + }, + { + "epoch": 3.5801038753495806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8961 + }, + { + "epoch": 3.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8962 + }, + { + "epoch": 3.5809029165001998, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8963 + }, + { + "epoch": 3.5813024370755095, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8964 + }, + { + "epoch": 3.581701957650819, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8965 + }, + { + "epoch": 3.5821014782261287, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8966 + }, + { + "epoch": 3.5825009988014385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8967 + }, + { + "epoch": 3.582900519376748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8968 + }, + { + "epoch": 3.5833000399520576, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8969 + }, + { + "epoch": 3.583699560527367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8970 + }, + { + "epoch": 3.5840990811026767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8971 + }, + { + "epoch": 3.5844986016779865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8972 + }, + { + "epoch": 3.5848981222532963, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8973 + }, + { + "epoch": 3.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8974 + }, + { + "epoch": 3.5856971634039154, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 8975 + }, + { + "epoch": 3.5860966839792248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8976 + }, + { + "epoch": 3.5864962045545346, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8977 + }, + { + "epoch": 3.5868957251298443, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8978 + }, + { + "epoch": 3.5872952457051537, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8979 + }, + { + "epoch": 3.5876947662804635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8980 + }, + { + "epoch": 3.588094286855773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8981 + }, + { + "epoch": 3.5884938074310826, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 8982 + }, + { + "epoch": 3.5888933280063924, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8983 + }, + { + "epoch": 3.589292848581702, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8984 + }, + { + "epoch": 3.5896923691570115, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 8985 + }, + { + "epoch": 3.5900918897323213, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8986 + }, + { + "epoch": 3.5904914103076306, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 8987 + }, + { + "epoch": 3.5908909308829404, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8988 + }, + { + "epoch": 3.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8989 + }, + { + "epoch": 3.59168997203356, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8990 + }, + { + "epoch": 3.5920894926088693, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8991 + }, + { + "epoch": 3.592489013184179, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8992 + }, + { + "epoch": 3.5928885337594885, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8993 + }, + { + "epoch": 3.5932880543347983, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8994 + }, + { + "epoch": 3.593687574910108, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8995 + }, + { + "epoch": 3.5940870954854174, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 8996 + }, + { + "epoch": 3.594486616060727, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8997 + }, + { + "epoch": 3.5948861366360365, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8998 + }, + { + "epoch": 3.5952856572113463, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8999 + }, + { + "epoch": 3.595685177786656, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9000 + }, + { + "epoch": 3.596084698361966, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9001 + }, + { + "epoch": 3.596484218937275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9002 + }, + { + "epoch": 3.596883739512585, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 9003 + }, + { + "epoch": 3.5972832600878943, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9004 + }, + { + "epoch": 3.597682780663204, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9005 + }, + { + "epoch": 3.598082301238514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9006 + }, + { + "epoch": 3.5984818218138233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9007 + }, + { + "epoch": 3.598881342389133, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9008 + }, + { + "epoch": 3.599280862964443, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 9009 + }, + { + "epoch": 3.599680383539752, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9010 + }, + { + "epoch": 3.600079904115062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9011 + }, + { + "epoch": 3.6004794246903717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9012 + }, + { + "epoch": 3.600878945265681, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9013 + }, + { + "epoch": 3.601278465840991, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9014 + }, + { + "epoch": 3.6016779864163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9015 + }, + { + "epoch": 3.60207750699161, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9016 + }, + { + "epoch": 3.60247702756692, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9017 + }, + { + "epoch": 3.6028765481422296, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9018 + }, + { + "epoch": 3.603276068717539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9019 + }, + { + "epoch": 3.6036755892928487, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9020 + }, + { + "epoch": 3.604075109868158, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9021 + }, + { + "epoch": 3.604474630443468, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9022 + }, + { + "epoch": 3.6048741510187776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9023 + }, + { + "epoch": 3.605273671594087, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9024 + }, + { + "epoch": 3.6056731921693967, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9025 + }, + { + "epoch": 3.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9026 + }, + { + "epoch": 3.606472233320016, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9027 + }, + { + "epoch": 3.6068717538953257, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9028 + }, + { + "epoch": 3.6072712744706354, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9029 + }, + { + "epoch": 3.607670795045945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9030 + }, + { + "epoch": 3.6080703156212546, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9031 + }, + { + "epoch": 3.608469836196564, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9032 + }, + { + "epoch": 3.6088693567718737, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9033 + }, + { + "epoch": 3.6092688773471835, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 9034 + }, + { + "epoch": 3.6096683979224933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9035 + }, + { + "epoch": 3.6100679184978026, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9036 + }, + { + "epoch": 3.6104674390731124, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9037 + }, + { + "epoch": 3.6108669596484217, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9038 + }, + { + "epoch": 3.6112664802237315, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 9039 + }, + { + "epoch": 3.6116660007990413, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9040 + }, + { + "epoch": 3.6120655213743507, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9041 + }, + { + "epoch": 3.6124650419496604, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9042 + }, + { + "epoch": 3.61286456252497, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9043 + }, + { + "epoch": 3.6132640831002796, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9044 + }, + { + "epoch": 3.6136636036755894, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9045 + }, + { + "epoch": 3.614063124250899, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9046 + }, + { + "epoch": 3.6144626448262085, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9047 + }, + { + "epoch": 3.6148621654015183, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9048 + }, + { + "epoch": 3.6152616859768276, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9049 + }, + { + "epoch": 3.6156612065521374, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9050 + }, + { + "epoch": 3.616060727127447, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9051 + }, + { + "epoch": 3.6164602477027565, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9052 + }, + { + "epoch": 3.6168597682780663, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9053 + }, + { + "epoch": 3.617259288853376, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9054 + }, + { + "epoch": 3.6176588094286855, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9055 + }, + { + "epoch": 3.6180583300039952, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9056 + }, + { + "epoch": 3.618457850579305, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9057 + }, + { + "epoch": 3.6188573711546144, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9058 + }, + { + "epoch": 3.619256891729924, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9059 + }, + { + "epoch": 3.6196564123052335, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9060 + }, + { + "epoch": 3.6200559328805433, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9061 + }, + { + "epoch": 3.620455453455853, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9062 + }, + { + "epoch": 3.620854974031163, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 9063 + }, + { + "epoch": 3.621254494606472, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9064 + }, + { + "epoch": 3.621654015181782, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9065 + }, + { + "epoch": 3.6220535357570913, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9066 + }, + { + "epoch": 3.622453056332401, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9067 + }, + { + "epoch": 3.622852576907711, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9068 + }, + { + "epoch": 3.6232520974830202, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9069 + }, + { + "epoch": 3.62365161805833, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9070 + }, + { + "epoch": 3.6240511386336394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9071 + }, + { + "epoch": 3.624450659208949, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9072 + }, + { + "epoch": 3.624850179784259, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9073 + }, + { + "epoch": 3.6252497003595687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9074 + }, + { + "epoch": 3.625649220934878, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 9075 + }, + { + "epoch": 3.626048741510188, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9076 + }, + { + "epoch": 3.626448262085497, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9077 + }, + { + "epoch": 3.626847782660807, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9078 + }, + { + "epoch": 3.6272473032361168, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9079 + }, + { + "epoch": 3.6276468238114266, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9080 + }, + { + "epoch": 3.628046344386736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9081 + }, + { + "epoch": 3.6284458649620457, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9082 + }, + { + "epoch": 3.628845385537355, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9083 + }, + { + "epoch": 3.629244906112665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9084 + }, + { + "epoch": 3.6296444266879746, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9085 + }, + { + "epoch": 3.630043947263284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9086 + }, + { + "epoch": 3.6304434678385937, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9087 + }, + { + "epoch": 3.630842988413903, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9088 + }, + { + "epoch": 3.631242508989213, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9089 + }, + { + "epoch": 3.6316420295645226, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9090 + }, + { + "epoch": 3.6320415501398324, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9091 + }, + { + "epoch": 3.6324410707151418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 9092 + }, + { + "epoch": 3.6328405912904516, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9093 + }, + { + "epoch": 3.633240111865761, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9094 + }, + { + "epoch": 3.6336396324410707, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9095 + }, + { + "epoch": 3.6340391530163805, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9096 + }, + { + "epoch": 3.63443867359169, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9097 + }, + { + "epoch": 3.6348381941669996, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9098 + }, + { + "epoch": 3.6352377147423094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9099 + }, + { + "epoch": 3.6356372353176187, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9100 + }, + { + "epoch": 3.6360367558929285, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9101 + }, + { + "epoch": 3.6364362764682383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9102 + }, + { + "epoch": 3.6368357970435476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9103 + }, + { + "epoch": 3.6372353176188574, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9104 + }, + { + "epoch": 3.6376348381941668, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9105 + }, + { + "epoch": 3.6380343587694766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9106 + }, + { + "epoch": 3.6384338793447863, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9107 + }, + { + "epoch": 3.638833399920096, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9108 + }, + { + "epoch": 3.6392329204954055, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9109 + }, + { + "epoch": 3.6396324410707153, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9110 + }, + { + "epoch": 3.6400319616460246, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9111 + }, + { + "epoch": 3.6404314822213344, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9112 + }, + { + "epoch": 3.640831002796644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9113 + }, + { + "epoch": 3.6412305233719535, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9114 + }, + { + "epoch": 3.6416300439472633, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9115 + }, + { + "epoch": 3.6420295645225726, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 9116 + }, + { + "epoch": 3.6424290850978824, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9117 + }, + { + "epoch": 3.642828605673192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9118 + }, + { + "epoch": 3.643228126248502, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9119 + }, + { + "epoch": 3.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9120 + }, + { + "epoch": 3.644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9121 + }, + { + "epoch": 3.6444266879744305, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9122 + }, + { + "epoch": 3.6448262085497403, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9123 + }, + { + "epoch": 3.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9124 + }, + { + "epoch": 3.64562524970036, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9125 + }, + { + "epoch": 3.646024770275669, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9126 + }, + { + "epoch": 3.646424290850979, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9127 + }, + { + "epoch": 3.6468238114262883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9128 + }, + { + "epoch": 3.647223332001598, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9129 + }, + { + "epoch": 3.647622852576908, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9130 + }, + { + "epoch": 3.6480223731522172, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9131 + }, + { + "epoch": 3.648421893727527, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9132 + }, + { + "epoch": 3.6488214143028364, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 9133 + }, + { + "epoch": 3.649220934878146, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9134 + }, + { + "epoch": 3.649620455453456, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9135 + }, + { + "epoch": 3.6500199760287657, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9136 + }, + { + "epoch": 3.650419496604075, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9137 + }, + { + "epoch": 3.650819017179385, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9138 + }, + { + "epoch": 3.651218537754694, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9139 + }, + { + "epoch": 3.651618058330004, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 9140 + }, + { + "epoch": 3.6520175789053138, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9141 + }, + { + "epoch": 3.652417099480623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9142 + }, + { + "epoch": 3.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9143 + }, + { + "epoch": 3.6532161406312427, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9144 + }, + { + "epoch": 3.653615661206552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 9145 + }, + { + "epoch": 3.654015181781862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9146 + }, + { + "epoch": 3.6544147023571716, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9147 + }, + { + "epoch": 3.654814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9148 + }, + { + "epoch": 3.6552137435077907, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9149 + }, + { + "epoch": 3.6556132640831, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9150 + }, + { + "epoch": 3.65601278465841, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9151 + }, + { + "epoch": 3.6564123052337196, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9152 + }, + { + "epoch": 3.6568118258090294, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9153 + }, + { + "epoch": 3.6572113463843388, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9154 + }, + { + "epoch": 3.6576108669596485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 9155 + }, + { + "epoch": 3.658010387534958, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9156 + }, + { + "epoch": 3.6584099081102677, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9157 + }, + { + "epoch": 3.6588094286855775, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9158 + }, + { + "epoch": 3.659208949260887, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9159 + }, + { + "epoch": 3.6596084698361966, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9160 + }, + { + "epoch": 3.660007990411506, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9161 + }, + { + "epoch": 3.6604075109868157, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9162 + }, + { + "epoch": 3.6608070315621255, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9163 + }, + { + "epoch": 3.6612065521374353, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9164 + }, + { + "epoch": 3.6616060727127446, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9165 + }, + { + "epoch": 3.6620055932880544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9166 + }, + { + "epoch": 3.6624051138633638, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9167 + }, + { + "epoch": 3.6628046344386735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9168 + }, + { + "epoch": 3.6632041550139833, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9169 + }, + { + "epoch": 3.663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9170 + }, + { + "epoch": 3.6640031961646025, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9171 + }, + { + "epoch": 3.6644027167399122, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9172 + }, + { + "epoch": 3.6648022373152216, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9173 + }, + { + "epoch": 3.6652017578905314, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9174 + }, + { + "epoch": 3.665601278465841, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9175 + }, + { + "epoch": 3.6660007990411505, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9176 + }, + { + "epoch": 3.6664003196164603, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9177 + }, + { + "epoch": 3.6667998401917696, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9178 + }, + { + "epoch": 3.6671993607670794, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9179 + }, + { + "epoch": 3.667598881342389, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9180 + }, + { + "epoch": 3.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9181 + }, + { + "epoch": 3.6683979224930083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9182 + }, + { + "epoch": 3.668797443068318, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9183 + }, + { + "epoch": 3.6691969636436275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9184 + }, + { + "epoch": 3.6695964842189372, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9185 + }, + { + "epoch": 3.669996004794247, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9186 + }, + { + "epoch": 3.6703955253695564, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9187 + }, + { + "epoch": 3.670795045944866, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9188 + }, + { + "epoch": 3.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9189 + }, + { + "epoch": 3.6715940870954853, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9190 + }, + { + "epoch": 3.671993607670795, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9191 + }, + { + "epoch": 3.672393128246105, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9192 + }, + { + "epoch": 3.672792648821414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9193 + }, + { + "epoch": 3.673192169396724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9194 + }, + { + "epoch": 3.6735916899720333, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9195 + }, + { + "epoch": 3.673991210547343, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9196 + }, + { + "epoch": 3.674390731122653, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9197 + }, + { + "epoch": 3.6747902516979627, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9198 + }, + { + "epoch": 3.675189772273272, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9199 + }, + { + "epoch": 3.675589292848582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9200 + }, + { + "epoch": 3.675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 9201 + }, + { + "epoch": 3.676388333999201, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9202 + }, + { + "epoch": 3.6767878545745107, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9203 + }, + { + "epoch": 3.67718737514982, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9204 + }, + { + "epoch": 3.67758689572513, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9205 + }, + { + "epoch": 3.677986416300439, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9206 + }, + { + "epoch": 3.678385936875749, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9207 + }, + { + "epoch": 3.678785457451059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9208 + }, + { + "epoch": 3.6791849780263686, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9209 + }, + { + "epoch": 3.679584498601678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9210 + }, + { + "epoch": 3.6799840191769877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9211 + }, + { + "epoch": 3.680383539752297, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9212 + }, + { + "epoch": 3.680783060327607, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9213 + }, + { + "epoch": 3.6811825809029166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9214 + }, + { + "epoch": 3.6815821014782264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9215 + }, + { + "epoch": 3.6819816220535357, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9216 + }, + { + "epoch": 3.6823811426288455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9217 + }, + { + "epoch": 3.682780663204155, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9218 + }, + { + "epoch": 3.6831801837794647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9219 + }, + { + "epoch": 3.6835797043547744, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9220 + }, + { + "epoch": 3.683979224930084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9221 + }, + { + "epoch": 3.6843787455053936, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9222 + }, + { + "epoch": 3.684778266080703, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9223 + }, + { + "epoch": 3.6851777866560127, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9224 + }, + { + "epoch": 3.6855773072313225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9225 + }, + { + "epoch": 3.6859768278066323, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9226 + }, + { + "epoch": 3.6863763483819416, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9227 + }, + { + "epoch": 3.6867758689572514, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9228 + }, + { + "epoch": 3.6871753895325607, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9229 + }, + { + "epoch": 3.6875749101078705, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9230 + }, + { + "epoch": 3.6879744306831803, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9231 + }, + { + "epoch": 3.6883739512584897, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9232 + }, + { + "epoch": 3.6887734718337994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9233 + }, + { + "epoch": 3.6891729924091092, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9234 + }, + { + "epoch": 3.6895725129844186, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9235 + }, + { + "epoch": 3.6899720335597284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9236 + }, + { + "epoch": 3.690371554135038, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9237 + }, + { + "epoch": 3.6907710747103475, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9238 + }, + { + "epoch": 3.6911705952856573, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9239 + }, + { + "epoch": 3.6915701158609666, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9240 + }, + { + "epoch": 3.6919696364362764, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9241 + }, + { + "epoch": 3.692369157011586, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9242 + }, + { + "epoch": 3.692768677586896, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9243 + }, + { + "epoch": 3.6931681981622053, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9244 + }, + { + "epoch": 3.693567718737515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9245 + }, + { + "epoch": 3.6939672393128244, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9246 + }, + { + "epoch": 3.6943667598881342, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9247 + }, + { + "epoch": 3.694766280463444, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9248 + }, + { + "epoch": 3.6951658010387534, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 9249 + }, + { + "epoch": 3.695565321614063, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9250 + }, + { + "epoch": 3.6959648421893725, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 9251 + }, + { + "epoch": 3.6963643627646823, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 9252 + }, + { + "epoch": 3.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9253 + }, + { + "epoch": 3.697163403915302, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9254 + }, + { + "epoch": 3.697562924490611, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9255 + }, + { + "epoch": 3.697962445065921, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9256 + }, + { + "epoch": 3.6983619656412303, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9257 + }, + { + "epoch": 3.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9258 + }, + { + "epoch": 3.69916100679185, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9259 + }, + { + "epoch": 3.6995605273671597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9260 + }, + { + "epoch": 3.699960047942469, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9261 + }, + { + "epoch": 3.700359568517779, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 9262 + }, + { + "epoch": 3.700759089093088, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9263 + }, + { + "epoch": 3.701158609668398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9264 + }, + { + "epoch": 3.7015581302437077, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9265 + }, + { + "epoch": 3.701957650819017, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9266 + }, + { + "epoch": 3.702357171394327, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9267 + }, + { + "epoch": 3.702756691969636, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9268 + }, + { + "epoch": 3.703156212544946, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9269 + }, + { + "epoch": 3.7035557331202558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9270 + }, + { + "epoch": 3.7039552536955656, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9271 + }, + { + "epoch": 3.704354774270875, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9272 + }, + { + "epoch": 3.7047542948461847, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9273 + }, + { + "epoch": 3.705153815421494, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9274 + }, + { + "epoch": 3.705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9275 + }, + { + "epoch": 3.7059528565721136, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9276 + }, + { + "epoch": 3.706352377147423, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9277 + }, + { + "epoch": 3.7067518977227327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9278 + }, + { + "epoch": 3.7071514182980425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9279 + }, + { + "epoch": 3.707550938873352, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9280 + }, + { + "epoch": 3.7079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9281 + }, + { + "epoch": 3.7083499800239714, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9282 + }, + { + "epoch": 3.7087495005992808, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9283 + }, + { + "epoch": 3.7091490211745906, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9284 + }, + { + "epoch": 3.7095485417499, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9285 + }, + { + "epoch": 3.7099480623252097, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9286 + }, + { + "epoch": 3.7103475829005195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9287 + }, + { + "epoch": 3.7107471034758293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9288 + }, + { + "epoch": 3.7111466240511386, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9289 + }, + { + "epoch": 3.7115461446264484, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9290 + }, + { + "epoch": 3.7119456652017577, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9291 + }, + { + "epoch": 3.7123451857770675, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9292 + }, + { + "epoch": 3.7127447063523773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9293 + }, + { + "epoch": 3.7131442269276866, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9294 + }, + { + "epoch": 3.7135437475029964, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9295 + }, + { + "epoch": 3.7139432680783058, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 9296 + }, + { + "epoch": 3.7143427886536156, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9297 + }, + { + "epoch": 3.7147423092289253, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9298 + }, + { + "epoch": 3.715141829804235, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9299 + }, + { + "epoch": 3.7155413503795445, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9300 + }, + { + "epoch": 3.7159408709548543, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 9301 + }, + { + "epoch": 3.7163403915301636, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9302 + }, + { + "epoch": 3.7167399121054734, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 9303 + }, + { + "epoch": 3.717139432680783, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9304 + }, + { + "epoch": 3.717538953256093, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9305 + }, + { + "epoch": 3.7179384738314023, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9306 + }, + { + "epoch": 3.718337994406712, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 9307 + }, + { + "epoch": 3.7187375149820214, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9308 + }, + { + "epoch": 3.719137035557331, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9309 + }, + { + "epoch": 3.719536556132641, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9310 + }, + { + "epoch": 3.7199360767079503, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9311 + }, + { + "epoch": 3.72033559728326, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 9312 + }, + { + "epoch": 3.7207351178585695, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9313 + }, + { + "epoch": 3.7211346384338793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9314 + }, + { + "epoch": 3.721534159009189, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9315 + }, + { + "epoch": 3.721933679584499, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9316 + }, + { + "epoch": 3.722333200159808, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9317 + }, + { + "epoch": 3.722732720735118, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9318 + }, + { + "epoch": 3.7231322413104273, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9319 + }, + { + "epoch": 3.723531761885737, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 9320 + }, + { + "epoch": 3.723931282461047, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9321 + }, + { + "epoch": 3.724330803036356, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9322 + }, + { + "epoch": 3.724730323611666, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9323 + }, + { + "epoch": 3.725129844186976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9324 + }, + { + "epoch": 3.725529364762285, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9325 + }, + { + "epoch": 3.725928885337595, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9326 + }, + { + "epoch": 3.7263284059129047, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9327 + }, + { + "epoch": 3.726727926488214, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9328 + }, + { + "epoch": 3.727127447063524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9329 + }, + { + "epoch": 3.727526967638833, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 9330 + }, + { + "epoch": 3.727926488214143, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9331 + }, + { + "epoch": 3.7283260087894528, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9332 + }, + { + "epoch": 3.7287255293647625, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9333 + }, + { + "epoch": 3.729125049940072, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9334 + }, + { + "epoch": 3.7295245705153817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 9335 + }, + { + "epoch": 3.729924091090691, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9336 + }, + { + "epoch": 3.730323611666001, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9337 + }, + { + "epoch": 3.7307231322413106, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9338 + }, + { + "epoch": 3.73112265281662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9339 + }, + { + "epoch": 3.7315221733919297, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9340 + }, + { + "epoch": 3.731921693967239, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9341 + }, + { + "epoch": 3.732321214542549, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9342 + }, + { + "epoch": 3.7327207351178586, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9343 + }, + { + "epoch": 3.7331202556931684, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9344 + }, + { + "epoch": 3.7335197762684778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9345 + }, + { + "epoch": 3.7339192968437875, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9346 + }, + { + "epoch": 3.734318817419097, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9347 + }, + { + "epoch": 3.7347183379944067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9348 + }, + { + "epoch": 3.7351178585697165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9349 + }, + { + "epoch": 3.7355173791450262, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9350 + }, + { + "epoch": 3.7359168997203356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9351 + }, + { + "epoch": 3.7363164202956454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9352 + }, + { + "epoch": 3.7367159408709547, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9353 + }, + { + "epoch": 3.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 9354 + }, + { + "epoch": 3.7375149820215743, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9355 + }, + { + "epoch": 3.7379145025968836, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9356 + }, + { + "epoch": 3.7383140231721934, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9357 + }, + { + "epoch": 3.7387135437475028, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 9358 + }, + { + "epoch": 3.7391130643228125, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 9359 + }, + { + "epoch": 3.7395125848981223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9360 + }, + { + "epoch": 3.739912105473432, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9361 + }, + { + "epoch": 3.7403116260487415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9362 + }, + { + "epoch": 3.7407111466240512, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 9363 + }, + { + "epoch": 3.7411106671993606, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9364 + }, + { + "epoch": 3.7415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9365 + }, + { + "epoch": 3.74190970834998, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9366 + }, + { + "epoch": 3.7423092289252895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9367 + }, + { + "epoch": 3.7427087495005993, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9368 + }, + { + "epoch": 3.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 9369 + }, + { + "epoch": 3.7435077906512184, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 9370 + }, + { + "epoch": 3.743907311226528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9371 + }, + { + "epoch": 3.744306831801838, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9372 + }, + { + "epoch": 3.7447063523771473, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 9373 + }, + { + "epoch": 3.745105872952457, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9374 + }, + { + "epoch": 3.7455053935277665, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9375 + }, + { + "epoch": 3.7459049141030762, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9376 + }, + { + "epoch": 3.746304434678386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9377 + }, + { + "epoch": 3.746703955253696, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9378 + }, + { + "epoch": 3.747103475829005, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9379 + }, + { + "epoch": 3.747502996404315, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9380 + }, + { + "epoch": 3.7479025169796243, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9381 + }, + { + "epoch": 3.748302037554934, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9382 + }, + { + "epoch": 3.748701558130244, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9383 + }, + { + "epoch": 3.749101078705553, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9384 + }, + { + "epoch": 3.749500599280863, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9385 + }, + { + "epoch": 3.7499001198561723, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9386 + }, + { + "epoch": 3.750299640431482, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9387 + }, + { + "epoch": 3.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9388 + }, + { + "epoch": 3.7510986815821017, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9389 + }, + { + "epoch": 3.751498202157411, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9390 + }, + { + "epoch": 3.751897722732721, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9391 + }, + { + "epoch": 3.75229724330803, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9392 + }, + { + "epoch": 3.75269676388334, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9393 + }, + { + "epoch": 3.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9394 + }, + { + "epoch": 3.7534958050339595, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9395 + }, + { + "epoch": 3.753895325609269, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9396 + }, + { + "epoch": 3.7542948461845786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9397 + }, + { + "epoch": 3.754694366759888, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9398 + }, + { + "epoch": 3.7550938873351978, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9399 + }, + { + "epoch": 3.7554934079105076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9400 + }, + { + "epoch": 3.755892928485817, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9401 + }, + { + "epoch": 3.7562924490611267, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9402 + }, + { + "epoch": 3.756691969636436, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9403 + }, + { + "epoch": 3.757091490211746, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9404 + }, + { + "epoch": 3.7574910107870556, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9405 + }, + { + "epoch": 3.7578905313623654, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9406 + }, + { + "epoch": 3.7582900519376747, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9407 + }, + { + "epoch": 3.7586895725129845, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9408 + }, + { + "epoch": 3.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9409 + }, + { + "epoch": 3.7594886136636037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9410 + }, + { + "epoch": 3.7598881342389134, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9411 + }, + { + "epoch": 3.760287654814223, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9412 + }, + { + "epoch": 3.7606871753895326, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9413 + }, + { + "epoch": 3.7610866959648424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9414 + }, + { + "epoch": 3.7614862165401517, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9415 + }, + { + "epoch": 3.7618857371154615, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9416 + }, + { + "epoch": 3.7622852576907713, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9417 + }, + { + "epoch": 3.7626847782660806, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9418 + }, + { + "epoch": 3.7630842988413904, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9419 + }, + { + "epoch": 3.7634838194166997, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9420 + }, + { + "epoch": 3.7638833399920095, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9421 + }, + { + "epoch": 3.7642828605673193, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9422 + }, + { + "epoch": 3.764682381142629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9423 + }, + { + "epoch": 3.7650819017179384, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9424 + }, + { + "epoch": 3.7654814222932482, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9425 + }, + { + "epoch": 3.7658809428685576, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9426 + }, + { + "epoch": 3.7662804634438674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9427 + }, + { + "epoch": 3.766679984019177, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9428 + }, + { + "epoch": 3.7670795045944865, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9429 + }, + { + "epoch": 3.7674790251697963, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9430 + }, + { + "epoch": 3.7678785457451056, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9431 + }, + { + "epoch": 3.7682780663204154, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9432 + }, + { + "epoch": 3.768677586895725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9433 + }, + { + "epoch": 3.769077107471035, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9434 + }, + { + "epoch": 3.7694766280463443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9435 + }, + { + "epoch": 3.769876148621654, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9436 + }, + { + "epoch": 3.7702756691969634, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9437 + }, + { + "epoch": 3.7706751897722732, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9438 + }, + { + "epoch": 3.771074710347583, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9439 + }, + { + "epoch": 3.771474230922893, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 9440 + }, + { + "epoch": 3.771873751498202, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9441 + }, + { + "epoch": 3.772273272073512, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 9442 + }, + { + "epoch": 3.7726727926488213, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9443 + }, + { + "epoch": 3.773072313224131, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9444 + }, + { + "epoch": 3.773471833799441, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9445 + }, + { + "epoch": 3.77387135437475, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9446 + }, + { + "epoch": 3.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9447 + }, + { + "epoch": 3.7746703955253693, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9448 + }, + { + "epoch": 3.775069916100679, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 9449 + }, + { + "epoch": 3.775469436675989, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9450 + }, + { + "epoch": 3.7758689572512987, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9451 + }, + { + "epoch": 3.776268477826608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9452 + }, + { + "epoch": 3.776667998401918, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9453 + }, + { + "epoch": 3.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9454 + }, + { + "epoch": 3.777467039552537, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9455 + }, + { + "epoch": 3.7778665601278467, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9456 + }, + { + "epoch": 3.778266080703156, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9457 + }, + { + "epoch": 3.778665601278466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9458 + }, + { + "epoch": 3.7790651218537756, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9459 + }, + { + "epoch": 3.779464642429085, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9460 + }, + { + "epoch": 3.7798641630043948, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9461 + }, + { + "epoch": 3.7802636835797045, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9462 + }, + { + "epoch": 3.780663204155014, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9463 + }, + { + "epoch": 3.7810627247303237, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9464 + }, + { + "epoch": 3.781462245305633, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9465 + }, + { + "epoch": 3.781861765880943, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9466 + }, + { + "epoch": 3.7822612864562526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9467 + }, + { + "epoch": 3.7826608070315624, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9468 + }, + { + "epoch": 3.7830603276068717, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9469 + }, + { + "epoch": 3.7834598481821815, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9470 + }, + { + "epoch": 3.783859368757491, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 9471 + }, + { + "epoch": 3.7842588893328006, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9472 + }, + { + "epoch": 3.7846584099081104, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9473 + }, + { + "epoch": 3.7850579304834198, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9474 + }, + { + "epoch": 3.7854574510587295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9475 + }, + { + "epoch": 3.785856971634039, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9476 + }, + { + "epoch": 3.7862564922093487, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9477 + }, + { + "epoch": 3.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9478 + }, + { + "epoch": 3.7870555333599683, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9479 + }, + { + "epoch": 3.7874550539352776, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9480 + }, + { + "epoch": 3.7878545745105874, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9481 + }, + { + "epoch": 3.7882540950858967, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 9482 + }, + { + "epoch": 3.7886536156612065, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9483 + }, + { + "epoch": 3.7890531362365163, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9484 + }, + { + "epoch": 3.789452656811826, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9485 + }, + { + "epoch": 3.7898521773871354, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 9486 + }, + { + "epoch": 3.790251697962445, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9487 + }, + { + "epoch": 3.7906512185377546, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9488 + }, + { + "epoch": 3.7910507391130643, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9489 + }, + { + "epoch": 3.791450259688374, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 9490 + }, + { + "epoch": 3.7918497802636835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9491 + }, + { + "epoch": 3.7922493008389933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9492 + }, + { + "epoch": 3.7926488214143026, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9493 + }, + { + "epoch": 3.7930483419896124, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9494 + }, + { + "epoch": 3.793447862564922, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9495 + }, + { + "epoch": 3.793847383140232, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9496 + }, + { + "epoch": 3.7942469037155413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9497 + }, + { + "epoch": 3.794646424290851, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9498 + }, + { + "epoch": 3.7950459448661604, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9499 + }, + { + "epoch": 3.79544546544147, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9500 + }, + { + "epoch": 3.79584498601678, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9501 + }, + { + "epoch": 3.7962445065920893, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9502 + }, + { + "epoch": 3.796644027167399, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9503 + }, + { + "epoch": 3.797043547742709, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9504 + }, + { + "epoch": 3.7974430683180183, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9505 + }, + { + "epoch": 3.797842588893328, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9506 + }, + { + "epoch": 3.798242109468638, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9507 + }, + { + "epoch": 3.798641630043947, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9508 + }, + { + "epoch": 3.799041150619257, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 9509 + }, + { + "epoch": 3.7994406711945663, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 9510 + }, + { + "epoch": 3.799840191769876, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9511 + }, + { + "epoch": 3.800239712345186, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9512 + }, + { + "epoch": 3.8006392329204957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9513 + }, + { + "epoch": 3.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9514 + }, + { + "epoch": 3.801438274071115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9515 + }, + { + "epoch": 3.801837794646424, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9516 + }, + { + "epoch": 3.802237315221734, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9517 + }, + { + "epoch": 3.8026368357970437, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9518 + }, + { + "epoch": 3.803036356372353, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9519 + }, + { + "epoch": 3.803435876947663, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9520 + }, + { + "epoch": 3.803835397522972, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9521 + }, + { + "epoch": 3.804234918098282, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9522 + }, + { + "epoch": 3.8046344386735917, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9523 + }, + { + "epoch": 3.8050339592489015, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9524 + }, + { + "epoch": 3.805433479824211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9525 + }, + { + "epoch": 3.8058330003995207, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9526 + }, + { + "epoch": 3.80623252097483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9527 + }, + { + "epoch": 3.80663204155014, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9528 + }, + { + "epoch": 3.8070315621254496, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9529 + }, + { + "epoch": 3.8074310827007594, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9530 + }, + { + "epoch": 3.8078306032760687, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9531 + }, + { + "epoch": 3.8082301238513785, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9532 + }, + { + "epoch": 3.808629644426688, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9533 + }, + { + "epoch": 3.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9534 + }, + { + "epoch": 3.8094286855773074, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9535 + }, + { + "epoch": 3.8098282061526167, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9536 + }, + { + "epoch": 3.8102277267279265, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9537 + }, + { + "epoch": 3.810627247303236, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9538 + }, + { + "epoch": 3.8110267678785457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9539 + }, + { + "epoch": 3.8114262884538554, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9540 + }, + { + "epoch": 3.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9541 + }, + { + "epoch": 3.8122253296044746, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9542 + }, + { + "epoch": 3.8126248501797844, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 9543 + }, + { + "epoch": 3.8130243707550937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9544 + }, + { + "epoch": 3.8134238913304035, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9545 + }, + { + "epoch": 3.8138234119057133, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9546 + }, + { + "epoch": 3.8142229324810226, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9547 + }, + { + "epoch": 3.8146224530563324, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9548 + }, + { + "epoch": 3.815021973631642, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9549 + }, + { + "epoch": 3.8154214942069515, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9550 + }, + { + "epoch": 3.8158210147822613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9551 + }, + { + "epoch": 3.816220535357571, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9552 + }, + { + "epoch": 3.8166200559328805, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9553 + }, + { + "epoch": 3.8170195765081902, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 9554 + }, + { + "epoch": 3.8174190970834996, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9555 + }, + { + "epoch": 3.8178186176588094, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9556 + }, + { + "epoch": 3.818218138234119, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9557 + }, + { + "epoch": 3.818617658809429, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9558 + }, + { + "epoch": 3.8190171793847383, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9559 + }, + { + "epoch": 3.819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9560 + }, + { + "epoch": 3.8198162205353574, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9561 + }, + { + "epoch": 3.820215741110667, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9562 + }, + { + "epoch": 3.820615261685977, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9563 + }, + { + "epoch": 3.8210147822612863, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9564 + }, + { + "epoch": 3.821414302836596, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9565 + }, + { + "epoch": 3.8218138234119055, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9566 + }, + { + "epoch": 3.8222133439872152, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9567 + }, + { + "epoch": 3.822612864562525, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9568 + }, + { + "epoch": 3.823012385137835, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 9569 + }, + { + "epoch": 3.823411905713144, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9570 + }, + { + "epoch": 3.823811426288454, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9571 + }, + { + "epoch": 3.8242109468637633, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9572 + }, + { + "epoch": 3.824610467439073, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9573 + }, + { + "epoch": 3.825009988014383, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9574 + }, + { + "epoch": 3.8254095085896926, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9575 + }, + { + "epoch": 3.825809029165002, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9576 + }, + { + "epoch": 3.8262085497403118, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 9577 + }, + { + "epoch": 3.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9578 + }, + { + "epoch": 3.827007590890931, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9579 + }, + { + "epoch": 3.8274071114662407, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9580 + }, + { + "epoch": 3.82780663204155, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9581 + }, + { + "epoch": 3.82820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9582 + }, + { + "epoch": 3.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9583 + }, + { + "epoch": 3.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 9584 + }, + { + "epoch": 3.8294047143427887, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9585 + }, + { + "epoch": 3.8298042349180985, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9586 + }, + { + "epoch": 3.830203755493408, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9587 + }, + { + "epoch": 3.8306032760687176, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9588 + }, + { + "epoch": 3.831002796644027, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9589 + }, + { + "epoch": 3.8314023172193368, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9590 + }, + { + "epoch": 3.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9591 + }, + { + "epoch": 3.832201358369956, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9592 + }, + { + "epoch": 3.8326008789452657, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9593 + }, + { + "epoch": 3.8330003995205755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9594 + }, + { + "epoch": 3.833399920095885, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9595 + }, + { + "epoch": 3.8337994406711946, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9596 + }, + { + "epoch": 3.8341989612465044, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9597 + }, + { + "epoch": 3.8345984818218137, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 9598 + }, + { + "epoch": 3.8349980023971235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9599 + }, + { + "epoch": 3.835397522972433, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9600 + }, + { + "epoch": 3.8357970435477426, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9601 + }, + { + "epoch": 3.8361965641230524, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9602 + }, + { + "epoch": 3.836596084698362, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9603 + }, + { + "epoch": 3.8369956052736716, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9604 + }, + { + "epoch": 3.8373951258489813, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9605 + }, + { + "epoch": 3.8377946464242907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9606 + }, + { + "epoch": 3.8381941669996005, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9607 + }, + { + "epoch": 3.8385936875749103, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9608 + }, + { + "epoch": 3.8389932081502196, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9609 + }, + { + "epoch": 3.8393927287255294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9610 + }, + { + "epoch": 3.8397922493008387, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9611 + }, + { + "epoch": 3.8401917698761485, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9612 + }, + { + "epoch": 3.8405912904514583, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9613 + }, + { + "epoch": 3.840990811026768, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9614 + }, + { + "epoch": 3.8413903316020774, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9615 + }, + { + "epoch": 3.841789852177387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9616 + }, + { + "epoch": 3.8421893727526966, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9617 + }, + { + "epoch": 3.8425888933280063, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9618 + }, + { + "epoch": 3.842988413903316, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 9619 + }, + { + "epoch": 3.843387934478626, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9620 + }, + { + "epoch": 3.8437874550539353, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9621 + }, + { + "epoch": 3.844186975629245, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9622 + }, + { + "epoch": 3.8445864962045544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9623 + }, + { + "epoch": 3.844986016779864, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 9624 + }, + { + "epoch": 3.845385537355174, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9625 + }, + { + "epoch": 3.8457850579304833, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 9626 + }, + { + "epoch": 3.846184578505793, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9627 + }, + { + "epoch": 3.8465840990811024, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9628 + }, + { + "epoch": 3.8469836196564122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9629 + }, + { + "epoch": 3.847383140231722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9630 + }, + { + "epoch": 3.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9631 + }, + { + "epoch": 3.848182181382341, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9632 + }, + { + "epoch": 3.848581701957651, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9633 + }, + { + "epoch": 3.8489812225329603, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 9634 + }, + { + "epoch": 3.84938074310827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9635 + }, + { + "epoch": 3.84978026368358, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 9636 + }, + { + "epoch": 3.850179784258889, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9637 + }, + { + "epoch": 3.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9638 + }, + { + "epoch": 3.8509788254095088, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9639 + }, + { + "epoch": 3.851378345984818, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9640 + }, + { + "epoch": 3.851777866560128, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9641 + }, + { + "epoch": 3.8521773871354377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 9642 + }, + { + "epoch": 3.852576907710747, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9643 + }, + { + "epoch": 3.852976428286057, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9644 + }, + { + "epoch": 3.853375948861366, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9645 + }, + { + "epoch": 3.853775469436676, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9646 + }, + { + "epoch": 3.8541749900119857, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9647 + }, + { + "epoch": 3.8545745105872955, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9648 + }, + { + "epoch": 3.854974031162605, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9649 + }, + { + "epoch": 3.8553735517379146, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9650 + }, + { + "epoch": 3.855773072313224, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9651 + }, + { + "epoch": 3.8561725928885338, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9652 + }, + { + "epoch": 3.8565721134638435, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9653 + }, + { + "epoch": 3.856971634039153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9654 + }, + { + "epoch": 3.8573711546144627, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9655 + }, + { + "epoch": 3.857770675189772, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9656 + }, + { + "epoch": 3.858170195765082, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9657 + }, + { + "epoch": 3.8585697163403916, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 9658 + }, + { + "epoch": 3.8589692369157014, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9659 + }, + { + "epoch": 3.8593687574910107, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9660 + }, + { + "epoch": 3.8597682780663205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9661 + }, + { + "epoch": 3.86016779864163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9662 + }, + { + "epoch": 3.8605673192169396, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9663 + }, + { + "epoch": 3.8609668397922494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9664 + }, + { + "epoch": 3.861366360367559, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9665 + }, + { + "epoch": 3.8617658809428685, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9666 + }, + { + "epoch": 3.8621654015181783, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9667 + }, + { + "epoch": 3.8625649220934877, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9668 + }, + { + "epoch": 3.8629644426687975, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9669 + }, + { + "epoch": 3.8633639632441072, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9670 + }, + { + "epoch": 3.8637634838194166, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9671 + }, + { + "epoch": 3.8641630043947264, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9672 + }, + { + "epoch": 3.8645625249700357, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9673 + }, + { + "epoch": 3.8649620455453455, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9674 + }, + { + "epoch": 3.8653615661206553, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9675 + }, + { + "epoch": 3.865761086695965, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 9676 + }, + { + "epoch": 3.8661606072712744, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 9677 + }, + { + "epoch": 3.866560127846584, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9678 + }, + { + "epoch": 3.8669596484218935, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9679 + }, + { + "epoch": 3.8673591689972033, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9680 + }, + { + "epoch": 3.867758689572513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 9681 + }, + { + "epoch": 3.8681582101478225, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9682 + }, + { + "epoch": 3.8685577307231322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9683 + }, + { + "epoch": 3.868957251298442, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9684 + }, + { + "epoch": 3.8693567718737514, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9685 + }, + { + "epoch": 3.869756292449061, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9686 + }, + { + "epoch": 3.870155813024371, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9687 + }, + { + "epoch": 3.8705553335996803, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 9688 + }, + { + "epoch": 3.87095485417499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9689 + }, + { + "epoch": 3.8713543747502994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9690 + }, + { + "epoch": 3.871753895325609, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9691 + }, + { + "epoch": 3.872153415900919, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9692 + }, + { + "epoch": 3.872552936476229, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9693 + }, + { + "epoch": 3.872952457051538, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9694 + }, + { + "epoch": 3.873351977626848, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9695 + }, + { + "epoch": 3.8737514982021573, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9696 + }, + { + "epoch": 3.874151018777467, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9697 + }, + { + "epoch": 3.874550539352777, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 9698 + }, + { + "epoch": 3.874950059928086, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9699 + }, + { + "epoch": 3.875349580503396, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9700 + }, + { + "epoch": 3.8757491010787053, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9701 + }, + { + "epoch": 3.876148621654015, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9702 + }, + { + "epoch": 3.876548142229325, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9703 + }, + { + "epoch": 3.8769476628046347, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9704 + }, + { + "epoch": 3.877347183379944, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9705 + }, + { + "epoch": 3.877746703955254, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9706 + }, + { + "epoch": 3.878146224530563, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9707 + }, + { + "epoch": 3.878545745105873, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9708 + }, + { + "epoch": 3.8789452656811827, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9709 + }, + { + "epoch": 3.8793447862564925, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9710 + }, + { + "epoch": 3.879744306831802, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9711 + }, + { + "epoch": 3.8801438274071116, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9712 + }, + { + "epoch": 3.880543347982421, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9713 + }, + { + "epoch": 3.8809428685577307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9714 + }, + { + "epoch": 3.8813423891330405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9715 + }, + { + "epoch": 3.88174190970835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9716 + }, + { + "epoch": 3.8821414302836597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9717 + }, + { + "epoch": 3.882540950858969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9718 + }, + { + "epoch": 3.882940471434279, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9719 + }, + { + "epoch": 3.8833399920095886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9720 + }, + { + "epoch": 3.8837395125848984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9721 + }, + { + "epoch": 3.8841390331602077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9722 + }, + { + "epoch": 3.8845385537355175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9723 + }, + { + "epoch": 3.884938074310827, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9724 + }, + { + "epoch": 3.8853375948861366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9725 + }, + { + "epoch": 3.8857371154614464, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9726 + }, + { + "epoch": 3.8861366360367557, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9727 + }, + { + "epoch": 3.8865361566120655, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9728 + }, + { + "epoch": 3.8869356771873753, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9729 + }, + { + "epoch": 3.8873351977626847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9730 + }, + { + "epoch": 3.8877347183379944, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9731 + }, + { + "epoch": 3.8881342389133042, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9732 + }, + { + "epoch": 3.8885337594886136, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9733 + }, + { + "epoch": 3.8889332800639234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9734 + }, + { + "epoch": 3.8893328006392327, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9735 + }, + { + "epoch": 3.8897323212145425, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9736 + }, + { + "epoch": 3.8901318417898523, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9737 + }, + { + "epoch": 3.890531362365162, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9738 + }, + { + "epoch": 3.8909308829404714, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9739 + }, + { + "epoch": 3.891330403515781, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9740 + }, + { + "epoch": 3.8917299240910905, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9741 + }, + { + "epoch": 3.8921294446664003, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9742 + }, + { + "epoch": 3.89252896524171, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9743 + }, + { + "epoch": 3.8929284858170194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9744 + }, + { + "epoch": 3.8933280063923292, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 9745 + }, + { + "epoch": 3.8937275269676386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9746 + }, + { + "epoch": 3.8941270475429484, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9747 + }, + { + "epoch": 3.894526568118258, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9748 + }, + { + "epoch": 3.894926088693568, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9749 + }, + { + "epoch": 3.8953256092688773, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9750 + }, + { + "epoch": 3.895725129844187, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9751 + }, + { + "epoch": 3.8961246504194964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9752 + }, + { + "epoch": 3.896524170994806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9753 + }, + { + "epoch": 3.896923691570116, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9754 + }, + { + "epoch": 3.8973232121454258, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9755 + }, + { + "epoch": 3.897722732720735, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9756 + }, + { + "epoch": 3.898122253296045, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9757 + }, + { + "epoch": 3.8985217738713542, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9758 + }, + { + "epoch": 3.898921294446664, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9759 + }, + { + "epoch": 3.899320815021974, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9760 + }, + { + "epoch": 3.899720335597283, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 9761 + }, + { + "epoch": 3.900119856172593, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9762 + }, + { + "epoch": 3.9005193767479023, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9763 + }, + { + "epoch": 3.900918897323212, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9764 + }, + { + "epoch": 3.901318417898522, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9765 + }, + { + "epoch": 3.9017179384738316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9766 + }, + { + "epoch": 3.902117459049141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9767 + }, + { + "epoch": 3.9025169796244508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9768 + }, + { + "epoch": 3.90291650019976, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9769 + }, + { + "epoch": 3.90331602077507, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9770 + }, + { + "epoch": 3.9037155413503797, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9771 + }, + { + "epoch": 3.904115061925689, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9772 + }, + { + "epoch": 3.904514582500999, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9773 + }, + { + "epoch": 3.9049141030763086, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9774 + }, + { + "epoch": 3.905313623651618, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9775 + }, + { + "epoch": 3.9057131442269277, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9776 + }, + { + "epoch": 3.9061126648022375, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9777 + }, + { + "epoch": 3.906512185377547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9778 + }, + { + "epoch": 3.9069117059528566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9779 + }, + { + "epoch": 3.907311226528166, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9780 + }, + { + "epoch": 3.9077107471034758, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9781 + }, + { + "epoch": 3.9081102676787856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9782 + }, + { + "epoch": 3.9085097882540953, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9783 + }, + { + "epoch": 3.9089093088294047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9784 + }, + { + "epoch": 3.9093088294047145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9785 + }, + { + "epoch": 3.909708349980024, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9786 + }, + { + "epoch": 3.9101078705553336, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9787 + }, + { + "epoch": 3.9105073911306434, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9788 + }, + { + "epoch": 3.9109069117059527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9789 + }, + { + "epoch": 3.9113064322812625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9790 + }, + { + "epoch": 3.911705952856572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9791 + }, + { + "epoch": 3.9121054734318816, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9792 + }, + { + "epoch": 3.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9793 + }, + { + "epoch": 3.912904514582501, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9794 + }, + { + "epoch": 3.9133040351578106, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9795 + }, + { + "epoch": 3.9137035557331203, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9796 + }, + { + "epoch": 3.9141030763084297, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 9797 + }, + { + "epoch": 3.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9798 + }, + { + "epoch": 3.9149021174590493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9799 + }, + { + "epoch": 3.915301638034359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9800 + }, + { + "epoch": 3.9157011586096684, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9801 + }, + { + "epoch": 3.916100679184978, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9802 + }, + { + "epoch": 3.9165001997602875, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9803 + }, + { + "epoch": 3.9168997203355973, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9804 + }, + { + "epoch": 3.917299240910907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9805 + }, + { + "epoch": 3.9176987614862164, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9806 + }, + { + "epoch": 3.918098282061526, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9807 + }, + { + "epoch": 3.9184978026368356, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 9808 + }, + { + "epoch": 3.9188973232121453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 9809 + }, + { + "epoch": 3.919296843787455, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9810 + }, + { + "epoch": 3.919696364362765, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9811 + }, + { + "epoch": 3.9200958849380743, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9812 + }, + { + "epoch": 3.920495405513384, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9813 + }, + { + "epoch": 3.9208949260886934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9814 + }, + { + "epoch": 3.921294446664003, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9815 + }, + { + "epoch": 3.921693967239313, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9816 + }, + { + "epoch": 3.9220934878146223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9817 + }, + { + "epoch": 3.922493008389932, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9818 + }, + { + "epoch": 3.922892528965242, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9819 + }, + { + "epoch": 3.923292049540551, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9820 + }, + { + "epoch": 3.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9821 + }, + { + "epoch": 3.924091090691171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9822 + }, + { + "epoch": 3.92449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9823 + }, + { + "epoch": 3.92489013184179, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9824 + }, + { + "epoch": 3.9252896524170993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9825 + }, + { + "epoch": 3.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9826 + }, + { + "epoch": 3.926088693567719, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9827 + }, + { + "epoch": 3.9264882141430286, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9828 + }, + { + "epoch": 3.926887734718338, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9829 + }, + { + "epoch": 3.9272872552936477, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9830 + }, + { + "epoch": 3.927686775868957, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9831 + }, + { + "epoch": 3.928086296444267, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 9832 + }, + { + "epoch": 3.9284858170195767, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9833 + }, + { + "epoch": 3.928885337594886, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9834 + }, + { + "epoch": 3.929284858170196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9835 + }, + { + "epoch": 3.929684378745505, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9836 + }, + { + "epoch": 3.930083899320815, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9837 + }, + { + "epoch": 3.9304834198961247, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9838 + }, + { + "epoch": 3.9308829404714345, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9839 + }, + { + "epoch": 3.931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9840 + }, + { + "epoch": 3.9316819816220536, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9841 + }, + { + "epoch": 3.932081502197363, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9842 + }, + { + "epoch": 3.9324810227726728, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9843 + }, + { + "epoch": 3.9328805433479825, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9844 + }, + { + "epoch": 3.9332800639232923, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9845 + }, + { + "epoch": 3.9336795844986017, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9846 + }, + { + "epoch": 3.9340791050739115, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9847 + }, + { + "epoch": 3.934478625649221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9848 + }, + { + "epoch": 3.9348781462245306, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9849 + }, + { + "epoch": 3.9352776667998404, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9850 + }, + { + "epoch": 3.9356771873751497, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9851 + }, + { + "epoch": 3.9360767079504595, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9852 + }, + { + "epoch": 3.936476228525769, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9853 + }, + { + "epoch": 3.9368757491010786, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9854 + }, + { + "epoch": 3.9372752696763884, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9855 + }, + { + "epoch": 3.937674790251698, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9856 + }, + { + "epoch": 3.9380743108270075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9857 + }, + { + "epoch": 3.9384738314023173, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9858 + }, + { + "epoch": 3.9388733519776267, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9859 + }, + { + "epoch": 3.9392728725529365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9860 + }, + { + "epoch": 3.9396723931282462, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9861 + }, + { + "epoch": 3.9400719137035556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9862 + }, + { + "epoch": 3.9404714342788654, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9863 + }, + { + "epoch": 3.9408709548541747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9864 + }, + { + "epoch": 3.9412704754294845, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9865 + }, + { + "epoch": 3.9416699960047943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9866 + }, + { + "epoch": 3.942069516580104, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9867 + }, + { + "epoch": 3.9424690371554134, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9868 + }, + { + "epoch": 3.942868557730723, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9869 + }, + { + "epoch": 3.9432680783060325, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9870 + }, + { + "epoch": 3.9436675988813423, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9871 + }, + { + "epoch": 3.944067119456652, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9872 + }, + { + "epoch": 3.944466640031962, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9873 + }, + { + "epoch": 3.9448661606072712, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9874 + }, + { + "epoch": 3.945265681182581, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9875 + }, + { + "epoch": 3.9456652017578904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9876 + }, + { + "epoch": 3.9460647223332, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 9877 + }, + { + "epoch": 3.94646424290851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9878 + }, + { + "epoch": 3.9468637634838193, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9879 + }, + { + "epoch": 3.947263284059129, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9880 + }, + { + "epoch": 3.9476628046344384, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9881 + }, + { + "epoch": 3.948062325209748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9882 + }, + { + "epoch": 3.948461845785058, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9883 + }, + { + "epoch": 3.9488613663603678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9884 + }, + { + "epoch": 3.949260886935677, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9885 + }, + { + "epoch": 3.949660407510987, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9886 + }, + { + "epoch": 3.9500599280862962, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9887 + }, + { + "epoch": 3.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9888 + }, + { + "epoch": 3.950858969236916, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9889 + }, + { + "epoch": 3.9512584898122256, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9890 + }, + { + "epoch": 3.951658010387535, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9891 + }, + { + "epoch": 3.9520575309628447, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9892 + }, + { + "epoch": 3.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9893 + }, + { + "epoch": 3.952856572113464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9894 + }, + { + "epoch": 3.9532560926887736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9895 + }, + { + "epoch": 3.953655613264083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9896 + }, + { + "epoch": 3.9540551338393928, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9897 + }, + { + "epoch": 3.954454654414702, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9898 + }, + { + "epoch": 3.954854174990012, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9899 + }, + { + "epoch": 3.9552536955653217, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9900 + }, + { + "epoch": 3.9556532161406315, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9901 + }, + { + "epoch": 3.956052736715941, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9902 + }, + { + "epoch": 3.9564522572912506, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9903 + }, + { + "epoch": 3.95685177786656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9904 + }, + { + "epoch": 3.9572512984418697, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9905 + }, + { + "epoch": 3.9576508190171795, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9906 + }, + { + "epoch": 3.958050339592489, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9907 + }, + { + "epoch": 3.9584498601677987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9908 + }, + { + "epoch": 3.958849380743108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9909 + }, + { + "epoch": 3.959248901318418, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9910 + }, + { + "epoch": 3.9596484218937276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 9911 + }, + { + "epoch": 3.9600479424690374, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9912 + }, + { + "epoch": 3.9604474630443467, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9913 + }, + { + "epoch": 3.9608469836196565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9914 + }, + { + "epoch": 3.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9915 + }, + { + "epoch": 3.9616460247702756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9916 + }, + { + "epoch": 3.9620455453455854, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9917 + }, + { + "epoch": 3.962445065920895, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9918 + }, + { + "epoch": 3.9628445864962045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9919 + }, + { + "epoch": 3.9632441070715143, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9920 + }, + { + "epoch": 3.9636436276468237, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9921 + }, + { + "epoch": 3.9640431482221334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9922 + }, + { + "epoch": 3.9644426687974432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9923 + }, + { + "epoch": 3.9648421893727526, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9924 + }, + { + "epoch": 3.9652417099480624, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9925 + }, + { + "epoch": 3.9656412305233717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 9926 + }, + { + "epoch": 3.9660407510986815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9927 + }, + { + "epoch": 3.9664402716739913, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9928 + }, + { + "epoch": 3.966839792249301, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9929 + }, + { + "epoch": 3.9672393128246104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9930 + }, + { + "epoch": 3.96763883339992, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9931 + }, + { + "epoch": 3.9680383539752295, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9932 + }, + { + "epoch": 3.9684378745505393, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9933 + }, + { + "epoch": 3.968837395125849, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 9934 + }, + { + "epoch": 3.969236915701159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 9935 + }, + { + "epoch": 3.9696364362764682, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9936 + }, + { + "epoch": 3.970035956851778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 9937 + }, + { + "epoch": 3.9704354774270874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9938 + }, + { + "epoch": 3.970834998002397, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9939 + }, + { + "epoch": 3.971234518577707, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9940 + }, + { + "epoch": 3.9716340391530163, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9941 + }, + { + "epoch": 3.972033559728326, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9942 + }, + { + "epoch": 3.9724330803036354, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9943 + }, + { + "epoch": 3.972832600878945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 9944 + }, + { + "epoch": 3.973232121454255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9945 + }, + { + "epoch": 3.9736316420295648, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9946 + }, + { + "epoch": 3.974031162604874, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9947 + }, + { + "epoch": 3.974430683180184, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9948 + }, + { + "epoch": 3.9748302037554932, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9949 + }, + { + "epoch": 3.975229724330803, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9950 + }, + { + "epoch": 3.975629244906113, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9951 + }, + { + "epoch": 3.976028765481422, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9952 + }, + { + "epoch": 3.976428286056732, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9953 + }, + { + "epoch": 3.9768278066320413, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9954 + }, + { + "epoch": 3.977227327207351, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9955 + }, + { + "epoch": 3.977626847782661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 9956 + }, + { + "epoch": 3.9780263683579706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9957 + }, + { + "epoch": 3.97842588893328, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9958 + }, + { + "epoch": 3.9788254095085898, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9959 + }, + { + "epoch": 3.979224930083899, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9960 + }, + { + "epoch": 3.979624450659209, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9961 + }, + { + "epoch": 3.9800239712345187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9962 + }, + { + "epoch": 3.9804234918098285, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9963 + }, + { + "epoch": 3.980823012385138, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 9964 + }, + { + "epoch": 3.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9965 + }, + { + "epoch": 3.981622053535757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9966 + }, + { + "epoch": 3.9820215741110667, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9967 + }, + { + "epoch": 3.9824210946863765, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9968 + }, + { + "epoch": 3.982820615261686, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9969 + }, + { + "epoch": 3.9832201358369956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9970 + }, + { + "epoch": 3.983619656412305, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9971 + }, + { + "epoch": 3.9840191769876148, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9972 + }, + { + "epoch": 3.9844186975629245, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9973 + }, + { + "epoch": 3.9848182181382343, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9974 + }, + { + "epoch": 3.9852177387135437, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9975 + }, + { + "epoch": 3.9856172592888535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9976 + }, + { + "epoch": 3.986016779864163, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9977 + }, + { + "epoch": 3.9864163004394726, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9978 + }, + { + "epoch": 3.9868158210147824, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9979 + }, + { + "epoch": 3.987215341590092, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9980 + }, + { + "epoch": 3.9876148621654015, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9981 + }, + { + "epoch": 3.9880143827407113, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9982 + }, + { + "epoch": 3.9884139033160206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9983 + }, + { + "epoch": 3.9888134238913304, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9984 + }, + { + "epoch": 3.98921294446664, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9985 + }, + { + "epoch": 3.9896124650419496, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9986 + }, + { + "epoch": 3.9900119856172593, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 9987 + }, + { + "epoch": 3.9904115061925687, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9988 + }, + { + "epoch": 3.9908110267678785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9989 + }, + { + "epoch": 3.9912105473431883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9990 + }, + { + "epoch": 3.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9991 + }, + { + "epoch": 3.9920095884938074, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9992 + }, + { + "epoch": 3.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9993 + }, + { + "epoch": 3.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9994 + }, + { + "epoch": 3.9932081502197363, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9995 + }, + { + "epoch": 3.993607670795046, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9996 + }, + { + "epoch": 3.9940071913703554, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9997 + }, + { + "epoch": 3.994406711945665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9998 + }, + { + "epoch": 3.9948062325209746, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9999 + }, + { + "epoch": 3.9952057530962843, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10000 + }, + { + "epoch": 3.995605273671594, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10001 + }, + { + "epoch": 3.996004794246904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10002 + }, + { + "epoch": 3.9964043148222133, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10003 + }, + { + "epoch": 3.996803835397523, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10004 + }, + { + "epoch": 3.9972033559728324, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10005 + }, + { + "epoch": 3.997602876548142, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10006 + }, + { + "epoch": 3.998002397123452, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10007 + }, + { + "epoch": 3.9984019176987617, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10008 + }, + { + "epoch": 3.998801438274071, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10009 + }, + { + "epoch": 3.999200958849381, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 10010 + }, + { + "epoch": 3.99960047942469, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 10011 + }, + { + "epoch": 4.0, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10012 + }, + { + "epoch": 4.00039952057531, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10013 + }, + { + "epoch": 4.00079904115062, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10014 + }, + { + "epoch": 4.0011985617259285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10015 + }, + { + "epoch": 4.001598082301238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10016 + }, + { + "epoch": 4.001997602876548, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10017 + }, + { + "epoch": 4.002397123451858, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10018 + }, + { + "epoch": 4.002796644027168, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10019 + }, + { + "epoch": 4.003196164602477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10020 + }, + { + "epoch": 4.003595685177786, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10021 + }, + { + "epoch": 4.003995205753096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10022 + }, + { + "epoch": 4.004394726328406, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10023 + }, + { + "epoch": 4.004794246903716, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10024 + }, + { + "epoch": 4.0051937674790254, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10025 + }, + { + "epoch": 4.005593288054335, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10026 + }, + { + "epoch": 4.005992808629644, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10027 + }, + { + "epoch": 4.006392329204954, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10028 + }, + { + "epoch": 4.006791849780264, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10029 + }, + { + "epoch": 4.0071913703555735, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10030 + }, + { + "epoch": 4.007590890930883, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10031 + }, + { + "epoch": 4.007990411506192, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10032 + }, + { + "epoch": 4.008389932081502, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10033 + }, + { + "epoch": 4.008789452656812, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10034 + }, + { + "epoch": 4.0091889732321215, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10035 + }, + { + "epoch": 4.009588493807431, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 10036 + }, + { + "epoch": 4.009988014382741, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10037 + }, + { + "epoch": 4.01038753495805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 10038 + }, + { + "epoch": 4.01078705553336, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10039 + }, + { + "epoch": 4.01118657610867, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10040 + }, + { + "epoch": 4.011586096683979, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10041 + }, + { + "epoch": 4.011985617259289, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10042 + }, + { + "epoch": 4.012385137834598, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10043 + }, + { + "epoch": 4.012784658409908, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10044 + }, + { + "epoch": 4.013184178985218, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10045 + }, + { + "epoch": 4.013583699560527, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10046 + }, + { + "epoch": 4.013983220135837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10047 + }, + { + "epoch": 4.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10048 + }, + { + "epoch": 4.014782261286456, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 10049 + }, + { + "epoch": 4.015181781861766, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10050 + }, + { + "epoch": 4.0155813024370755, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10051 + }, + { + "epoch": 4.015980823012385, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10052 + }, + { + "epoch": 4.016380343587695, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10053 + }, + { + "epoch": 4.016779864163005, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10054 + }, + { + "epoch": 4.017179384738314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10055 + }, + { + "epoch": 4.0175789053136235, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10056 + }, + { + "epoch": 4.017978425888933, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10057 + }, + { + "epoch": 4.018377946464243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10058 + }, + { + "epoch": 4.018777467039553, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10059 + }, + { + "epoch": 4.019176987614862, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 10060 + }, + { + "epoch": 4.0195765081901715, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10061 + }, + { + "epoch": 4.019976028765481, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10062 + }, + { + "epoch": 4.020375549340791, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10063 + }, + { + "epoch": 4.020775069916101, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10064 + }, + { + "epoch": 4.021174590491411, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10065 + }, + { + "epoch": 4.02157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10066 + }, + { + "epoch": 4.021973631642029, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10067 + }, + { + "epoch": 4.022373152217339, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10068 + }, + { + "epoch": 4.022772672792649, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10069 + }, + { + "epoch": 4.023172193367959, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10070 + }, + { + "epoch": 4.0235717139432685, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10071 + }, + { + "epoch": 4.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10072 + }, + { + "epoch": 4.024370755093887, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10073 + }, + { + "epoch": 4.024770275669197, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10074 + }, + { + "epoch": 4.025169796244507, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10075 + }, + { + "epoch": 4.025569316819817, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10076 + }, + { + "epoch": 4.0259688373951255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10077 + }, + { + "epoch": 4.026368357970435, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10078 + }, + { + "epoch": 4.026767878545745, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 10079 + }, + { + "epoch": 4.027167399121055, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10080 + }, + { + "epoch": 4.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10081 + }, + { + "epoch": 4.027966440271674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10082 + }, + { + "epoch": 4.028365960846983, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10083 + }, + { + "epoch": 4.028765481422293, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10084 + }, + { + "epoch": 4.029165001997603, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10085 + }, + { + "epoch": 4.029564522572913, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10086 + }, + { + "epoch": 4.029964043148222, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10087 + }, + { + "epoch": 4.030363563723531, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10088 + }, + { + "epoch": 4.030763084298841, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10089 + }, + { + "epoch": 4.031162604874151, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10090 + }, + { + "epoch": 4.031562125449461, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 10091 + }, + { + "epoch": 4.0319616460247705, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10092 + }, + { + "epoch": 4.03236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10093 + }, + { + "epoch": 4.032760687175389, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10094 + }, + { + "epoch": 4.033160207750699, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10095 + }, + { + "epoch": 4.033559728326009, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10096 + }, + { + "epoch": 4.0339592489013185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10097 + }, + { + "epoch": 4.034358769476628, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10098 + }, + { + "epoch": 4.034758290051938, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10099 + }, + { + "epoch": 4.035157810627247, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10100 + }, + { + "epoch": 4.035557331202557, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10101 + }, + { + "epoch": 4.035956851777867, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10102 + }, + { + "epoch": 4.036356372353176, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10103 + }, + { + "epoch": 4.036755892928486, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10104 + }, + { + "epoch": 4.037155413503795, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10105 + }, + { + "epoch": 4.037554934079105, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10106 + }, + { + "epoch": 4.037954454654415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 10107 + }, + { + "epoch": 4.038353975229724, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10108 + }, + { + "epoch": 4.038753495805034, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10109 + }, + { + "epoch": 4.039153016380344, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10110 + }, + { + "epoch": 4.039552536955653, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10111 + }, + { + "epoch": 4.039952057530963, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10112 + }, + { + "epoch": 4.040351578106272, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10113 + }, + { + "epoch": 4.040751098681582, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10114 + }, + { + "epoch": 4.041150619256892, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10115 + }, + { + "epoch": 4.041550139832202, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10116 + }, + { + "epoch": 4.041949660407511, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10117 + }, + { + "epoch": 4.0423491809828205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10118 + }, + { + "epoch": 4.04274870155813, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10119 + }, + { + "epoch": 4.04314822213344, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10120 + }, + { + "epoch": 4.04354774270875, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10121 + }, + { + "epoch": 4.043947263284059, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10122 + }, + { + "epoch": 4.0443467838593685, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10123 + }, + { + "epoch": 4.044746304434678, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10124 + }, + { + "epoch": 4.045145825009988, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10125 + }, + { + "epoch": 4.045545345585298, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10126 + }, + { + "epoch": 4.045944866160608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10127 + }, + { + "epoch": 4.046344386735917, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10128 + }, + { + "epoch": 4.046743907311226, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10129 + }, + { + "epoch": 4.047143427886536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10130 + }, + { + "epoch": 4.047542948461846, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10131 + }, + { + "epoch": 4.047942469037156, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10132 + }, + { + "epoch": 4.048341989612465, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10133 + }, + { + "epoch": 4.048741510187774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10134 + }, + { + "epoch": 4.049141030763084, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10135 + }, + { + "epoch": 4.049540551338394, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10136 + }, + { + "epoch": 4.049940071913704, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10137 + }, + { + "epoch": 4.0503395924890135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10138 + }, + { + "epoch": 4.050739113064322, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10139 + }, + { + "epoch": 4.051138633639632, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 10140 + }, + { + "epoch": 4.051538154214942, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10141 + }, + { + "epoch": 4.051937674790252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10142 + }, + { + "epoch": 4.052337195365562, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10143 + }, + { + "epoch": 4.052736715940871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10144 + }, + { + "epoch": 4.05313623651618, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10145 + }, + { + "epoch": 4.05353575709149, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10146 + }, + { + "epoch": 4.0539352776668, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10147 + }, + { + "epoch": 4.05433479824211, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10148 + }, + { + "epoch": 4.054734318817419, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10149 + }, + { + "epoch": 4.055133839392728, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10150 + }, + { + "epoch": 4.055533359968038, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10151 + }, + { + "epoch": 4.055932880543348, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10152 + }, + { + "epoch": 4.056332401118658, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10153 + }, + { + "epoch": 4.0567319216939675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10154 + }, + { + "epoch": 4.057131442269277, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10155 + }, + { + "epoch": 4.057530962844586, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10156 + }, + { + "epoch": 4.057930483419896, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10157 + }, + { + "epoch": 4.058330003995206, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10158 + }, + { + "epoch": 4.0587295245705155, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10159 + }, + { + "epoch": 4.059129045145825, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10160 + }, + { + "epoch": 4.059528565721135, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10161 + }, + { + "epoch": 4.059928086296444, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10162 + }, + { + "epoch": 4.060327606871754, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10163 + }, + { + "epoch": 4.0607271274470635, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10164 + }, + { + "epoch": 4.061126648022373, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10165 + }, + { + "epoch": 4.061526168597683, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10166 + }, + { + "epoch": 4.061925689172992, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10167 + }, + { + "epoch": 4.062325209748302, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10168 + }, + { + "epoch": 4.062724730323612, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10169 + }, + { + "epoch": 4.063124250898921, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10170 + }, + { + "epoch": 4.063523771474231, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10171 + }, + { + "epoch": 4.063923292049541, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10172 + }, + { + "epoch": 4.06432281262485, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10173 + }, + { + "epoch": 4.06472233320016, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10174 + }, + { + "epoch": 4.065121853775469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10175 + }, + { + "epoch": 4.065521374350779, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10176 + }, + { + "epoch": 4.065920894926089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10177 + }, + { + "epoch": 4.066320415501398, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10178 + }, + { + "epoch": 4.066719936076708, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10179 + }, + { + "epoch": 4.0671194566520175, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10180 + }, + { + "epoch": 4.067518977227327, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10181 + }, + { + "epoch": 4.067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10182 + }, + { + "epoch": 4.068318018377947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10183 + }, + { + "epoch": 4.068717538953256, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10184 + }, + { + "epoch": 4.0691170595285655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10185 + }, + { + "epoch": 4.069516580103875, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10186 + }, + { + "epoch": 4.069916100679185, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10187 + }, + { + "epoch": 4.070315621254495, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10188 + }, + { + "epoch": 4.070715141829805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10189 + }, + { + "epoch": 4.0711146624051135, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10190 + }, + { + "epoch": 4.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10191 + }, + { + "epoch": 4.071913703555733, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10192 + }, + { + "epoch": 4.072313224131043, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10193 + }, + { + "epoch": 4.072712744706353, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10194 + }, + { + "epoch": 4.073112265281662, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10195 + }, + { + "epoch": 4.073511785856971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10196 + }, + { + "epoch": 4.073911306432281, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10197 + }, + { + "epoch": 4.074310827007591, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10198 + }, + { + "epoch": 4.074710347582901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10199 + }, + { + "epoch": 4.0751098681582105, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10200 + }, + { + "epoch": 4.075509388733519, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10201 + }, + { + "epoch": 4.075908909308829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 10202 + }, + { + "epoch": 4.076308429884139, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10203 + }, + { + "epoch": 4.076707950459449, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 10204 + }, + { + "epoch": 4.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10205 + }, + { + "epoch": 4.077506991610068, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10206 + }, + { + "epoch": 4.077906512185377, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10207 + }, + { + "epoch": 4.078306032760687, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10208 + }, + { + "epoch": 4.078705553335997, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10209 + }, + { + "epoch": 4.079105073911307, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10210 + }, + { + "epoch": 4.079504594486616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10211 + }, + { + "epoch": 4.079904115061925, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10212 + }, + { + "epoch": 4.080303635637235, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10213 + }, + { + "epoch": 4.080703156212545, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10214 + }, + { + "epoch": 4.081102676787855, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10215 + }, + { + "epoch": 4.081502197363164, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10216 + }, + { + "epoch": 4.081901717938474, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10217 + }, + { + "epoch": 4.082301238513783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10218 + }, + { + "epoch": 4.082700759089093, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10219 + }, + { + "epoch": 4.083100279664403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10220 + }, + { + "epoch": 4.0834998002397125, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10221 + }, + { + "epoch": 4.083899320815022, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10222 + }, + { + "epoch": 4.084298841390331, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10223 + }, + { + "epoch": 4.084698361965641, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10224 + }, + { + "epoch": 4.085097882540951, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 10225 + }, + { + "epoch": 4.0854974031162605, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10226 + }, + { + "epoch": 4.08589692369157, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10227 + }, + { + "epoch": 4.08629644426688, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10228 + }, + { + "epoch": 4.086695964842189, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10229 + }, + { + "epoch": 4.087095485417499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10230 + }, + { + "epoch": 4.087495005992809, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10231 + }, + { + "epoch": 4.087894526568118, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10232 + }, + { + "epoch": 4.088294047143428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10233 + }, + { + "epoch": 4.088693567718738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10234 + }, + { + "epoch": 4.089093088294047, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10235 + }, + { + "epoch": 4.089492608869357, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10236 + }, + { + "epoch": 4.089892129444666, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10237 + }, + { + "epoch": 4.090291650019976, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10238 + }, + { + "epoch": 4.090691170595286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10239 + }, + { + "epoch": 4.091090691170595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10240 + }, + { + "epoch": 4.091490211745905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10241 + }, + { + "epoch": 4.0918897323212144, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10242 + }, + { + "epoch": 4.092289252896524, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10243 + }, + { + "epoch": 4.092688773471834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10244 + }, + { + "epoch": 4.093088294047144, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10245 + }, + { + "epoch": 4.093487814622453, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10246 + }, + { + "epoch": 4.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10247 + }, + { + "epoch": 4.094286855773072, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10248 + }, + { + "epoch": 4.094686376348382, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 10249 + }, + { + "epoch": 4.095085896923692, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10250 + }, + { + "epoch": 4.095485417499002, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10251 + }, + { + "epoch": 4.0958849380743105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10252 + }, + { + "epoch": 4.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10253 + }, + { + "epoch": 4.09668397922493, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10254 + }, + { + "epoch": 4.09708349980024, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10255 + }, + { + "epoch": 4.09748302037555, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10256 + }, + { + "epoch": 4.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10257 + }, + { + "epoch": 4.098282061526168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10258 + }, + { + "epoch": 4.098681582101478, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10259 + }, + { + "epoch": 4.099081102676788, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10260 + }, + { + "epoch": 4.099480623252098, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10261 + }, + { + "epoch": 4.0998801438274075, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10262 + }, + { + "epoch": 4.100279664402716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10263 + }, + { + "epoch": 4.100679184978026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10264 + }, + { + "epoch": 4.101078705553336, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10265 + }, + { + "epoch": 4.101478226128646, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10266 + }, + { + "epoch": 4.1018777467039556, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 10267 + }, + { + "epoch": 4.1022772672792644, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10268 + }, + { + "epoch": 4.102676787854574, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10269 + }, + { + "epoch": 4.103076308429884, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10270 + }, + { + "epoch": 4.103475829005194, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10271 + }, + { + "epoch": 4.103875349580504, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10272 + }, + { + "epoch": 4.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 10273 + }, + { + "epoch": 4.104674390731122, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10274 + }, + { + "epoch": 4.105073911306432, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10275 + }, + { + "epoch": 4.105473431881742, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10276 + }, + { + "epoch": 4.105872952457052, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10277 + }, + { + "epoch": 4.106272473032361, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 10278 + }, + { + "epoch": 4.106671993607671, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10279 + }, + { + "epoch": 4.10707151418298, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10280 + }, + { + "epoch": 4.10747103475829, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10281 + }, + { + "epoch": 4.1078705553336, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10282 + }, + { + "epoch": 4.1082700759089095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10283 + }, + { + "epoch": 4.108669596484219, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10284 + }, + { + "epoch": 4.109069117059528, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10285 + }, + { + "epoch": 4.109468637634838, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10286 + }, + { + "epoch": 4.109868158210148, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10287 + }, + { + "epoch": 4.1102676787854575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10288 + }, + { + "epoch": 4.110667199360767, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10289 + }, + { + "epoch": 4.111066719936077, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10290 + }, + { + "epoch": 4.111466240511386, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10291 + }, + { + "epoch": 4.111865761086696, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10292 + }, + { + "epoch": 4.112265281662006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10293 + }, + { + "epoch": 4.112664802237315, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10294 + }, + { + "epoch": 4.113064322812625, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 10295 + }, + { + "epoch": 4.113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10296 + }, + { + "epoch": 4.113863363963244, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10297 + }, + { + "epoch": 4.114262884538554, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10298 + }, + { + "epoch": 4.114662405113863, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10299 + }, + { + "epoch": 4.115061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10300 + }, + { + "epoch": 4.115461446264483, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10301 + }, + { + "epoch": 4.115860966839792, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10302 + }, + { + "epoch": 4.116260487415102, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10303 + }, + { + "epoch": 4.116660007990411, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10304 + }, + { + "epoch": 4.117059528565721, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10305 + }, + { + "epoch": 4.117459049141031, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10306 + }, + { + "epoch": 4.117858569716341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10307 + }, + { + "epoch": 4.11825809029165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10308 + }, + { + "epoch": 4.1186576108669595, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10309 + }, + { + "epoch": 4.119057131442269, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10310 + }, + { + "epoch": 4.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 10311 + }, + { + "epoch": 4.119856172592889, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10312 + }, + { + "epoch": 4.120255693168198, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10313 + }, + { + "epoch": 4.1206552137435075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 10314 + }, + { + "epoch": 4.121054734318817, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10315 + }, + { + "epoch": 4.121454254894127, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10316 + }, + { + "epoch": 4.121853775469437, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10317 + }, + { + "epoch": 4.122253296044747, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10318 + }, + { + "epoch": 4.122652816620056, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10319 + }, + { + "epoch": 4.123052337195365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10320 + }, + { + "epoch": 4.123451857770675, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10321 + }, + { + "epoch": 4.123851378345985, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10322 + }, + { + "epoch": 4.124250898921295, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10323 + }, + { + "epoch": 4.1246504194966045, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10324 + }, + { + "epoch": 4.125049940071913, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10325 + }, + { + "epoch": 4.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10326 + }, + { + "epoch": 4.125848981222533, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10327 + }, + { + "epoch": 4.126248501797843, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 10328 + }, + { + "epoch": 4.1266480223731525, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10329 + }, + { + "epoch": 4.127047542948461, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10330 + }, + { + "epoch": 4.127447063523771, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10331 + }, + { + "epoch": 4.127846584099081, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10332 + }, + { + "epoch": 4.128246104674391, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10333 + }, + { + "epoch": 4.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10334 + }, + { + "epoch": 4.12904514582501, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10335 + }, + { + "epoch": 4.129444666400319, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10336 + }, + { + "epoch": 4.129844186975629, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 10337 + }, + { + "epoch": 4.130243707550939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10338 + }, + { + "epoch": 4.130643228126249, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10339 + }, + { + "epoch": 4.131042748701558, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10340 + }, + { + "epoch": 4.131442269276867, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10341 + }, + { + "epoch": 4.131841789852177, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10342 + }, + { + "epoch": 4.132241310427487, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10343 + }, + { + "epoch": 4.132640831002797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10344 + }, + { + "epoch": 4.1330403515781065, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10345 + }, + { + "epoch": 4.133439872153416, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10346 + }, + { + "epoch": 4.133839392728725, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10347 + }, + { + "epoch": 4.134238913304035, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10348 + }, + { + "epoch": 4.134638433879345, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10349 + }, + { + "epoch": 4.1350379544546545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10350 + }, + { + "epoch": 4.135437475029964, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 10351 + }, + { + "epoch": 4.135836995605274, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10352 + }, + { + "epoch": 4.136236516180583, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10353 + }, + { + "epoch": 4.136636036755893, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10354 + }, + { + "epoch": 4.1370355573312025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10355 + }, + { + "epoch": 4.137435077906512, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10356 + }, + { + "epoch": 4.137834598481822, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10357 + }, + { + "epoch": 4.138234119057131, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10358 + }, + { + "epoch": 4.138633639632441, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10359 + }, + { + "epoch": 4.139033160207751, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10360 + }, + { + "epoch": 4.13943268078306, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10361 + }, + { + "epoch": 4.13983220135837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10362 + }, + { + "epoch": 4.14023172193368, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10363 + }, + { + "epoch": 4.140631242508989, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10364 + }, + { + "epoch": 4.141030763084299, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10365 + }, + { + "epoch": 4.141430283659608, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10366 + }, + { + "epoch": 4.141829804234918, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10367 + }, + { + "epoch": 4.142229324810228, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10368 + }, + { + "epoch": 4.142628845385538, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10369 + }, + { + "epoch": 4.143028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10370 + }, + { + "epoch": 4.1434278865361565, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10371 + }, + { + "epoch": 4.143827407111466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10372 + }, + { + "epoch": 4.144226927686776, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10373 + }, + { + "epoch": 4.144626448262086, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10374 + }, + { + "epoch": 4.145025968837395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10375 + }, + { + "epoch": 4.1454254894127045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10376 + }, + { + "epoch": 4.145825009988014, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10377 + }, + { + "epoch": 4.146224530563324, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10378 + }, + { + "epoch": 4.146624051138634, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10379 + }, + { + "epoch": 4.147023571713944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10380 + }, + { + "epoch": 4.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10381 + }, + { + "epoch": 4.147822612864562, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10382 + }, + { + "epoch": 4.148222133439872, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10383 + }, + { + "epoch": 4.148621654015182, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10384 + }, + { + "epoch": 4.149021174590492, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10385 + }, + { + "epoch": 4.1494206951658015, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 10386 + }, + { + "epoch": 4.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10387 + }, + { + "epoch": 4.15021973631642, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10388 + }, + { + "epoch": 4.15061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10389 + }, + { + "epoch": 4.15101877746704, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10390 + }, + { + "epoch": 4.1514182980423495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10391 + }, + { + "epoch": 4.151817818617658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10392 + }, + { + "epoch": 4.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10393 + }, + { + "epoch": 4.152616859768278, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 10394 + }, + { + "epoch": 4.153016380343588, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10395 + }, + { + "epoch": 4.153415900918898, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 10396 + }, + { + "epoch": 4.153815421494207, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10397 + }, + { + "epoch": 4.154214942069516, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10398 + }, + { + "epoch": 4.154614462644826, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10399 + }, + { + "epoch": 4.155013983220136, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10400 + }, + { + "epoch": 4.155413503795446, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10401 + }, + { + "epoch": 4.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10402 + }, + { + "epoch": 4.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10403 + }, + { + "epoch": 4.156612065521374, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10404 + }, + { + "epoch": 4.157011586096684, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10405 + }, + { + "epoch": 4.157411106671994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10406 + }, + { + "epoch": 4.157810627247303, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10407 + }, + { + "epoch": 4.158210147822613, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10408 + }, + { + "epoch": 4.158609668397922, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10409 + }, + { + "epoch": 4.159009188973232, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10410 + }, + { + "epoch": 4.159408709548542, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10411 + }, + { + "epoch": 4.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10412 + }, + { + "epoch": 4.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10413 + }, + { + "epoch": 4.160607271274471, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10414 + }, + { + "epoch": 4.16100679184978, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10415 + }, + { + "epoch": 4.16140631242509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10416 + }, + { + "epoch": 4.1618058330003995, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10417 + }, + { + "epoch": 4.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10418 + }, + { + "epoch": 4.162604874151019, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10419 + }, + { + "epoch": 4.163004394726328, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10420 + }, + { + "epoch": 4.163403915301638, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10421 + }, + { + "epoch": 4.163803435876948, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10422 + }, + { + "epoch": 4.164202956452257, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10423 + }, + { + "epoch": 4.164602477027567, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10424 + }, + { + "epoch": 4.165001997602877, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10425 + }, + { + "epoch": 4.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10426 + }, + { + "epoch": 4.165801038753496, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10427 + }, + { + "epoch": 4.166200559328805, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10428 + }, + { + "epoch": 4.166600079904115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10429 + }, + { + "epoch": 4.166999600479425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10430 + }, + { + "epoch": 4.167399121054734, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10431 + }, + { + "epoch": 4.167798641630044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10432 + }, + { + "epoch": 4.168198162205353, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 10433 + }, + { + "epoch": 4.168597682780663, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10434 + }, + { + "epoch": 4.168997203355973, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10435 + }, + { + "epoch": 4.169396723931283, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10436 + }, + { + "epoch": 4.169796244506592, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10437 + }, + { + "epoch": 4.1701957650819015, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10438 + }, + { + "epoch": 4.170595285657211, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10439 + }, + { + "epoch": 4.170994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10440 + }, + { + "epoch": 4.171394326807831, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10441 + }, + { + "epoch": 4.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10442 + }, + { + "epoch": 4.1721933679584495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10443 + }, + { + "epoch": 4.172592888533759, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10444 + }, + { + "epoch": 4.172992409109069, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10445 + }, + { + "epoch": 4.173391929684379, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 10446 + }, + { + "epoch": 4.173791450259689, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10447 + }, + { + "epoch": 4.174190970834998, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10448 + }, + { + "epoch": 4.174590491410307, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10449 + }, + { + "epoch": 4.174990011985617, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10450 + }, + { + "epoch": 4.175389532560927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10451 + }, + { + "epoch": 4.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10452 + }, + { + "epoch": 4.1761885737115465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10453 + }, + { + "epoch": 4.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 10454 + }, + { + "epoch": 4.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10455 + }, + { + "epoch": 4.177387135437475, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10456 + }, + { + "epoch": 4.177786656012785, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10457 + }, + { + "epoch": 4.1781861765880945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10458 + }, + { + "epoch": 4.178585697163404, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10459 + }, + { + "epoch": 4.178985217738713, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10460 + }, + { + "epoch": 4.179384738314023, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10461 + }, + { + "epoch": 4.179784258889333, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10462 + }, + { + "epoch": 4.180183779464643, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10463 + }, + { + "epoch": 4.180583300039952, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10464 + }, + { + "epoch": 4.180982820615261, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10465 + }, + { + "epoch": 4.181382341190571, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10466 + }, + { + "epoch": 4.181781861765881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10467 + }, + { + "epoch": 4.182181382341191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10468 + }, + { + "epoch": 4.1825809029165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10469 + }, + { + "epoch": 4.18298042349181, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10470 + }, + { + "epoch": 4.183379944067119, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10471 + }, + { + "epoch": 4.183779464642429, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10472 + }, + { + "epoch": 4.184178985217739, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10473 + }, + { + "epoch": 4.1845785057930485, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10474 + }, + { + "epoch": 4.184978026368358, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10475 + }, + { + "epoch": 4.185377546943668, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10476 + }, + { + "epoch": 4.185777067518977, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10477 + }, + { + "epoch": 4.186176588094287, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 10478 + }, + { + "epoch": 4.1865761086695965, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10479 + }, + { + "epoch": 4.186975629244906, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10480 + }, + { + "epoch": 4.187375149820216, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10481 + }, + { + "epoch": 4.187774670395525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10482 + }, + { + "epoch": 4.188174190970835, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10483 + }, + { + "epoch": 4.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10484 + }, + { + "epoch": 4.188973232121454, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10485 + }, + { + "epoch": 4.189372752696764, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10486 + }, + { + "epoch": 4.189772273272074, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10487 + }, + { + "epoch": 4.190171793847383, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10488 + }, + { + "epoch": 4.190571314422693, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10489 + }, + { + "epoch": 4.190970834998002, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10490 + }, + { + "epoch": 4.191370355573312, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10491 + }, + { + "epoch": 4.191769876148622, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 10492 + }, + { + "epoch": 4.192169396723932, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10493 + }, + { + "epoch": 4.192568917299241, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10494 + }, + { + "epoch": 4.19296843787455, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10495 + }, + { + "epoch": 4.19336795844986, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10496 + }, + { + "epoch": 4.19376747902517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 10497 + }, + { + "epoch": 4.19416699960048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10498 + }, + { + "epoch": 4.194566520175789, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10499 + }, + { + "epoch": 4.1949660407510985, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10500 + }, + { + "epoch": 4.195365561326408, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10501 + }, + { + "epoch": 4.195765081901718, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10502 + }, + { + "epoch": 4.196164602477028, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10503 + }, + { + "epoch": 4.196564123052338, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10504 + }, + { + "epoch": 4.1969636436276465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10505 + }, + { + "epoch": 4.197363164202956, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10506 + }, + { + "epoch": 4.197762684778266, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10507 + }, + { + "epoch": 4.198162205353576, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10508 + }, + { + "epoch": 4.198561725928886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10509 + }, + { + "epoch": 4.1989612465041946, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10510 + }, + { + "epoch": 4.199360767079504, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10511 + }, + { + "epoch": 4.199760287654814, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10512 + }, + { + "epoch": 4.200159808230124, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10513 + }, + { + "epoch": 4.200559328805434, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10514 + }, + { + "epoch": 4.2009588493807435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10515 + }, + { + "epoch": 4.201358369956052, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10516 + }, + { + "epoch": 4.201757890531362, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10517 + }, + { + "epoch": 4.202157411106672, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 10518 + }, + { + "epoch": 4.202556931681982, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10519 + }, + { + "epoch": 4.2029564522572915, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10520 + }, + { + "epoch": 4.2033559728326, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10521 + }, + { + "epoch": 4.20375549340791, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10522 + }, + { + "epoch": 4.20415501398322, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10523 + }, + { + "epoch": 4.20455453455853, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10524 + }, + { + "epoch": 4.20495405513384, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10525 + }, + { + "epoch": 4.205353575709149, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10526 + }, + { + "epoch": 4.205753096284458, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10527 + }, + { + "epoch": 4.206152616859768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10528 + }, + { + "epoch": 4.206552137435078, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10529 + }, + { + "epoch": 4.206951658010388, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10530 + }, + { + "epoch": 4.207351178585697, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10531 + }, + { + "epoch": 4.207750699161007, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 10532 + }, + { + "epoch": 4.208150219736316, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10533 + }, + { + "epoch": 4.208549740311626, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10534 + }, + { + "epoch": 4.208949260886936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10535 + }, + { + "epoch": 4.2093487814622454, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10536 + }, + { + "epoch": 4.209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 10537 + }, + { + "epoch": 4.210147822612864, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10538 + }, + { + "epoch": 4.210547343188174, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10539 + }, + { + "epoch": 4.210946863763484, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10540 + }, + { + "epoch": 4.2113463843387935, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10541 + }, + { + "epoch": 4.211745904914103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10542 + }, + { + "epoch": 4.212145425489413, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10543 + }, + { + "epoch": 4.212544946064722, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10544 + }, + { + "epoch": 4.212944466640032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10545 + }, + { + "epoch": 4.2133439872153415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10546 + }, + { + "epoch": 4.213743507790651, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10547 + }, + { + "epoch": 4.214143028365961, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10548 + }, + { + "epoch": 4.214542548941271, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10549 + }, + { + "epoch": 4.21494206951658, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10550 + }, + { + "epoch": 4.21534159009189, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10551 + }, + { + "epoch": 4.215741110667199, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10552 + }, + { + "epoch": 4.216140631242509, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10553 + }, + { + "epoch": 4.216540151817819, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10554 + }, + { + "epoch": 4.216939672393128, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10555 + }, + { + "epoch": 4.217339192968438, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10556 + }, + { + "epoch": 4.217738713543747, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10557 + }, + { + "epoch": 4.218138234119057, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10558 + }, + { + "epoch": 4.218537754694367, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10559 + }, + { + "epoch": 4.218937275269677, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10560 + }, + { + "epoch": 4.219336795844986, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10561 + }, + { + "epoch": 4.2197363164202955, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10562 + }, + { + "epoch": 4.220135836995605, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10563 + }, + { + "epoch": 4.220535357570915, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10564 + }, + { + "epoch": 4.220934878146225, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10565 + }, + { + "epoch": 4.221334398721535, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10566 + }, + { + "epoch": 4.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10567 + }, + { + "epoch": 4.222133439872153, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10568 + }, + { + "epoch": 4.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10569 + }, + { + "epoch": 4.222932481022773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10570 + }, + { + "epoch": 4.223332001598083, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10571 + }, + { + "epoch": 4.2237315221733915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10572 + }, + { + "epoch": 4.224131042748701, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10573 + }, + { + "epoch": 4.224530563324011, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10574 + }, + { + "epoch": 4.224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10575 + }, + { + "epoch": 4.225329604474631, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10576 + }, + { + "epoch": 4.2257291250499405, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10577 + }, + { + "epoch": 4.226128645625249, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 10578 + }, + { + "epoch": 4.226528166200559, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10579 + }, + { + "epoch": 4.226927686775869, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10580 + }, + { + "epoch": 4.227327207351179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10581 + }, + { + "epoch": 4.2277267279264885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10582 + }, + { + "epoch": 4.228126248501797, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10583 + }, + { + "epoch": 4.228525769077107, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10584 + }, + { + "epoch": 4.228925289652417, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10585 + }, + { + "epoch": 4.229324810227727, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10586 + }, + { + "epoch": 4.229724330803037, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 10587 + }, + { + "epoch": 4.230123851378346, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10588 + }, + { + "epoch": 4.230523371953655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10589 + }, + { + "epoch": 4.230922892528965, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10590 + }, + { + "epoch": 4.231322413104275, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10591 + }, + { + "epoch": 4.231721933679585, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 10592 + }, + { + "epoch": 4.232121454254894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10593 + }, + { + "epoch": 4.232520974830204, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10594 + }, + { + "epoch": 4.232920495405513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10595 + }, + { + "epoch": 4.233320015980823, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10596 + }, + { + "epoch": 4.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10597 + }, + { + "epoch": 4.234119057131442, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10598 + }, + { + "epoch": 4.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10599 + }, + { + "epoch": 4.234918098282061, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10600 + }, + { + "epoch": 4.235317618857371, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10601 + }, + { + "epoch": 4.235717139432681, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10602 + }, + { + "epoch": 4.2361166600079905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10603 + }, + { + "epoch": 4.2365161805833, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10604 + }, + { + "epoch": 4.23691570115861, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 10605 + }, + { + "epoch": 4.237315221733919, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10606 + }, + { + "epoch": 4.237714742309229, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10607 + }, + { + "epoch": 4.2381142628845385, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10608 + }, + { + "epoch": 4.238513783459848, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10609 + }, + { + "epoch": 4.238913304035158, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10610 + }, + { + "epoch": 4.239312824610467, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10611 + }, + { + "epoch": 4.239712345185777, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 10612 + }, + { + "epoch": 4.240111865761087, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10613 + }, + { + "epoch": 4.240511386336396, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10614 + }, + { + "epoch": 4.240910906911706, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 10615 + }, + { + "epoch": 4.241310427487016, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10616 + }, + { + "epoch": 4.241709948062325, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10617 + }, + { + "epoch": 4.242109468637635, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10618 + }, + { + "epoch": 4.242508989212944, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10619 + }, + { + "epoch": 4.242908509788254, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10620 + }, + { + "epoch": 4.243308030363564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10621 + }, + { + "epoch": 4.243707550938874, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 10622 + }, + { + "epoch": 4.244107071514183, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10623 + }, + { + "epoch": 4.244506592089492, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10624 + }, + { + "epoch": 4.244906112664802, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10625 + }, + { + "epoch": 4.245305633240112, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10626 + }, + { + "epoch": 4.245705153815422, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10627 + }, + { + "epoch": 4.246104674390731, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 10628 + }, + { + "epoch": 4.2465041949660405, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10629 + }, + { + "epoch": 4.24690371554135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10630 + }, + { + "epoch": 4.24730323611666, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10631 + }, + { + "epoch": 4.24770275669197, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10632 + }, + { + "epoch": 4.24810227726728, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10633 + }, + { + "epoch": 4.2485017978425885, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10634 + }, + { + "epoch": 4.248901318417898, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10635 + }, + { + "epoch": 4.249300838993208, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10636 + }, + { + "epoch": 4.249700359568518, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10637 + }, + { + "epoch": 4.250099880143828, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10638 + }, + { + "epoch": 4.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10639 + }, + { + "epoch": 4.250898921294446, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10640 + }, + { + "epoch": 4.251298441869756, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10641 + }, + { + "epoch": 4.251697962445066, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10642 + }, + { + "epoch": 4.252097483020376, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10643 + }, + { + "epoch": 4.2524970035956855, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10644 + }, + { + "epoch": 4.252896524170994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10645 + }, + { + "epoch": 4.253296044746304, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10646 + }, + { + "epoch": 4.253695565321614, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10647 + }, + { + "epoch": 4.254095085896924, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10648 + }, + { + "epoch": 4.2544946064722335, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10649 + }, + { + "epoch": 4.254894127047543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 10650 + }, + { + "epoch": 4.255293647622852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10651 + }, + { + "epoch": 4.255693168198162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10652 + }, + { + "epoch": 4.256092688773472, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10653 + }, + { + "epoch": 4.256492209348782, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10654 + }, + { + "epoch": 4.256891729924091, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10655 + }, + { + "epoch": 4.257291250499401, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10656 + }, + { + "epoch": 4.25769077107471, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10657 + }, + { + "epoch": 4.25809029165002, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10658 + }, + { + "epoch": 4.25848981222533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10659 + }, + { + "epoch": 4.258889332800639, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10660 + }, + { + "epoch": 4.259288853375949, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10661 + }, + { + "epoch": 4.259688373951258, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10662 + }, + { + "epoch": 4.260087894526568, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10663 + }, + { + "epoch": 4.260487415101878, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10664 + }, + { + "epoch": 4.2608869356771875, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10665 + }, + { + "epoch": 4.261286456252497, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10666 + }, + { + "epoch": 4.261685976827807, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10667 + }, + { + "epoch": 4.262085497403116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10668 + }, + { + "epoch": 4.262485017978426, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10669 + }, + { + "epoch": 4.2628845385537355, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 10670 + }, + { + "epoch": 4.263284059129045, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10671 + }, + { + "epoch": 4.263683579704355, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10672 + }, + { + "epoch": 4.264083100279665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10673 + }, + { + "epoch": 4.264482620854974, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10674 + }, + { + "epoch": 4.2648821414302835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10675 + }, + { + "epoch": 4.265281662005593, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10676 + }, + { + "epoch": 4.265681182580903, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10677 + }, + { + "epoch": 4.266080703156213, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10678 + }, + { + "epoch": 4.266480223731522, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10679 + }, + { + "epoch": 4.266879744306832, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10680 + }, + { + "epoch": 4.267279264882141, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 10681 + }, + { + "epoch": 4.267678785457451, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10682 + }, + { + "epoch": 4.268078306032761, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 10683 + }, + { + "epoch": 4.26847782660807, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10684 + }, + { + "epoch": 4.26887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 10685 + }, + { + "epoch": 4.269276867758689, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 10686 + }, + { + "epoch": 4.269676388333999, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 10687 + }, + { + "epoch": 4.270075908909309, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10688 + }, + { + "epoch": 4.270475429484619, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10689 + }, + { + "epoch": 4.270874950059928, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10690 + }, + { + "epoch": 4.2712744706352375, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10691 + }, + { + "epoch": 4.271673991210547, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 10692 + }, + { + "epoch": 4.272073511785857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10693 + }, + { + "epoch": 4.272473032361167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10694 + }, + { + "epoch": 4.272872552936477, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10695 + }, + { + "epoch": 4.2732720735117855, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10696 + }, + { + "epoch": 4.273671594087095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10697 + }, + { + "epoch": 4.274071114662405, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10698 + }, + { + "epoch": 4.274470635237715, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10699 + }, + { + "epoch": 4.274870155813025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10700 + }, + { + "epoch": 4.2752696763883336, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10701 + }, + { + "epoch": 4.275669196963643, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10702 + }, + { + "epoch": 4.276068717538953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10703 + }, + { + "epoch": 4.276468238114263, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10704 + }, + { + "epoch": 4.276867758689573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 10705 + }, + { + "epoch": 4.2772672792648825, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10706 + }, + { + "epoch": 4.277666799840191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10707 + }, + { + "epoch": 4.278066320415501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10708 + }, + { + "epoch": 4.278465840990811, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10709 + }, + { + "epoch": 4.278865361566121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10710 + }, + { + "epoch": 4.2792648821414305, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10711 + }, + { + "epoch": 4.27966440271674, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10712 + }, + { + "epoch": 4.280063923292049, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10713 + }, + { + "epoch": 4.280463443867359, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10714 + }, + { + "epoch": 4.280862964442669, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10715 + }, + { + "epoch": 4.281262485017979, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10716 + }, + { + "epoch": 4.281662005593288, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10717 + }, + { + "epoch": 4.282061526168597, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10718 + }, + { + "epoch": 4.282461046743907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10719 + }, + { + "epoch": 4.282860567319217, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10720 + }, + { + "epoch": 4.283260087894527, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10721 + }, + { + "epoch": 4.283659608469836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10722 + }, + { + "epoch": 4.284059129045146, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10723 + }, + { + "epoch": 4.284458649620455, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10724 + }, + { + "epoch": 4.284858170195765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 10725 + }, + { + "epoch": 4.285257690771075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10726 + }, + { + "epoch": 4.285657211346384, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 10727 + }, + { + "epoch": 4.286056731921694, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10728 + }, + { + "epoch": 4.286456252497004, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10729 + }, + { + "epoch": 4.286855773072313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10730 + }, + { + "epoch": 4.287255293647623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10731 + }, + { + "epoch": 4.2876548142229325, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10732 + }, + { + "epoch": 4.288054334798242, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10733 + }, + { + "epoch": 4.288453855373552, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10734 + }, + { + "epoch": 4.288853375948861, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10735 + }, + { + "epoch": 4.289252896524171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 10736 + }, + { + "epoch": 4.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10737 + }, + { + "epoch": 4.29005193767479, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10738 + }, + { + "epoch": 4.2904514582501, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10739 + }, + { + "epoch": 4.29085097882541, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10740 + }, + { + "epoch": 4.291250499400719, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10741 + }, + { + "epoch": 4.291650019976029, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10742 + }, + { + "epoch": 4.292049540551338, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10743 + }, + { + "epoch": 4.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10744 + }, + { + "epoch": 4.292848581701958, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10745 + }, + { + "epoch": 4.293248102277268, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10746 + }, + { + "epoch": 4.293647622852577, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10747 + }, + { + "epoch": 4.294047143427886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10748 + }, + { + "epoch": 4.294446664003196, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10749 + }, + { + "epoch": 4.294846184578506, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10750 + }, + { + "epoch": 4.295245705153816, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10751 + }, + { + "epoch": 4.295645225729125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 10752 + }, + { + "epoch": 4.2960447463044344, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10753 + }, + { + "epoch": 4.296444266879744, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10754 + }, + { + "epoch": 4.296843787455054, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10755 + }, + { + "epoch": 4.297243308030364, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 10756 + }, + { + "epoch": 4.297642828605674, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10757 + }, + { + "epoch": 4.2980423491809825, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10758 + }, + { + "epoch": 4.298441869756292, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10759 + }, + { + "epoch": 4.298841390331602, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10760 + }, + { + "epoch": 4.299240910906912, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10761 + }, + { + "epoch": 4.299640431482222, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 10762 + }, + { + "epoch": 4.300039952057531, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10763 + }, + { + "epoch": 4.30043947263284, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10764 + }, + { + "epoch": 4.30083899320815, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10765 + }, + { + "epoch": 4.30123851378346, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10766 + }, + { + "epoch": 4.30163803435877, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 10767 + }, + { + "epoch": 4.3020375549340795, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10768 + }, + { + "epoch": 4.302437075509388, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10769 + }, + { + "epoch": 4.302836596084698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10770 + }, + { + "epoch": 4.303236116660008, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10771 + }, + { + "epoch": 4.303635637235318, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10772 + }, + { + "epoch": 4.3040351578106275, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10773 + }, + { + "epoch": 4.304434678385936, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10774 + }, + { + "epoch": 4.304834198961246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10775 + }, + { + "epoch": 4.305233719536556, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10776 + }, + { + "epoch": 4.305633240111866, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10777 + }, + { + "epoch": 4.3060327606871756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10778 + }, + { + "epoch": 4.306432281262485, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10779 + }, + { + "epoch": 4.306831801837794, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10780 + }, + { + "epoch": 4.307231322413104, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10781 + }, + { + "epoch": 4.307630842988414, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10782 + }, + { + "epoch": 4.308030363563724, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10783 + }, + { + "epoch": 4.308429884139033, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10784 + }, + { + "epoch": 4.308829404714343, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10785 + }, + { + "epoch": 4.309228925289652, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10786 + }, + { + "epoch": 4.309628445864962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10787 + }, + { + "epoch": 4.310027966440272, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10788 + }, + { + "epoch": 4.310427487015581, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10789 + }, + { + "epoch": 4.310827007590891, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10790 + }, + { + "epoch": 4.3112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10791 + }, + { + "epoch": 4.31162604874151, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10792 + }, + { + "epoch": 4.31202556931682, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10793 + }, + { + "epoch": 4.3124250898921295, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10794 + }, + { + "epoch": 4.312824610467439, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10795 + }, + { + "epoch": 4.313224131042749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10796 + }, + { + "epoch": 4.313623651618058, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10797 + }, + { + "epoch": 4.314023172193368, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10798 + }, + { + "epoch": 4.3144226927686775, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10799 + }, + { + "epoch": 4.314822213343987, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10800 + }, + { + "epoch": 4.315221733919297, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10801 + }, + { + "epoch": 4.315621254494607, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10802 + }, + { + "epoch": 4.316020775069916, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10803 + }, + { + "epoch": 4.316420295645226, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 10804 + }, + { + "epoch": 4.316819816220535, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10805 + }, + { + "epoch": 4.317219336795845, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10806 + }, + { + "epoch": 4.317618857371155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10807 + }, + { + "epoch": 4.318018377946464, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10808 + }, + { + "epoch": 4.318417898521774, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10809 + }, + { + "epoch": 4.318817419097083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10810 + }, + { + "epoch": 4.319216939672393, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10811 + }, + { + "epoch": 4.319616460247703, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10812 + }, + { + "epoch": 4.320015980823013, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10813 + }, + { + "epoch": 4.320415501398322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10814 + }, + { + "epoch": 4.320815021973631, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10815 + }, + { + "epoch": 4.321214542548941, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10816 + }, + { + "epoch": 4.321614063124251, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10817 + }, + { + "epoch": 4.322013583699561, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10818 + }, + { + "epoch": 4.322413104274871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10819 + }, + { + "epoch": 4.3228126248501795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10820 + }, + { + "epoch": 4.323212145425489, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10821 + }, + { + "epoch": 4.323611666000799, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 10822 + }, + { + "epoch": 4.324011186576109, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 10823 + }, + { + "epoch": 4.324410707151419, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10824 + }, + { + "epoch": 4.3248102277267275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10825 + }, + { + "epoch": 4.325209748302037, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10826 + }, + { + "epoch": 4.325609268877347, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10827 + }, + { + "epoch": 4.326008789452657, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10828 + }, + { + "epoch": 4.326408310027967, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10829 + }, + { + "epoch": 4.3268078306032765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10830 + }, + { + "epoch": 4.327207351178585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10831 + }, + { + "epoch": 4.327606871753895, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10832 + }, + { + "epoch": 4.328006392329205, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10833 + }, + { + "epoch": 4.328405912904515, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10834 + }, + { + "epoch": 4.3288054334798245, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10835 + }, + { + "epoch": 4.329204954055134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10836 + }, + { + "epoch": 4.329604474630443, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10837 + }, + { + "epoch": 4.330003995205753, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10838 + }, + { + "epoch": 4.330403515781063, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10839 + }, + { + "epoch": 4.3308030363563725, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10840 + }, + { + "epoch": 4.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10841 + }, + { + "epoch": 4.331602077506991, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10842 + }, + { + "epoch": 4.332001598082301, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10843 + }, + { + "epoch": 4.332401118657611, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10844 + }, + { + "epoch": 4.332800639232921, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10845 + }, + { + "epoch": 4.33320015980823, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10846 + }, + { + "epoch": 4.33359968038354, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10847 + }, + { + "epoch": 4.333999200958849, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10848 + }, + { + "epoch": 4.334398721534159, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10849 + }, + { + "epoch": 4.334798242109469, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10850 + }, + { + "epoch": 4.335197762684778, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10851 + }, + { + "epoch": 4.335597283260088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10852 + }, + { + "epoch": 4.335996803835398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10853 + }, + { + "epoch": 4.336396324410707, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10854 + }, + { + "epoch": 4.336795844986017, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10855 + }, + { + "epoch": 4.3371953655613265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10856 + }, + { + "epoch": 4.337594886136636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10857 + }, + { + "epoch": 4.337994406711946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10858 + }, + { + "epoch": 4.338393927287255, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10859 + }, + { + "epoch": 4.338793447862565, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10860 + }, + { + "epoch": 4.3391929684378745, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10861 + }, + { + "epoch": 4.339592489013184, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10862 + }, + { + "epoch": 4.339992009588494, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 10863 + }, + { + "epoch": 4.340391530163803, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 10864 + }, + { + "epoch": 4.340791050739113, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10865 + }, + { + "epoch": 4.3411905713144225, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10866 + }, + { + "epoch": 4.341590091889732, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10867 + }, + { + "epoch": 4.341989612465042, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10868 + }, + { + "epoch": 4.342389133040352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10869 + }, + { + "epoch": 4.342788653615661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10870 + }, + { + "epoch": 4.343188174190971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10871 + }, + { + "epoch": 4.34358769476628, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 10872 + }, + { + "epoch": 4.34398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10873 + }, + { + "epoch": 4.3443867359169, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10874 + }, + { + "epoch": 4.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10875 + }, + { + "epoch": 4.345185777067519, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10876 + }, + { + "epoch": 4.345585297642828, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10877 + }, + { + "epoch": 4.345984818218138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10878 + }, + { + "epoch": 4.346384338793448, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10879 + }, + { + "epoch": 4.346783859368758, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10880 + }, + { + "epoch": 4.347183379944067, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10881 + }, + { + "epoch": 4.3475829005193765, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10882 + }, + { + "epoch": 4.347982421094686, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10883 + }, + { + "epoch": 4.348381941669996, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10884 + }, + { + "epoch": 4.348781462245306, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10885 + }, + { + "epoch": 4.349180982820616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10886 + }, + { + "epoch": 4.3495805033959245, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10887 + }, + { + "epoch": 4.349980023971234, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10888 + }, + { + "epoch": 4.350379544546544, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10889 + }, + { + "epoch": 4.350779065121854, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10890 + }, + { + "epoch": 4.351178585697164, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10891 + }, + { + "epoch": 4.351578106272473, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10892 + }, + { + "epoch": 4.351977626847782, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 10893 + }, + { + "epoch": 4.352377147423092, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10894 + }, + { + "epoch": 4.352776667998402, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10895 + }, + { + "epoch": 4.353176188573712, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10896 + }, + { + "epoch": 4.3535757091490215, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10897 + }, + { + "epoch": 4.35397522972433, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10898 + }, + { + "epoch": 4.35437475029964, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10899 + }, + { + "epoch": 4.35477427087495, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10900 + }, + { + "epoch": 4.35517379145026, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10901 + }, + { + "epoch": 4.3555733120255695, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10902 + }, + { + "epoch": 4.355972832600879, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10903 + }, + { + "epoch": 4.356372353176188, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10904 + }, + { + "epoch": 4.356771873751498, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 10905 + }, + { + "epoch": 4.357171394326808, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10906 + }, + { + "epoch": 4.357570914902118, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10907 + }, + { + "epoch": 4.357970435477427, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10908 + }, + { + "epoch": 4.358369956052737, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10909 + }, + { + "epoch": 4.358769476628046, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10910 + }, + { + "epoch": 4.359168997203356, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10911 + }, + { + "epoch": 4.359568517778666, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10912 + }, + { + "epoch": 4.359968038353975, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10913 + }, + { + "epoch": 4.360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10914 + }, + { + "epoch": 4.360767079504594, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10915 + }, + { + "epoch": 4.361166600079904, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10916 + }, + { + "epoch": 4.361566120655214, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10917 + }, + { + "epoch": 4.361965641230523, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10918 + }, + { + "epoch": 4.362365161805833, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10919 + }, + { + "epoch": 4.362764682381143, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10920 + }, + { + "epoch": 4.363164202956452, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10921 + }, + { + "epoch": 4.363563723531762, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10922 + }, + { + "epoch": 4.3639632441070715, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10923 + }, + { + "epoch": 4.364362764682381, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10924 + }, + { + "epoch": 4.364762285257691, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10925 + }, + { + "epoch": 4.365161805833001, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10926 + }, + { + "epoch": 4.36556132640831, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10927 + }, + { + "epoch": 4.3659608469836195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10928 + }, + { + "epoch": 4.366360367558929, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10929 + }, + { + "epoch": 4.366759888134239, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10930 + }, + { + "epoch": 4.367159408709549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10931 + }, + { + "epoch": 4.367558929284858, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10932 + }, + { + "epoch": 4.367958449860168, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10933 + }, + { + "epoch": 4.368357970435477, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10934 + }, + { + "epoch": 4.368757491010787, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10935 + }, + { + "epoch": 4.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10936 + }, + { + "epoch": 4.369556532161407, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10937 + }, + { + "epoch": 4.369956052736716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10938 + }, + { + "epoch": 4.370355573312025, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10939 + }, + { + "epoch": 4.370755093887335, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10940 + }, + { + "epoch": 4.371154614462645, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10941 + }, + { + "epoch": 4.371554135037955, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10942 + }, + { + "epoch": 4.3719536556132645, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10943 + }, + { + "epoch": 4.372353176188573, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10944 + }, + { + "epoch": 4.372752696763883, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 10945 + }, + { + "epoch": 4.373152217339193, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10946 + }, + { + "epoch": 4.373551737914503, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10947 + }, + { + "epoch": 4.373951258489813, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10948 + }, + { + "epoch": 4.3743507790651215, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10949 + }, + { + "epoch": 4.374750299640431, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 10950 + }, + { + "epoch": 4.375149820215741, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 10951 + }, + { + "epoch": 4.375549340791051, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10952 + }, + { + "epoch": 4.375948861366361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10953 + }, + { + "epoch": 4.3763483819416695, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 10954 + }, + { + "epoch": 4.376747902516979, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10955 + }, + { + "epoch": 4.377147423092289, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10956 + }, + { + "epoch": 4.377546943667599, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10957 + }, + { + "epoch": 4.377946464242909, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10958 + }, + { + "epoch": 4.3783459848182185, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10959 + }, + { + "epoch": 4.378745505393527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10960 + }, + { + "epoch": 4.379145025968837, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10961 + }, + { + "epoch": 4.379544546544147, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 10962 + }, + { + "epoch": 4.379944067119457, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10963 + }, + { + "epoch": 4.3803435876947665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10964 + }, + { + "epoch": 4.380743108270076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10965 + }, + { + "epoch": 4.381142628845385, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10966 + }, + { + "epoch": 4.381542149420695, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10967 + }, + { + "epoch": 4.381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10968 + }, + { + "epoch": 4.3823411905713145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10969 + }, + { + "epoch": 4.382740711146624, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10970 + }, + { + "epoch": 4.383140231721933, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10971 + }, + { + "epoch": 4.383539752297243, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10972 + }, + { + "epoch": 4.383939272872553, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10973 + }, + { + "epoch": 4.384338793447863, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10974 + }, + { + "epoch": 4.384738314023172, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10975 + }, + { + "epoch": 4.385137834598482, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10976 + }, + { + "epoch": 4.385537355173791, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10977 + }, + { + "epoch": 4.385936875749101, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10978 + }, + { + "epoch": 4.386336396324411, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10979 + }, + { + "epoch": 4.38673591689972, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10980 + }, + { + "epoch": 4.38713543747503, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10981 + }, + { + "epoch": 4.38753495805034, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10982 + }, + { + "epoch": 4.387934478625649, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10983 + }, + { + "epoch": 4.388333999200959, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10984 + }, + { + "epoch": 4.3887335197762685, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10985 + }, + { + "epoch": 4.389133040351578, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 10986 + }, + { + "epoch": 4.389532560926888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 10987 + }, + { + "epoch": 4.389932081502197, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10988 + }, + { + "epoch": 4.390331602077507, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10989 + }, + { + "epoch": 4.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 10990 + }, + { + "epoch": 4.391130643228126, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10991 + }, + { + "epoch": 4.391530163803436, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10992 + }, + { + "epoch": 4.391929684378746, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10993 + }, + { + "epoch": 4.392329204954055, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 10994 + }, + { + "epoch": 4.3927287255293646, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10995 + }, + { + "epoch": 4.393128246104674, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10996 + }, + { + "epoch": 4.393527766679984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10997 + }, + { + "epoch": 4.393927287255294, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10998 + }, + { + "epoch": 4.394326807830604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10999 + }, + { + "epoch": 4.394726328405913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11000 + }, + { + "epoch": 4.395125848981222, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11001 + }, + { + "epoch": 4.395525369556532, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11002 + }, + { + "epoch": 4.395924890131842, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 11003 + }, + { + "epoch": 4.396324410707152, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11004 + }, + { + "epoch": 4.396723931282461, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11005 + }, + { + "epoch": 4.39712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11006 + }, + { + "epoch": 4.39752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11007 + }, + { + "epoch": 4.39792249300839, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11008 + }, + { + "epoch": 4.3983220135837, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11009 + }, + { + "epoch": 4.39872153415901, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11010 + }, + { + "epoch": 4.3991210547343185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11011 + }, + { + "epoch": 4.399520575309628, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11012 + }, + { + "epoch": 4.399920095884938, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11013 + }, + { + "epoch": 4.400319616460248, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11014 + }, + { + "epoch": 4.400719137035558, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11015 + }, + { + "epoch": 4.401118657610867, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11016 + }, + { + "epoch": 4.401518178186176, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11017 + }, + { + "epoch": 4.401917698761486, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11018 + }, + { + "epoch": 4.402317219336796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11019 + }, + { + "epoch": 4.402716739912106, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11020 + }, + { + "epoch": 4.4031162604874154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11021 + }, + { + "epoch": 4.403515781062724, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11022 + }, + { + "epoch": 4.403915301638034, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11023 + }, + { + "epoch": 4.404314822213344, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11024 + }, + { + "epoch": 4.404714342788654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11025 + }, + { + "epoch": 4.4051138633639635, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11026 + }, + { + "epoch": 4.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11027 + }, + { + "epoch": 4.405912904514582, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11028 + }, + { + "epoch": 4.406312425089892, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11029 + }, + { + "epoch": 4.406711945665202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11030 + }, + { + "epoch": 4.4071114662405115, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11031 + }, + { + "epoch": 4.407510986815821, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11032 + }, + { + "epoch": 4.407910507391131, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 11033 + }, + { + "epoch": 4.40831002796644, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11034 + }, + { + "epoch": 4.40870954854175, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11035 + }, + { + "epoch": 4.40910906911706, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11036 + }, + { + "epoch": 4.409508589692369, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11037 + }, + { + "epoch": 4.409908110267679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11038 + }, + { + "epoch": 4.410307630842988, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11039 + }, + { + "epoch": 4.410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11040 + }, + { + "epoch": 4.411106671993608, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11041 + }, + { + "epoch": 4.411506192568917, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11042 + }, + { + "epoch": 4.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11043 + }, + { + "epoch": 4.412305233719536, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11044 + }, + { + "epoch": 4.412704754294846, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11045 + }, + { + "epoch": 4.413104274870156, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11046 + }, + { + "epoch": 4.4135037954454654, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11047 + }, + { + "epoch": 4.413903316020775, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11048 + }, + { + "epoch": 4.414302836596085, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11049 + }, + { + "epoch": 4.414702357171394, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11050 + }, + { + "epoch": 4.415101877746704, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11051 + }, + { + "epoch": 4.4155013983220135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11052 + }, + { + "epoch": 4.415900918897323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11053 + }, + { + "epoch": 4.416300439472633, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11054 + }, + { + "epoch": 4.416699960047943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11055 + }, + { + "epoch": 4.417099480623252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11056 + }, + { + "epoch": 4.4174990011985615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11057 + }, + { + "epoch": 4.417898521773871, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11058 + }, + { + "epoch": 4.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11059 + }, + { + "epoch": 4.418697562924491, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11060 + }, + { + "epoch": 4.4190970834998, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11061 + }, + { + "epoch": 4.41949660407511, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11062 + }, + { + "epoch": 4.419896124650419, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11063 + }, + { + "epoch": 4.420295645225729, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11064 + }, + { + "epoch": 4.420695165801039, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11065 + }, + { + "epoch": 4.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11066 + }, + { + "epoch": 4.421494206951658, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11067 + }, + { + "epoch": 4.421893727526967, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11068 + }, + { + "epoch": 4.422293248102277, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11069 + }, + { + "epoch": 4.422692768677587, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11070 + }, + { + "epoch": 4.423092289252897, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11071 + }, + { + "epoch": 4.4234918098282066, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11072 + }, + { + "epoch": 4.4238913304035155, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11073 + }, + { + "epoch": 4.424290850978825, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11074 + }, + { + "epoch": 4.424690371554135, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11075 + }, + { + "epoch": 4.425089892129445, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11076 + }, + { + "epoch": 4.425489412704755, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11077 + }, + { + "epoch": 4.4258889332800635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11078 + }, + { + "epoch": 4.426288453855373, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11079 + }, + { + "epoch": 4.426687974430683, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11080 + }, + { + "epoch": 4.427087495005993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11081 + }, + { + "epoch": 4.427487015581303, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11082 + }, + { + "epoch": 4.427886536156612, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11083 + }, + { + "epoch": 4.428286056731921, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11084 + }, + { + "epoch": 4.428685577307231, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11085 + }, + { + "epoch": 4.429085097882541, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11086 + }, + { + "epoch": 4.429484618457851, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11087 + }, + { + "epoch": 4.4298841390331605, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 11088 + }, + { + "epoch": 4.43028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11089 + }, + { + "epoch": 4.430683180183779, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11090 + }, + { + "epoch": 4.431082700759089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11091 + }, + { + "epoch": 4.431482221334399, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 11092 + }, + { + "epoch": 4.4318817419097085, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11093 + }, + { + "epoch": 4.432281262485018, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11094 + }, + { + "epoch": 4.432680783060327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11095 + }, + { + "epoch": 4.433080303635637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11096 + }, + { + "epoch": 4.433479824210947, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11097 + }, + { + "epoch": 4.433879344786257, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 11098 + }, + { + "epoch": 4.434278865361566, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11099 + }, + { + "epoch": 4.434678385936876, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11100 + }, + { + "epoch": 4.435077906512185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11101 + }, + { + "epoch": 4.435477427087495, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11102 + }, + { + "epoch": 4.435876947662805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11103 + }, + { + "epoch": 4.436276468238114, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11104 + }, + { + "epoch": 4.436675988813424, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11105 + }, + { + "epoch": 4.437075509388734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11106 + }, + { + "epoch": 4.437475029964043, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 11107 + }, + { + "epoch": 4.437874550539353, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 11108 + }, + { + "epoch": 4.438274071114662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11109 + }, + { + "epoch": 4.438673591689972, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11110 + }, + { + "epoch": 4.439073112265282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11111 + }, + { + "epoch": 4.439472632840591, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11112 + }, + { + "epoch": 4.439872153415901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11113 + }, + { + "epoch": 4.4402716739912105, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11114 + }, + { + "epoch": 4.44067119456652, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11115 + }, + { + "epoch": 4.44107071514183, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11116 + }, + { + "epoch": 4.44147023571714, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11117 + }, + { + "epoch": 4.441869756292449, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11118 + }, + { + "epoch": 4.4422692768677585, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 11119 + }, + { + "epoch": 4.442668797443068, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11120 + }, + { + "epoch": 4.443068318018378, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11121 + }, + { + "epoch": 4.443467838593688, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11122 + }, + { + "epoch": 4.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11123 + }, + { + "epoch": 4.444266879744307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11124 + }, + { + "epoch": 4.444666400319616, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11125 + }, + { + "epoch": 4.445065920894926, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11126 + }, + { + "epoch": 4.445465441470236, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11127 + }, + { + "epoch": 4.445864962045546, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11128 + }, + { + "epoch": 4.446264482620855, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11129 + }, + { + "epoch": 4.446664003196164, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11130 + }, + { + "epoch": 4.447063523771474, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11131 + }, + { + "epoch": 4.447463044346784, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11132 + }, + { + "epoch": 4.447862564922094, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11133 + }, + { + "epoch": 4.448262085497403, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11134 + }, + { + "epoch": 4.448661606072712, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11135 + }, + { + "epoch": 4.449061126648022, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11136 + }, + { + "epoch": 4.449460647223332, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 11137 + }, + { + "epoch": 4.449860167798642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11138 + }, + { + "epoch": 4.450259688373952, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11139 + }, + { + "epoch": 4.4506592089492605, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 11140 + }, + { + "epoch": 4.45105872952457, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11141 + }, + { + "epoch": 4.45145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11142 + }, + { + "epoch": 4.45185777067519, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11143 + }, + { + "epoch": 4.4522572912505, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11144 + }, + { + "epoch": 4.452656811825809, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11145 + }, + { + "epoch": 4.453056332401118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11146 + }, + { + "epoch": 4.453455852976428, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11147 + }, + { + "epoch": 4.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11148 + }, + { + "epoch": 4.454254894127048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11149 + }, + { + "epoch": 4.4546544147023575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11150 + }, + { + "epoch": 4.455053935277666, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11151 + }, + { + "epoch": 4.455453455852976, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11152 + }, + { + "epoch": 4.455852976428286, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11153 + }, + { + "epoch": 4.456252497003596, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11154 + }, + { + "epoch": 4.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11155 + }, + { + "epoch": 4.457051538154215, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11156 + }, + { + "epoch": 4.457451058729524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11157 + }, + { + "epoch": 4.457850579304834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11158 + }, + { + "epoch": 4.458250099880144, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11159 + }, + { + "epoch": 4.4586496204554535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11160 + }, + { + "epoch": 4.459049141030763, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11161 + }, + { + "epoch": 4.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11162 + }, + { + "epoch": 4.459848182181382, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11163 + }, + { + "epoch": 4.460247702756692, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11164 + }, + { + "epoch": 4.460647223332002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11165 + }, + { + "epoch": 4.461046743907311, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11166 + }, + { + "epoch": 4.461446264482621, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11167 + }, + { + "epoch": 4.46184578505793, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11168 + }, + { + "epoch": 4.46224530563324, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11169 + }, + { + "epoch": 4.46264482620855, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11170 + }, + { + "epoch": 4.463044346783859, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11171 + }, + { + "epoch": 4.463443867359169, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 11172 + }, + { + "epoch": 4.463843387934479, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11173 + }, + { + "epoch": 4.464242908509788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11174 + }, + { + "epoch": 4.464642429085098, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11175 + }, + { + "epoch": 4.4650419496604075, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11176 + }, + { + "epoch": 4.465441470235717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11177 + }, + { + "epoch": 4.465840990811027, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11178 + }, + { + "epoch": 4.466240511386337, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11179 + }, + { + "epoch": 4.466640031961646, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11180 + }, + { + "epoch": 4.4670395525369555, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11181 + }, + { + "epoch": 4.467439073112265, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11182 + }, + { + "epoch": 4.467838593687575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11183 + }, + { + "epoch": 4.468238114262885, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11184 + }, + { + "epoch": 4.468637634838194, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 11185 + }, + { + "epoch": 4.4690371554135035, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11186 + }, + { + "epoch": 4.469436675988813, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11187 + }, + { + "epoch": 4.469836196564123, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11188 + }, + { + "epoch": 4.470235717139433, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11189 + }, + { + "epoch": 4.470635237714743, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 11190 + }, + { + "epoch": 4.471034758290052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11191 + }, + { + "epoch": 4.471434278865361, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11192 + }, + { + "epoch": 4.471833799440671, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11193 + }, + { + "epoch": 4.472233320015981, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11194 + }, + { + "epoch": 4.472632840591291, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11195 + }, + { + "epoch": 4.4730323611666005, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11196 + }, + { + "epoch": 4.473431881741909, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11197 + }, + { + "epoch": 4.473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11198 + }, + { + "epoch": 4.474230922892529, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11199 + }, + { + "epoch": 4.474630443467839, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11200 + }, + { + "epoch": 4.475029964043149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11201 + }, + { + "epoch": 4.4754294846184575, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 11202 + }, + { + "epoch": 4.475829005193767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 11203 + }, + { + "epoch": 4.476228525769077, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11204 + }, + { + "epoch": 4.476628046344387, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 11205 + }, + { + "epoch": 4.477027566919697, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11206 + }, + { + "epoch": 4.477427087495006, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11207 + }, + { + "epoch": 4.477826608070315, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11208 + }, + { + "epoch": 4.478226128645625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11209 + }, + { + "epoch": 4.478625649220935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11210 + }, + { + "epoch": 4.479025169796245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11211 + }, + { + "epoch": 4.479424690371554, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11212 + }, + { + "epoch": 4.479824210946864, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11213 + }, + { + "epoch": 4.480223731522173, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11214 + }, + { + "epoch": 4.480623252097483, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11215 + }, + { + "epoch": 4.481022772672793, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11216 + }, + { + "epoch": 4.4814222932481025, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11217 + }, + { + "epoch": 4.481821813823412, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11218 + }, + { + "epoch": 4.482221334398721, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11219 + }, + { + "epoch": 4.482620854974031, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11220 + }, + { + "epoch": 4.483020375549341, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11221 + }, + { + "epoch": 4.4834198961246505, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11222 + }, + { + "epoch": 4.48381941669996, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11223 + }, + { + "epoch": 4.484218937275269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11224 + }, + { + "epoch": 4.484618457850579, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11225 + }, + { + "epoch": 4.485017978425889, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11226 + }, + { + "epoch": 4.485417499001199, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11227 + }, + { + "epoch": 4.485817019576508, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11228 + }, + { + "epoch": 4.486216540151818, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11229 + }, + { + "epoch": 4.486616060727127, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11230 + }, + { + "epoch": 4.487015581302437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11231 + }, + { + "epoch": 4.487415101877747, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11232 + }, + { + "epoch": 4.487814622453056, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11233 + }, + { + "epoch": 4.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11234 + }, + { + "epoch": 4.488613663603676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11235 + }, + { + "epoch": 4.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11236 + }, + { + "epoch": 4.489412704754295, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11237 + }, + { + "epoch": 4.4898122253296044, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11238 + }, + { + "epoch": 4.490211745904914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11239 + }, + { + "epoch": 4.490611266480224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11240 + }, + { + "epoch": 4.491010787055533, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11241 + }, + { + "epoch": 4.491410307630843, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 11242 + }, + { + "epoch": 4.4918098282061525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11243 + }, + { + "epoch": 4.492209348781462, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11244 + }, + { + "epoch": 4.492608869356772, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11245 + }, + { + "epoch": 4.493008389932082, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11246 + }, + { + "epoch": 4.493407910507391, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11247 + }, + { + "epoch": 4.4938074310827005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11248 + }, + { + "epoch": 4.49420695165801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11249 + }, + { + "epoch": 4.49460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11250 + }, + { + "epoch": 4.49500599280863, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11251 + }, + { + "epoch": 4.49540551338394, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11252 + }, + { + "epoch": 4.495805033959249, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11253 + }, + { + "epoch": 4.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11254 + }, + { + "epoch": 4.496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11255 + }, + { + "epoch": 4.497003595685178, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11256 + }, + { + "epoch": 4.497403116260488, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11257 + }, + { + "epoch": 4.497802636835797, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 11258 + }, + { + "epoch": 4.498202157411106, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11259 + }, + { + "epoch": 4.498601677986416, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11260 + }, + { + "epoch": 4.499001198561726, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11261 + }, + { + "epoch": 4.499400719137036, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11262 + }, + { + "epoch": 4.4998002397123456, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11263 + }, + { + "epoch": 4.5001997602876544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11264 + }, + { + "epoch": 4.500599280862964, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11265 + }, + { + "epoch": 4.500998801438274, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11266 + }, + { + "epoch": 4.501398322013584, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11267 + }, + { + "epoch": 4.501797842588894, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11268 + }, + { + "epoch": 4.502197363164203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11269 + }, + { + "epoch": 4.502596883739512, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11270 + }, + { + "epoch": 4.502996404314822, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11271 + }, + { + "epoch": 4.503395924890132, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11272 + }, + { + "epoch": 4.503795445465442, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 11273 + }, + { + "epoch": 4.504194966040751, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11274 + }, + { + "epoch": 4.50459448661606, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11275 + }, + { + "epoch": 4.50499400719137, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11276 + }, + { + "epoch": 4.50539352776668, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11277 + }, + { + "epoch": 4.50579304834199, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11278 + }, + { + "epoch": 4.5061925689172995, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11279 + }, + { + "epoch": 4.506592089492609, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11280 + }, + { + "epoch": 4.506991610067918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11281 + }, + { + "epoch": 4.507391130643228, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11282 + }, + { + "epoch": 4.507790651218538, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11283 + }, + { + "epoch": 4.5081901717938475, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11284 + }, + { + "epoch": 4.508589692369157, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11285 + }, + { + "epoch": 4.508989212944467, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11286 + }, + { + "epoch": 4.509388733519776, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11287 + }, + { + "epoch": 4.509788254095086, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11288 + }, + { + "epoch": 4.5101877746703956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11289 + }, + { + "epoch": 4.510587295245705, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11290 + }, + { + "epoch": 4.510986815821015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11291 + }, + { + "epoch": 4.511386336396324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 11292 + }, + { + "epoch": 4.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11293 + }, + { + "epoch": 4.512185377546944, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11294 + }, + { + "epoch": 4.512584898122253, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11295 + }, + { + "epoch": 4.512984418697563, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11296 + }, + { + "epoch": 4.513383939272872, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11297 + }, + { + "epoch": 4.513783459848182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11298 + }, + { + "epoch": 4.514182980423492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11299 + }, + { + "epoch": 4.514582500998801, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11300 + }, + { + "epoch": 4.514982021574111, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11301 + }, + { + "epoch": 4.515381542149421, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.2975, + "step": 11302 + }, + { + "epoch": 4.515781062724731, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11303 + }, + { + "epoch": 4.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11304 + }, + { + "epoch": 4.5165801038753495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 11305 + }, + { + "epoch": 4.516979624450659, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 11306 + }, + { + "epoch": 4.517379145025969, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11307 + }, + { + "epoch": 4.517778665601279, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 11308 + }, + { + "epoch": 4.518178186176588, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11309 + }, + { + "epoch": 4.5185777067518975, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11310 + }, + { + "epoch": 4.518977227327207, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11311 + }, + { + "epoch": 4.519376747902517, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11312 + }, + { + "epoch": 4.519776268477827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11313 + }, + { + "epoch": 4.520175789053136, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11314 + }, + { + "epoch": 4.520575309628446, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11315 + }, + { + "epoch": 4.520974830203755, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 11316 + }, + { + "epoch": 4.521374350779065, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11317 + }, + { + "epoch": 4.521773871354375, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11318 + }, + { + "epoch": 4.522173391929685, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11319 + }, + { + "epoch": 4.5225729125049945, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 11320 + }, + { + "epoch": 4.522972433080303, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11321 + }, + { + "epoch": 4.523371953655613, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11322 + }, + { + "epoch": 4.523771474230923, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11323 + }, + { + "epoch": 4.524170994806233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11324 + }, + { + "epoch": 4.5245705153815425, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11325 + }, + { + "epoch": 4.524970035956851, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11326 + }, + { + "epoch": 4.525369556532161, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11327 + }, + { + "epoch": 4.525769077107471, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11328 + }, + { + "epoch": 4.526168597682781, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11329 + }, + { + "epoch": 4.526568118258091, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11330 + }, + { + "epoch": 4.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11331 + }, + { + "epoch": 4.527367159408709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11332 + }, + { + "epoch": 4.527766679984019, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11333 + }, + { + "epoch": 4.528166200559329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11334 + }, + { + "epoch": 4.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 11335 + }, + { + "epoch": 4.528965241709948, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11336 + }, + { + "epoch": 4.529364762285257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11337 + }, + { + "epoch": 4.529764282860567, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11338 + }, + { + "epoch": 4.530163803435877, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11339 + }, + { + "epoch": 4.530563324011187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11340 + }, + { + "epoch": 4.5309628445864965, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11341 + }, + { + "epoch": 4.531362365161806, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11342 + }, + { + "epoch": 4.531761885737115, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 11343 + }, + { + "epoch": 4.532161406312425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11344 + }, + { + "epoch": 4.532560926887735, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11345 + }, + { + "epoch": 4.5329604474630445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 11346 + }, + { + "epoch": 4.533359968038354, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11347 + }, + { + "epoch": 4.533759488613663, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11348 + }, + { + "epoch": 4.534159009188973, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11349 + }, + { + "epoch": 4.534558529764283, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 11350 + }, + { + "epoch": 4.5349580503395925, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11351 + }, + { + "epoch": 4.535357570914902, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11352 + }, + { + "epoch": 4.535757091490212, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 11353 + }, + { + "epoch": 4.536156612065521, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11354 + }, + { + "epoch": 4.536556132640831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11355 + }, + { + "epoch": 4.536955653216141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11356 + }, + { + "epoch": 4.53735517379145, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11357 + }, + { + "epoch": 4.53775469436676, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11358 + }, + { + "epoch": 4.53815421494207, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11359 + }, + { + "epoch": 4.538553735517379, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11360 + }, + { + "epoch": 4.538953256092689, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11361 + }, + { + "epoch": 4.539352776667998, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11362 + }, + { + "epoch": 4.539752297243308, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11363 + }, + { + "epoch": 4.540151817818618, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11364 + }, + { + "epoch": 4.540551338393927, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11365 + }, + { + "epoch": 4.540950858969237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11366 + }, + { + "epoch": 4.5413503795445465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11367 + }, + { + "epoch": 4.541749900119856, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11368 + }, + { + "epoch": 4.542149420695166, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11369 + }, + { + "epoch": 4.542548941270476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11370 + }, + { + "epoch": 4.542948461845785, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11371 + }, + { + "epoch": 4.5433479824210945, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11372 + }, + { + "epoch": 4.543747502996404, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11373 + }, + { + "epoch": 4.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11374 + }, + { + "epoch": 4.544546544147024, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11375 + }, + { + "epoch": 4.544946064722334, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11376 + }, + { + "epoch": 4.5453455852976425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11377 + }, + { + "epoch": 4.545745105872952, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11378 + }, + { + "epoch": 4.546144626448262, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11379 + }, + { + "epoch": 4.546544147023572, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11380 + }, + { + "epoch": 4.546943667598882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11381 + }, + { + "epoch": 4.547343188174191, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11382 + }, + { + "epoch": 4.5477427087495, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11383 + }, + { + "epoch": 4.54814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11384 + }, + { + "epoch": 4.54854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11385 + }, + { + "epoch": 4.54894127047543, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11386 + }, + { + "epoch": 4.549340791050739, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11387 + }, + { + "epoch": 4.549740311626048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11388 + }, + { + "epoch": 4.550139832201358, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11389 + }, + { + "epoch": 4.550539352776668, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11390 + }, + { + "epoch": 4.550938873351978, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11391 + }, + { + "epoch": 4.551338393927288, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11392 + }, + { + "epoch": 4.551737914502597, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11393 + }, + { + "epoch": 4.552137435077906, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11394 + }, + { + "epoch": 4.552536955653216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11395 + }, + { + "epoch": 4.552936476228526, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11396 + }, + { + "epoch": 4.553335996803836, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11397 + }, + { + "epoch": 4.553735517379145, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11398 + }, + { + "epoch": 4.554135037954454, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11399 + }, + { + "epoch": 4.554534558529764, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11400 + }, + { + "epoch": 4.554934079105074, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 11401 + }, + { + "epoch": 4.555333599680384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11402 + }, + { + "epoch": 4.555733120255693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11403 + }, + { + "epoch": 4.556132640831002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11404 + }, + { + "epoch": 4.556532161406312, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11405 + }, + { + "epoch": 4.556931681981622, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11406 + }, + { + "epoch": 4.557331202556932, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11407 + }, + { + "epoch": 4.5577307231322415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11408 + }, + { + "epoch": 4.558130243707551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 11409 + }, + { + "epoch": 4.558529764282861, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11410 + }, + { + "epoch": 4.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11411 + }, + { + "epoch": 4.55932880543348, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11412 + }, + { + "epoch": 4.5597283260087895, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11413 + }, + { + "epoch": 4.560127846584099, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11414 + }, + { + "epoch": 4.560527367159409, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 11415 + }, + { + "epoch": 4.560926887734718, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11416 + }, + { + "epoch": 4.561326408310028, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11417 + }, + { + "epoch": 4.561725928885338, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11418 + }, + { + "epoch": 4.562125449460647, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11419 + }, + { + "epoch": 4.562524970035957, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11420 + }, + { + "epoch": 4.562924490611266, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11421 + }, + { + "epoch": 4.563324011186576, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11422 + }, + { + "epoch": 4.563723531761886, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11423 + }, + { + "epoch": 4.564123052337195, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11424 + }, + { + "epoch": 4.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11425 + }, + { + "epoch": 4.564922093487815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11426 + }, + { + "epoch": 4.565321614063124, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11427 + }, + { + "epoch": 4.565721134638434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11428 + }, + { + "epoch": 4.566120655213743, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11429 + }, + { + "epoch": 4.566520175789053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11430 + }, + { + "epoch": 4.566919696364363, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11431 + }, + { + "epoch": 4.567319216939673, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11432 + }, + { + "epoch": 4.567718737514982, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11433 + }, + { + "epoch": 4.5681182580902915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11434 + }, + { + "epoch": 4.568517778665601, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11435 + }, + { + "epoch": 4.568917299240911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11436 + }, + { + "epoch": 4.569316819816221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11437 + }, + { + "epoch": 4.56971634039153, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11438 + }, + { + "epoch": 4.5701158609668395, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11439 + }, + { + "epoch": 4.570515381542149, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11440 + }, + { + "epoch": 4.570914902117459, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11441 + }, + { + "epoch": 4.571314422692769, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11442 + }, + { + "epoch": 4.571713943268079, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11443 + }, + { + "epoch": 4.572113463843388, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11444 + }, + { + "epoch": 4.572512984418697, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11445 + }, + { + "epoch": 4.572912504994007, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11446 + }, + { + "epoch": 4.573312025569317, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11447 + }, + { + "epoch": 4.573711546144627, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11448 + }, + { + "epoch": 4.5741110667199365, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11449 + }, + { + "epoch": 4.574510587295245, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11450 + }, + { + "epoch": 4.574910107870555, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11451 + }, + { + "epoch": 4.575309628445865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11452 + }, + { + "epoch": 4.575709149021175, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11453 + }, + { + "epoch": 4.5761086695964845, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11454 + }, + { + "epoch": 4.5765081901717934, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11455 + }, + { + "epoch": 4.576907710747103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11456 + }, + { + "epoch": 4.577307231322413, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11457 + }, + { + "epoch": 4.577706751897723, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11458 + }, + { + "epoch": 4.578106272473033, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11459 + }, + { + "epoch": 4.578505793048342, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11460 + }, + { + "epoch": 4.578905313623651, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11461 + }, + { + "epoch": 4.579304834198961, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11462 + }, + { + "epoch": 4.579704354774271, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11463 + }, + { + "epoch": 4.580103875349581, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11464 + }, + { + "epoch": 4.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11465 + }, + { + "epoch": 4.5809029165002, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11466 + }, + { + "epoch": 4.581302437075509, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11467 + }, + { + "epoch": 4.581701957650819, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11468 + }, + { + "epoch": 4.582101478226129, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11469 + }, + { + "epoch": 4.5825009988014385, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11470 + }, + { + "epoch": 4.582900519376748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11471 + }, + { + "epoch": 4.583300039952057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 11472 + }, + { + "epoch": 4.583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11473 + }, + { + "epoch": 4.584099081102677, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11474 + }, + { + "epoch": 4.5844986016779865, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11475 + }, + { + "epoch": 4.584898122253296, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11476 + }, + { + "epoch": 4.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11477 + }, + { + "epoch": 4.585697163403915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11478 + }, + { + "epoch": 4.586096683979225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11479 + }, + { + "epoch": 4.5864962045545346, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11480 + }, + { + "epoch": 4.586895725129844, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11481 + }, + { + "epoch": 4.587295245705154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11482 + }, + { + "epoch": 4.587694766280464, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11483 + }, + { + "epoch": 4.588094286855773, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11484 + }, + { + "epoch": 4.588493807431083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 11485 + }, + { + "epoch": 4.588893328006392, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11486 + }, + { + "epoch": 4.589292848581702, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 11487 + }, + { + "epoch": 4.589692369157012, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11488 + }, + { + "epoch": 4.590091889732321, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11489 + }, + { + "epoch": 4.590491410307631, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11490 + }, + { + "epoch": 4.59089093088294, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 11491 + }, + { + "epoch": 4.59129045145825, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11492 + }, + { + "epoch": 4.59168997203356, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11493 + }, + { + "epoch": 4.592089492608869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11494 + }, + { + "epoch": 4.592489013184179, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11495 + }, + { + "epoch": 4.5928885337594885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 11496 + }, + { + "epoch": 4.593288054334798, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11497 + }, + { + "epoch": 4.593687574910108, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11498 + }, + { + "epoch": 4.594087095485418, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11499 + }, + { + "epoch": 4.594486616060728, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11500 + }, + { + "epoch": 4.5948861366360365, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11501 + }, + { + "epoch": 4.595285657211346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11502 + }, + { + "epoch": 4.595685177786656, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11503 + }, + { + "epoch": 4.596084698361966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11504 + }, + { + "epoch": 4.596484218937276, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11505 + }, + { + "epoch": 4.5968837395125846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11506 + }, + { + "epoch": 4.597283260087894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11507 + }, + { + "epoch": 4.597682780663204, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11508 + }, + { + "epoch": 4.598082301238514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11509 + }, + { + "epoch": 4.598481821813824, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11510 + }, + { + "epoch": 4.598881342389133, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11511 + }, + { + "epoch": 4.599280862964442, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11512 + }, + { + "epoch": 4.599680383539752, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11513 + }, + { + "epoch": 4.600079904115062, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11514 + }, + { + "epoch": 4.600479424690372, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11515 + }, + { + "epoch": 4.6008789452656815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11516 + }, + { + "epoch": 4.60127846584099, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11517 + }, + { + "epoch": 4.6016779864163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11518 + }, + { + "epoch": 4.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11519 + }, + { + "epoch": 4.60247702756692, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11520 + }, + { + "epoch": 4.60287654814223, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11521 + }, + { + "epoch": 4.603276068717539, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11522 + }, + { + "epoch": 4.603675589292848, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11523 + }, + { + "epoch": 4.604075109868158, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11524 + }, + { + "epoch": 4.604474630443468, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11525 + }, + { + "epoch": 4.604874151018778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11526 + }, + { + "epoch": 4.605273671594087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11527 + }, + { + "epoch": 4.605673192169396, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11528 + }, + { + "epoch": 4.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11529 + }, + { + "epoch": 4.606472233320016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11530 + }, + { + "epoch": 4.606871753895326, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11531 + }, + { + "epoch": 4.6072712744706354, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 11532 + }, + { + "epoch": 4.607670795045945, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11533 + }, + { + "epoch": 4.608070315621254, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11534 + }, + { + "epoch": 4.608469836196564, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11535 + }, + { + "epoch": 4.608869356771874, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11536 + }, + { + "epoch": 4.6092688773471835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11537 + }, + { + "epoch": 4.609668397922493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11538 + }, + { + "epoch": 4.610067918497803, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11539 + }, + { + "epoch": 4.610467439073112, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11540 + }, + { + "epoch": 4.610866959648422, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11541 + }, + { + "epoch": 4.6112664802237315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11542 + }, + { + "epoch": 4.611666000799041, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11543 + }, + { + "epoch": 4.612065521374351, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11544 + }, + { + "epoch": 4.61246504194966, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11545 + }, + { + "epoch": 4.61286456252497, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11546 + }, + { + "epoch": 4.61326408310028, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11547 + }, + { + "epoch": 4.613663603675589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11548 + }, + { + "epoch": 4.614063124250899, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11549 + }, + { + "epoch": 4.614462644826209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11550 + }, + { + "epoch": 4.614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11551 + }, + { + "epoch": 4.615261685976828, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 11552 + }, + { + "epoch": 4.615661206552137, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11553 + }, + { + "epoch": 4.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11554 + }, + { + "epoch": 4.616460247702757, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11555 + }, + { + "epoch": 4.616859768278067, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11556 + }, + { + "epoch": 4.617259288853376, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11557 + }, + { + "epoch": 4.6176588094286855, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11558 + }, + { + "epoch": 4.618058330003995, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11559 + }, + { + "epoch": 4.618457850579305, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11560 + }, + { + "epoch": 4.618857371154615, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11561 + }, + { + "epoch": 4.619256891729924, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11562 + }, + { + "epoch": 4.6196564123052335, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11563 + }, + { + "epoch": 4.620055932880543, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11564 + }, + { + "epoch": 4.620455453455853, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 11565 + }, + { + "epoch": 4.620854974031163, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11566 + }, + { + "epoch": 4.621254494606472, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11567 + }, + { + "epoch": 4.6216540151817815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11568 + }, + { + "epoch": 4.622053535757091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11569 + }, + { + "epoch": 4.622453056332401, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11570 + }, + { + "epoch": 4.622852576907711, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11571 + }, + { + "epoch": 4.623252097483021, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 11572 + }, + { + "epoch": 4.6236516180583305, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11573 + }, + { + "epoch": 4.624051138633639, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11574 + }, + { + "epoch": 4.624450659208949, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11575 + }, + { + "epoch": 4.624850179784259, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11576 + }, + { + "epoch": 4.625249700359569, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11577 + }, + { + "epoch": 4.6256492209348785, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11578 + }, + { + "epoch": 4.626048741510187, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11579 + }, + { + "epoch": 4.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11580 + }, + { + "epoch": 4.626847782660807, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11581 + }, + { + "epoch": 4.627247303236117, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11582 + }, + { + "epoch": 4.627646823811427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 11583 + }, + { + "epoch": 4.6280463443867355, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11584 + }, + { + "epoch": 4.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11585 + }, + { + "epoch": 4.628845385537355, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11586 + }, + { + "epoch": 4.629244906112665, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11587 + }, + { + "epoch": 4.629644426687975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11588 + }, + { + "epoch": 4.630043947263284, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11589 + }, + { + "epoch": 4.630443467838594, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11590 + }, + { + "epoch": 4.630842988413903, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 11591 + }, + { + "epoch": 4.631242508989213, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11592 + }, + { + "epoch": 4.631642029564523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11593 + }, + { + "epoch": 4.632041550139832, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11594 + }, + { + "epoch": 4.632441070715142, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 11595 + }, + { + "epoch": 4.632840591290451, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11596 + }, + { + "epoch": 4.633240111865761, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11597 + }, + { + "epoch": 4.633639632441071, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11598 + }, + { + "epoch": 4.6340391530163805, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11599 + }, + { + "epoch": 4.63443867359169, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11600 + }, + { + "epoch": 4.634838194166999, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11601 + }, + { + "epoch": 4.635237714742309, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11602 + }, + { + "epoch": 4.635637235317619, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11603 + }, + { + "epoch": 4.6360367558929285, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11604 + }, + { + "epoch": 4.636436276468238, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11605 + }, + { + "epoch": 4.636835797043548, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11606 + }, + { + "epoch": 4.637235317618857, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11607 + }, + { + "epoch": 4.637634838194167, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11608 + }, + { + "epoch": 4.638034358769477, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11609 + }, + { + "epoch": 4.638433879344786, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 11610 + }, + { + "epoch": 4.638833399920096, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11611 + }, + { + "epoch": 4.639232920495406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11612 + }, + { + "epoch": 4.639632441070715, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11613 + }, + { + "epoch": 4.640031961646025, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11614 + }, + { + "epoch": 4.640431482221334, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11615 + }, + { + "epoch": 4.640831002796644, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11616 + }, + { + "epoch": 4.641230523371954, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11617 + }, + { + "epoch": 4.641630043947263, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11618 + }, + { + "epoch": 4.642029564522573, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11619 + }, + { + "epoch": 4.642429085097882, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11620 + }, + { + "epoch": 4.642828605673192, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11621 + }, + { + "epoch": 4.643228126248502, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11622 + }, + { + "epoch": 4.643627646823812, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11623 + }, + { + "epoch": 4.644027167399121, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11624 + }, + { + "epoch": 4.6444266879744305, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11625 + }, + { + "epoch": 4.64482620854974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11626 + }, + { + "epoch": 4.64522572912505, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11627 + }, + { + "epoch": 4.64562524970036, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11628 + }, + { + "epoch": 4.64602477027567, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11629 + }, + { + "epoch": 4.6464242908509785, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11630 + }, + { + "epoch": 4.646823811426288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11631 + }, + { + "epoch": 4.647223332001598, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11632 + }, + { + "epoch": 4.647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11633 + }, + { + "epoch": 4.648022373152218, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11634 + }, + { + "epoch": 4.648421893727527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11635 + }, + { + "epoch": 4.648821414302836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11636 + }, + { + "epoch": 4.649220934878146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11637 + }, + { + "epoch": 4.649620455453456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11638 + }, + { + "epoch": 4.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11639 + }, + { + "epoch": 4.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11640 + }, + { + "epoch": 4.650819017179384, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11641 + }, + { + "epoch": 4.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 11642 + }, + { + "epoch": 4.651618058330004, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11643 + }, + { + "epoch": 4.652017578905314, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11644 + }, + { + "epoch": 4.6524170994806235, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11645 + }, + { + "epoch": 4.652816620055933, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11646 + }, + { + "epoch": 4.653216140631242, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11647 + }, + { + "epoch": 4.653615661206552, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11648 + }, + { + "epoch": 4.654015181781862, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11649 + }, + { + "epoch": 4.654414702357172, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11650 + }, + { + "epoch": 4.654814222932481, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11651 + }, + { + "epoch": 4.65521374350779, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11652 + }, + { + "epoch": 4.6556132640831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11653 + }, + { + "epoch": 4.65601278465841, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11654 + }, + { + "epoch": 4.65641230523372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11655 + }, + { + "epoch": 4.656811825809029, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11656 + }, + { + "epoch": 4.657211346384338, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11657 + }, + { + "epoch": 4.657610866959648, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11658 + }, + { + "epoch": 4.658010387534958, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11659 + }, + { + "epoch": 4.658409908110268, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11660 + }, + { + "epoch": 4.6588094286855775, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11661 + }, + { + "epoch": 4.659208949260887, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11662 + }, + { + "epoch": 4.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11663 + }, + { + "epoch": 4.660007990411506, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11664 + }, + { + "epoch": 4.660407510986816, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11665 + }, + { + "epoch": 4.6608070315621255, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11666 + }, + { + "epoch": 4.661206552137435, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11667 + }, + { + "epoch": 4.661606072712745, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11668 + }, + { + "epoch": 4.662005593288054, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11669 + }, + { + "epoch": 4.662405113863364, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11670 + }, + { + "epoch": 4.6628046344386735, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11671 + }, + { + "epoch": 4.663204155013983, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11672 + }, + { + "epoch": 4.663603675589293, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11673 + }, + { + "epoch": 4.664003196164602, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11674 + }, + { + "epoch": 4.664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11675 + }, + { + "epoch": 4.664802237315222, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11676 + }, + { + "epoch": 4.665201757890531, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11677 + }, + { + "epoch": 4.665601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11678 + }, + { + "epoch": 4.666000799041151, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11679 + }, + { + "epoch": 4.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11680 + }, + { + "epoch": 4.66679984019177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11681 + }, + { + "epoch": 4.667199360767079, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11682 + }, + { + "epoch": 4.667598881342389, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11683 + }, + { + "epoch": 4.667998401917699, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11684 + }, + { + "epoch": 4.668397922493009, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11685 + }, + { + "epoch": 4.668797443068318, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11686 + }, + { + "epoch": 4.6691969636436275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11687 + }, + { + "epoch": 4.669596484218937, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11688 + }, + { + "epoch": 4.669996004794247, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11689 + }, + { + "epoch": 4.670395525369557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11690 + }, + { + "epoch": 4.670795045944866, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11691 + }, + { + "epoch": 4.6711945665201755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11692 + }, + { + "epoch": 4.671594087095485, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11693 + }, + { + "epoch": 4.671993607670795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11694 + }, + { + "epoch": 4.672393128246105, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11695 + }, + { + "epoch": 4.672792648821415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11696 + }, + { + "epoch": 4.6731921693967235, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11697 + }, + { + "epoch": 4.673591689972033, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11698 + }, + { + "epoch": 4.673991210547343, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11699 + }, + { + "epoch": 4.674390731122653, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 11700 + }, + { + "epoch": 4.674790251697963, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11701 + }, + { + "epoch": 4.6751897722732725, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11702 + }, + { + "epoch": 4.675589292848581, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11703 + }, + { + "epoch": 4.675988813423891, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11704 + }, + { + "epoch": 4.676388333999201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11705 + }, + { + "epoch": 4.676787854574511, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 11706 + }, + { + "epoch": 4.6771873751498205, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11707 + }, + { + "epoch": 4.677586895725129, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11708 + }, + { + "epoch": 4.677986416300439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11709 + }, + { + "epoch": 4.678385936875749, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11710 + }, + { + "epoch": 4.678785457451059, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11711 + }, + { + "epoch": 4.679184978026369, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11712 + }, + { + "epoch": 4.679584498601678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11713 + }, + { + "epoch": 4.679984019176987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11714 + }, + { + "epoch": 4.680383539752297, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11715 + }, + { + "epoch": 4.680783060327607, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11716 + }, + { + "epoch": 4.681182580902917, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11717 + }, + { + "epoch": 4.681582101478226, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11718 + }, + { + "epoch": 4.681981622053536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11719 + }, + { + "epoch": 4.682381142628845, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11720 + }, + { + "epoch": 4.682780663204155, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11721 + }, + { + "epoch": 4.683180183779465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11722 + }, + { + "epoch": 4.683579704354774, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 11723 + }, + { + "epoch": 4.683979224930084, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11724 + }, + { + "epoch": 4.684378745505393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11725 + }, + { + "epoch": 4.684778266080703, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11726 + }, + { + "epoch": 4.685177786656013, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11727 + }, + { + "epoch": 4.6855773072313225, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11728 + }, + { + "epoch": 4.685976827806632, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11729 + }, + { + "epoch": 4.686376348381942, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11730 + }, + { + "epoch": 4.686775868957251, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11731 + }, + { + "epoch": 4.687175389532561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11732 + }, + { + "epoch": 4.6875749101078705, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11733 + }, + { + "epoch": 4.68797443068318, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11734 + }, + { + "epoch": 4.68837395125849, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11735 + }, + { + "epoch": 4.6887734718338, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11736 + }, + { + "epoch": 4.689172992409109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11737 + }, + { + "epoch": 4.689572512984419, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11738 + }, + { + "epoch": 4.689972033559728, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11739 + }, + { + "epoch": 4.690371554135038, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 11740 + }, + { + "epoch": 4.690771074710348, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11741 + }, + { + "epoch": 4.691170595285657, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 11742 + }, + { + "epoch": 4.691570115860967, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11743 + }, + { + "epoch": 4.691969636436276, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11744 + }, + { + "epoch": 4.692369157011586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11745 + }, + { + "epoch": 4.692768677586896, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11746 + }, + { + "epoch": 4.693168198162205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11747 + }, + { + "epoch": 4.693567718737515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11748 + }, + { + "epoch": 4.6939672393128244, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 11749 + }, + { + "epoch": 4.694366759888134, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11750 + }, + { + "epoch": 4.694766280463444, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11751 + }, + { + "epoch": 4.695165801038754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11752 + }, + { + "epoch": 4.695565321614064, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11753 + }, + { + "epoch": 4.6959648421893725, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11754 + }, + { + "epoch": 4.696364362764682, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11755 + }, + { + "epoch": 4.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11756 + }, + { + "epoch": 4.697163403915302, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11757 + }, + { + "epoch": 4.697562924490612, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11758 + }, + { + "epoch": 4.6979624450659205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11759 + }, + { + "epoch": 4.69836196564123, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11760 + }, + { + "epoch": 4.69876148621654, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11761 + }, + { + "epoch": 4.69916100679185, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 11762 + }, + { + "epoch": 4.69956052736716, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11763 + }, + { + "epoch": 4.699960047942469, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11764 + }, + { + "epoch": 4.700359568517778, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11765 + }, + { + "epoch": 4.700759089093088, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11766 + }, + { + "epoch": 4.701158609668398, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11767 + }, + { + "epoch": 4.701558130243708, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11768 + }, + { + "epoch": 4.7019576508190175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11769 + }, + { + "epoch": 4.702357171394327, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11770 + }, + { + "epoch": 4.702756691969636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11771 + }, + { + "epoch": 4.703156212544946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11772 + }, + { + "epoch": 4.703555733120256, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 11773 + }, + { + "epoch": 4.7039552536955656, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11774 + }, + { + "epoch": 4.704354774270875, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11775 + }, + { + "epoch": 4.704754294846184, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11776 + }, + { + "epoch": 4.705153815421494, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11777 + }, + { + "epoch": 4.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11778 + }, + { + "epoch": 4.705952856572114, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11779 + }, + { + "epoch": 4.706352377147423, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11780 + }, + { + "epoch": 4.706751897722732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11781 + }, + { + "epoch": 4.707151418298042, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 11782 + }, + { + "epoch": 4.707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11783 + }, + { + "epoch": 4.707950459448662, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11784 + }, + { + "epoch": 4.708349980023971, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11785 + }, + { + "epoch": 4.708749500599281, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11786 + }, + { + "epoch": 4.70914902117459, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11787 + }, + { + "epoch": 4.7095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11788 + }, + { + "epoch": 4.70994806232521, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11789 + }, + { + "epoch": 4.7103475829005195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11790 + }, + { + "epoch": 4.710747103475829, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11791 + }, + { + "epoch": 4.711146624051139, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11792 + }, + { + "epoch": 4.711546144626448, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11793 + }, + { + "epoch": 4.711945665201758, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11794 + }, + { + "epoch": 4.7123451857770675, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 11795 + }, + { + "epoch": 4.712744706352377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11796 + }, + { + "epoch": 4.713144226927687, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11797 + }, + { + "epoch": 4.713543747502996, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11798 + }, + { + "epoch": 4.713943268078306, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11799 + }, + { + "epoch": 4.714342788653616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11800 + }, + { + "epoch": 4.714742309228925, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 11801 + }, + { + "epoch": 4.715141829804235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11802 + }, + { + "epoch": 4.715541350379545, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11803 + }, + { + "epoch": 4.715940870954854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11804 + }, + { + "epoch": 4.716340391530164, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11805 + }, + { + "epoch": 4.716739912105473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11806 + }, + { + "epoch": 4.717139432680783, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11807 + }, + { + "epoch": 4.717538953256093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11808 + }, + { + "epoch": 4.717938473831403, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11809 + }, + { + "epoch": 4.718337994406712, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11810 + }, + { + "epoch": 4.718737514982021, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11811 + }, + { + "epoch": 4.719137035557331, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11812 + }, + { + "epoch": 4.719536556132641, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11813 + }, + { + "epoch": 4.719936076707951, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11814 + }, + { + "epoch": 4.72033559728326, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11815 + }, + { + "epoch": 4.7207351178585695, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11816 + }, + { + "epoch": 4.721134638433879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11817 + }, + { + "epoch": 4.721534159009189, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11818 + }, + { + "epoch": 4.721933679584499, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11819 + }, + { + "epoch": 4.722333200159808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11820 + }, + { + "epoch": 4.7227327207351175, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11821 + }, + { + "epoch": 4.723132241310427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 11822 + }, + { + "epoch": 4.723531761885737, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11823 + }, + { + "epoch": 4.723931282461047, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11824 + }, + { + "epoch": 4.724330803036357, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11825 + }, + { + "epoch": 4.7247303236116664, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11826 + }, + { + "epoch": 4.725129844186975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11827 + }, + { + "epoch": 4.725529364762285, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11828 + }, + { + "epoch": 4.725928885337595, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11829 + }, + { + "epoch": 4.726328405912905, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11830 + }, + { + "epoch": 4.7267279264882145, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11831 + }, + { + "epoch": 4.727127447063523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11832 + }, + { + "epoch": 4.727526967638833, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11833 + }, + { + "epoch": 4.727926488214143, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11834 + }, + { + "epoch": 4.728326008789453, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11835 + }, + { + "epoch": 4.7287255293647625, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 11836 + }, + { + "epoch": 4.729125049940071, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11837 + }, + { + "epoch": 4.729524570515381, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11838 + }, + { + "epoch": 4.729924091090691, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11839 + }, + { + "epoch": 4.730323611666001, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11840 + }, + { + "epoch": 4.730723132241311, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11841 + }, + { + "epoch": 4.73112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11842 + }, + { + "epoch": 4.73152217339193, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11843 + }, + { + "epoch": 4.731921693967239, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11844 + }, + { + "epoch": 4.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11845 + }, + { + "epoch": 4.732720735117859, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11846 + }, + { + "epoch": 4.733120255693168, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11847 + }, + { + "epoch": 4.733519776268478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11848 + }, + { + "epoch": 4.733919296843787, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11849 + }, + { + "epoch": 4.734318817419097, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11850 + }, + { + "epoch": 4.734718337994407, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 11851 + }, + { + "epoch": 4.7351178585697165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11852 + }, + { + "epoch": 4.735517379145026, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11853 + }, + { + "epoch": 4.735916899720335, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11854 + }, + { + "epoch": 4.736316420295645, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11855 + }, + { + "epoch": 4.736715940870955, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11856 + }, + { + "epoch": 4.7371154614462645, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11857 + }, + { + "epoch": 4.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11858 + }, + { + "epoch": 4.737914502596884, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 11859 + }, + { + "epoch": 4.738314023172194, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11860 + }, + { + "epoch": 4.738713543747503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11861 + }, + { + "epoch": 4.7391130643228125, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11862 + }, + { + "epoch": 4.739512584898122, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11863 + }, + { + "epoch": 4.739912105473432, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11864 + }, + { + "epoch": 4.740311626048742, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11865 + }, + { + "epoch": 4.740711146624051, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11866 + }, + { + "epoch": 4.741110667199361, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11867 + }, + { + "epoch": 4.74151018777467, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11868 + }, + { + "epoch": 4.74190970834998, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11869 + }, + { + "epoch": 4.74230922892529, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11870 + }, + { + "epoch": 4.742708749500599, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11871 + }, + { + "epoch": 4.743108270075909, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11872 + }, + { + "epoch": 4.743507790651218, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11873 + }, + { + "epoch": 4.743907311226528, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11874 + }, + { + "epoch": 4.744306831801838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11875 + }, + { + "epoch": 4.744706352377148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11876 + }, + { + "epoch": 4.745105872952457, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11877 + }, + { + "epoch": 4.7455053935277665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11878 + }, + { + "epoch": 4.745904914103076, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11879 + }, + { + "epoch": 4.746304434678386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11880 + }, + { + "epoch": 4.746703955253696, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11881 + }, + { + "epoch": 4.747103475829006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11882 + }, + { + "epoch": 4.7475029964043145, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11883 + }, + { + "epoch": 4.747902516979624, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11884 + }, + { + "epoch": 4.748302037554934, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11885 + }, + { + "epoch": 4.748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11886 + }, + { + "epoch": 4.749101078705554, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11887 + }, + { + "epoch": 4.7495005992808625, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11888 + }, + { + "epoch": 4.749900119856172, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11889 + }, + { + "epoch": 4.750299640431482, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11890 + }, + { + "epoch": 4.750699161006792, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11891 + }, + { + "epoch": 4.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11892 + }, + { + "epoch": 4.7514982021574115, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11893 + }, + { + "epoch": 4.75189772273272, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11894 + }, + { + "epoch": 4.75229724330803, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11895 + }, + { + "epoch": 4.75269676388334, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11896 + }, + { + "epoch": 4.75309628445865, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11897 + }, + { + "epoch": 4.7534958050339595, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11898 + }, + { + "epoch": 4.753895325609269, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11899 + }, + { + "epoch": 4.754294846184578, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11900 + }, + { + "epoch": 4.754694366759888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11901 + }, + { + "epoch": 4.755093887335198, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11902 + }, + { + "epoch": 4.755493407910508, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11903 + }, + { + "epoch": 4.755892928485817, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11904 + }, + { + "epoch": 4.756292449061126, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11905 + }, + { + "epoch": 4.756691969636436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11906 + }, + { + "epoch": 4.757091490211746, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11907 + }, + { + "epoch": 4.757491010787056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11908 + }, + { + "epoch": 4.757890531362365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11909 + }, + { + "epoch": 4.758290051937674, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11910 + }, + { + "epoch": 4.758689572512984, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11911 + }, + { + "epoch": 4.759089093088294, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11912 + }, + { + "epoch": 4.759488613663604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11913 + }, + { + "epoch": 4.759888134238913, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11914 + }, + { + "epoch": 4.760287654814223, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11915 + }, + { + "epoch": 4.760687175389533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11916 + }, + { + "epoch": 4.761086695964842, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11917 + }, + { + "epoch": 4.761486216540152, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11918 + }, + { + "epoch": 4.7618857371154615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11919 + }, + { + "epoch": 4.762285257690771, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 11920 + }, + { + "epoch": 4.762684778266081, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11921 + }, + { + "epoch": 4.76308429884139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11922 + }, + { + "epoch": 4.7634838194167, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 11923 + }, + { + "epoch": 4.7638833399920095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11924 + }, + { + "epoch": 4.764282860567319, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11925 + }, + { + "epoch": 4.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11926 + }, + { + "epoch": 4.765081901717938, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11927 + }, + { + "epoch": 4.765481422293248, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11928 + }, + { + "epoch": 4.765880942868558, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11929 + }, + { + "epoch": 4.766280463443867, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11930 + }, + { + "epoch": 4.766679984019177, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11931 + }, + { + "epoch": 4.767079504594487, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11932 + }, + { + "epoch": 4.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11933 + }, + { + "epoch": 4.767878545745106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11934 + }, + { + "epoch": 4.768278066320415, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11935 + }, + { + "epoch": 4.768677586895725, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11936 + }, + { + "epoch": 4.769077107471035, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11937 + }, + { + "epoch": 4.769476628046345, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11938 + }, + { + "epoch": 4.769876148621654, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11939 + }, + { + "epoch": 4.770275669196963, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11940 + }, + { + "epoch": 4.770675189772273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11941 + }, + { + "epoch": 4.771074710347583, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11942 + }, + { + "epoch": 4.771474230922893, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11943 + }, + { + "epoch": 4.771873751498202, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11944 + }, + { + "epoch": 4.7722732720735115, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11945 + }, + { + "epoch": 4.772672792648821, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11946 + }, + { + "epoch": 4.773072313224131, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11947 + }, + { + "epoch": 4.773471833799441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11948 + }, + { + "epoch": 4.773871354374751, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11949 + }, + { + "epoch": 4.77427087495006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11950 + }, + { + "epoch": 4.774670395525369, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11951 + }, + { + "epoch": 4.775069916100679, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 11952 + }, + { + "epoch": 4.775469436675989, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11953 + }, + { + "epoch": 4.775868957251299, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11954 + }, + { + "epoch": 4.7762684778266085, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11955 + }, + { + "epoch": 4.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11956 + }, + { + "epoch": 4.777067518977227, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11957 + }, + { + "epoch": 4.777467039552537, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11958 + }, + { + "epoch": 4.777866560127847, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11959 + }, + { + "epoch": 4.7782660807031565, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11960 + }, + { + "epoch": 4.778665601278465, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11961 + }, + { + "epoch": 4.779065121853775, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11962 + }, + { + "epoch": 4.779464642429085, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11963 + }, + { + "epoch": 4.779864163004395, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11964 + }, + { + "epoch": 4.7802636835797045, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11965 + }, + { + "epoch": 4.780663204155014, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11966 + }, + { + "epoch": 4.781062724730323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 11967 + }, + { + "epoch": 4.781462245305633, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11968 + }, + { + "epoch": 4.781861765880943, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11969 + }, + { + "epoch": 4.782261286456253, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11970 + }, + { + "epoch": 4.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11971 + }, + { + "epoch": 4.783060327606872, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11972 + }, + { + "epoch": 4.783459848182181, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11973 + }, + { + "epoch": 4.783859368757491, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11974 + }, + { + "epoch": 4.784258889332801, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 11975 + }, + { + "epoch": 4.78465840990811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11976 + }, + { + "epoch": 4.78505793048342, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11977 + }, + { + "epoch": 4.785457451058729, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11978 + }, + { + "epoch": 4.785856971634039, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11979 + }, + { + "epoch": 4.786256492209349, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11980 + }, + { + "epoch": 4.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11981 + }, + { + "epoch": 4.787055533359968, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11982 + }, + { + "epoch": 4.787455053935278, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11983 + }, + { + "epoch": 4.787854574510587, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11984 + }, + { + "epoch": 4.788254095085897, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11985 + }, + { + "epoch": 4.7886536156612065, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11986 + }, + { + "epoch": 4.789053136236516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11987 + }, + { + "epoch": 4.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11988 + }, + { + "epoch": 4.789852177387136, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11989 + }, + { + "epoch": 4.790251697962445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11990 + }, + { + "epoch": 4.7906512185377546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11991 + }, + { + "epoch": 4.791050739113064, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11992 + }, + { + "epoch": 4.791450259688374, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11993 + }, + { + "epoch": 4.791849780263684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11994 + }, + { + "epoch": 4.792249300838993, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11995 + }, + { + "epoch": 4.792648821414303, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11996 + }, + { + "epoch": 4.793048341989612, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11997 + }, + { + "epoch": 4.793447862564922, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11998 + }, + { + "epoch": 4.793847383140232, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 11999 + }, + { + "epoch": 4.794246903715541, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12000 + }, + { + "epoch": 4.794646424290851, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12001 + }, + { + "epoch": 4.79504594486616, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12002 + }, + { + "epoch": 4.79544546544147, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12003 + }, + { + "epoch": 4.79584498601678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12004 + }, + { + "epoch": 4.79624450659209, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12005 + }, + { + "epoch": 4.7966440271674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12006 + }, + { + "epoch": 4.7970435477427085, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12007 + }, + { + "epoch": 4.797443068318018, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12008 + }, + { + "epoch": 4.797842588893328, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12009 + }, + { + "epoch": 4.798242109468638, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12010 + }, + { + "epoch": 4.798641630043948, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12011 + }, + { + "epoch": 4.7990411506192565, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12012 + }, + { + "epoch": 4.799440671194566, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 12013 + }, + { + "epoch": 4.799840191769876, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12014 + }, + { + "epoch": 4.800239712345186, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12015 + }, + { + "epoch": 4.800639232920496, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12016 + }, + { + "epoch": 4.801038753495805, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12017 + }, + { + "epoch": 4.801438274071114, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12018 + }, + { + "epoch": 4.801837794646424, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12019 + }, + { + "epoch": 4.802237315221734, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12020 + }, + { + "epoch": 4.802636835797044, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12021 + }, + { + "epoch": 4.8030363563723535, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12022 + }, + { + "epoch": 4.803435876947663, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12023 + }, + { + "epoch": 4.803835397522972, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12024 + }, + { + "epoch": 4.804234918098282, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12025 + }, + { + "epoch": 4.804634438673592, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12026 + }, + { + "epoch": 4.8050339592489015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12027 + }, + { + "epoch": 4.805433479824211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12028 + }, + { + "epoch": 4.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12029 + }, + { + "epoch": 4.80623252097483, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12030 + }, + { + "epoch": 4.80663204155014, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12031 + }, + { + "epoch": 4.80703156212545, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12032 + }, + { + "epoch": 4.807431082700759, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12033 + }, + { + "epoch": 4.807830603276068, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12034 + }, + { + "epoch": 4.808230123851378, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12035 + }, + { + "epoch": 4.808629644426688, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12036 + }, + { + "epoch": 4.809029165001998, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12037 + }, + { + "epoch": 4.809428685577307, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12038 + }, + { + "epoch": 4.809828206152617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12039 + }, + { + "epoch": 4.810227726727927, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12040 + }, + { + "epoch": 4.810627247303236, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12041 + }, + { + "epoch": 4.811026767878546, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12042 + }, + { + "epoch": 4.8114262884538554, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12043 + }, + { + "epoch": 4.811825809029165, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12044 + }, + { + "epoch": 4.812225329604475, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 12045 + }, + { + "epoch": 4.812624850179784, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12046 + }, + { + "epoch": 4.813024370755094, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12047 + }, + { + "epoch": 4.8134238913304035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12048 + }, + { + "epoch": 4.813823411905713, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12049 + }, + { + "epoch": 4.814222932481023, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12050 + }, + { + "epoch": 4.814622453056332, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12051 + }, + { + "epoch": 4.815021973631642, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12052 + }, + { + "epoch": 4.8154214942069515, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12053 + }, + { + "epoch": 4.815821014782261, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12054 + }, + { + "epoch": 4.816220535357571, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12055 + }, + { + "epoch": 4.816620055932881, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12056 + }, + { + "epoch": 4.81701957650819, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12057 + }, + { + "epoch": 4.8174190970835, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12058 + }, + { + "epoch": 4.817818617658809, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12059 + }, + { + "epoch": 4.818218138234119, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12060 + }, + { + "epoch": 4.818617658809429, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12061 + }, + { + "epoch": 4.819017179384739, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12062 + }, + { + "epoch": 4.819416699960048, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12063 + }, + { + "epoch": 4.819816220535357, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12064 + }, + { + "epoch": 4.820215741110667, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12065 + }, + { + "epoch": 4.820615261685977, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12066 + }, + { + "epoch": 4.821014782261287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12067 + }, + { + "epoch": 4.821414302836596, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12068 + }, + { + "epoch": 4.8218138234119055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12069 + }, + { + "epoch": 4.822213343987215, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12070 + }, + { + "epoch": 4.822612864562525, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12071 + }, + { + "epoch": 4.823012385137835, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12072 + }, + { + "epoch": 4.823411905713145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12073 + }, + { + "epoch": 4.8238114262884535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12074 + }, + { + "epoch": 4.824210946863763, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12075 + }, + { + "epoch": 4.824610467439073, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12076 + }, + { + "epoch": 4.825009988014383, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12077 + }, + { + "epoch": 4.825409508589693, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 12078 + }, + { + "epoch": 4.825809029165002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12079 + }, + { + "epoch": 4.826208549740311, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12080 + }, + { + "epoch": 4.826608070315621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12081 + }, + { + "epoch": 4.827007590890931, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12082 + }, + { + "epoch": 4.827407111466241, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12083 + }, + { + "epoch": 4.8278066320415505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12084 + }, + { + "epoch": 4.828206152616859, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12085 + }, + { + "epoch": 4.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12086 + }, + { + "epoch": 4.829005193767479, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12087 + }, + { + "epoch": 4.829404714342789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12088 + }, + { + "epoch": 4.8298042349180985, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12089 + }, + { + "epoch": 4.830203755493407, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12090 + }, + { + "epoch": 4.830603276068717, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12091 + }, + { + "epoch": 4.831002796644027, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12092 + }, + { + "epoch": 4.831402317219337, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12093 + }, + { + "epoch": 4.831801837794647, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 12094 + }, + { + "epoch": 4.832201358369956, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12095 + }, + { + "epoch": 4.832600878945266, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12096 + }, + { + "epoch": 4.833000399520575, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12097 + }, + { + "epoch": 4.833399920095885, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12098 + }, + { + "epoch": 4.833799440671195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12099 + }, + { + "epoch": 4.834198961246504, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12100 + }, + { + "epoch": 4.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12101 + }, + { + "epoch": 4.834998002397123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12102 + }, + { + "epoch": 4.835397522972433, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12103 + }, + { + "epoch": 4.835797043547743, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12104 + }, + { + "epoch": 4.836196564123052, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12105 + }, + { + "epoch": 4.836596084698362, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12106 + }, + { + "epoch": 4.836995605273671, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12107 + }, + { + "epoch": 4.837395125848981, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12108 + }, + { + "epoch": 4.837794646424291, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12109 + }, + { + "epoch": 4.8381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12110 + }, + { + "epoch": 4.83859368757491, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12111 + }, + { + "epoch": 4.83899320815022, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12112 + }, + { + "epoch": 4.83939272872553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12113 + }, + { + "epoch": 4.839792249300839, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12114 + }, + { + "epoch": 4.8401917698761485, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12115 + }, + { + "epoch": 4.840591290451458, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12116 + }, + { + "epoch": 4.840990811026768, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12117 + }, + { + "epoch": 4.841390331602078, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12118 + }, + { + "epoch": 4.841789852177387, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12119 + }, + { + "epoch": 4.842189372752697, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12120 + }, + { + "epoch": 4.842588893328006, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12121 + }, + { + "epoch": 4.842988413903316, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12122 + }, + { + "epoch": 4.843387934478626, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 12123 + }, + { + "epoch": 4.843787455053935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12124 + }, + { + "epoch": 4.844186975629245, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12125 + }, + { + "epoch": 4.844586496204554, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12126 + }, + { + "epoch": 4.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12127 + }, + { + "epoch": 4.845385537355174, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12128 + }, + { + "epoch": 4.845785057930484, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12129 + }, + { + "epoch": 4.8461845785057935, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 12130 + }, + { + "epoch": 4.846584099081102, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12131 + }, + { + "epoch": 4.846983619656412, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12132 + }, + { + "epoch": 4.847383140231722, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12133 + }, + { + "epoch": 4.847782660807032, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12134 + }, + { + "epoch": 4.848182181382342, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12135 + }, + { + "epoch": 4.8485817019576505, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12136 + }, + { + "epoch": 4.84898122253296, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 12137 + }, + { + "epoch": 4.84938074310827, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12138 + }, + { + "epoch": 4.84978026368358, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12139 + }, + { + "epoch": 4.85017978425889, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12140 + }, + { + "epoch": 4.8505793048341985, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12141 + }, + { + "epoch": 4.850978825409508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12142 + }, + { + "epoch": 4.851378345984818, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12143 + }, + { + "epoch": 4.851777866560128, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12144 + }, + { + "epoch": 4.852177387135438, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12145 + }, + { + "epoch": 4.8525769077107475, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12146 + }, + { + "epoch": 4.852976428286056, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12147 + }, + { + "epoch": 4.853375948861366, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12148 + }, + { + "epoch": 4.853775469436676, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12149 + }, + { + "epoch": 4.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12150 + }, + { + "epoch": 4.8545745105872955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12151 + }, + { + "epoch": 4.854974031162605, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 12152 + }, + { + "epoch": 4.855373551737914, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12153 + }, + { + "epoch": 4.855773072313224, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12154 + }, + { + "epoch": 4.856172592888534, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12155 + }, + { + "epoch": 4.8565721134638435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12156 + }, + { + "epoch": 4.856971634039153, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12157 + }, + { + "epoch": 4.857371154614462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12158 + }, + { + "epoch": 4.857770675189772, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12159 + }, + { + "epoch": 4.858170195765082, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12160 + }, + { + "epoch": 4.858569716340392, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12161 + }, + { + "epoch": 4.858969236915701, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12162 + }, + { + "epoch": 4.859368757491011, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12163 + }, + { + "epoch": 4.85976827806632, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12164 + }, + { + "epoch": 4.86016779864163, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12165 + }, + { + "epoch": 4.86056731921694, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12166 + }, + { + "epoch": 4.860966839792249, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 12167 + }, + { + "epoch": 4.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12168 + }, + { + "epoch": 4.861765880942869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12169 + }, + { + "epoch": 4.862165401518178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12170 + }, + { + "epoch": 4.862564922093488, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12171 + }, + { + "epoch": 4.8629644426687975, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12172 + }, + { + "epoch": 4.863363963244107, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12173 + }, + { + "epoch": 4.863763483819417, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12174 + }, + { + "epoch": 4.864163004394726, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12175 + }, + { + "epoch": 4.864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12176 + }, + { + "epoch": 4.8649620455453455, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 12177 + }, + { + "epoch": 4.865361566120655, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12178 + }, + { + "epoch": 4.865761086695965, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12179 + }, + { + "epoch": 4.866160607271274, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12180 + }, + { + "epoch": 4.866560127846584, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12181 + }, + { + "epoch": 4.8669596484218935, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12182 + }, + { + "epoch": 4.867359168997203, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12183 + }, + { + "epoch": 4.867758689572513, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12184 + }, + { + "epoch": 4.868158210147823, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12185 + }, + { + "epoch": 4.868557730723133, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12186 + }, + { + "epoch": 4.868957251298442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12187 + }, + { + "epoch": 4.869356771873751, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12188 + }, + { + "epoch": 4.869756292449061, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12189 + }, + { + "epoch": 4.870155813024371, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12190 + }, + { + "epoch": 4.870555333599681, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12191 + }, + { + "epoch": 4.87095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12192 + }, + { + "epoch": 4.871354374750299, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12193 + }, + { + "epoch": 4.871753895325609, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12194 + }, + { + "epoch": 4.872153415900919, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12195 + }, + { + "epoch": 4.872552936476229, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 12196 + }, + { + "epoch": 4.872952457051538, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12197 + }, + { + "epoch": 4.8733519776268475, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12198 + }, + { + "epoch": 4.873751498202157, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 12199 + }, + { + "epoch": 4.874151018777467, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12200 + }, + { + "epoch": 4.874550539352777, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 12201 + }, + { + "epoch": 4.874950059928087, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12202 + }, + { + "epoch": 4.875349580503396, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12203 + }, + { + "epoch": 4.875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12204 + }, + { + "epoch": 4.876148621654015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12205 + }, + { + "epoch": 4.876548142229325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12206 + }, + { + "epoch": 4.876947662804635, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 12207 + }, + { + "epoch": 4.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12208 + }, + { + "epoch": 4.877746703955253, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12209 + }, + { + "epoch": 4.878146224530563, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12210 + }, + { + "epoch": 4.878545745105873, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12211 + }, + { + "epoch": 4.878945265681183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12212 + }, + { + "epoch": 4.8793447862564925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12213 + }, + { + "epoch": 4.879744306831801, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12214 + }, + { + "epoch": 4.880143827407111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12215 + }, + { + "epoch": 4.880543347982421, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12216 + }, + { + "epoch": 4.880942868557731, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12217 + }, + { + "epoch": 4.8813423891330405, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 12218 + }, + { + "epoch": 4.88174190970835, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12219 + }, + { + "epoch": 4.88214143028366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12220 + }, + { + "epoch": 4.882540950858969, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12221 + }, + { + "epoch": 4.882940471434279, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12222 + }, + { + "epoch": 4.883339992009589, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12223 + }, + { + "epoch": 4.883739512584898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12224 + }, + { + "epoch": 4.884139033160208, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12225 + }, + { + "epoch": 4.884538553735517, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12226 + }, + { + "epoch": 4.884938074310827, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12227 + }, + { + "epoch": 4.885337594886137, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12228 + }, + { + "epoch": 4.885737115461446, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12229 + }, + { + "epoch": 4.886136636036756, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12230 + }, + { + "epoch": 4.886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12231 + }, + { + "epoch": 4.886935677187375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12232 + }, + { + "epoch": 4.887335197762685, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12233 + }, + { + "epoch": 4.887734718337994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12234 + }, + { + "epoch": 4.888134238913304, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12235 + }, + { + "epoch": 4.888533759488614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12236 + }, + { + "epoch": 4.888933280063923, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12237 + }, + { + "epoch": 4.889332800639233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12238 + }, + { + "epoch": 4.8897323212145425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12239 + }, + { + "epoch": 4.890131841789852, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12240 + }, + { + "epoch": 4.890531362365162, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12241 + }, + { + "epoch": 4.890930882940472, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12242 + }, + { + "epoch": 4.891330403515781, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12243 + }, + { + "epoch": 4.8917299240910905, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12244 + }, + { + "epoch": 4.8921294446664, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 12245 + }, + { + "epoch": 4.89252896524171, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12246 + }, + { + "epoch": 4.89292848581702, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12247 + }, + { + "epoch": 4.893328006392329, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12248 + }, + { + "epoch": 4.893727526967639, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12249 + }, + { + "epoch": 4.894127047542948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12250 + }, + { + "epoch": 4.894526568118258, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12251 + }, + { + "epoch": 4.894926088693568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12252 + }, + { + "epoch": 4.895325609268878, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12253 + }, + { + "epoch": 4.895725129844187, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12254 + }, + { + "epoch": 4.896124650419496, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12255 + }, + { + "epoch": 4.896524170994806, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12256 + }, + { + "epoch": 4.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12257 + }, + { + "epoch": 4.897323212145426, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12258 + }, + { + "epoch": 4.8977227327207356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12259 + }, + { + "epoch": 4.8981222532960444, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12260 + }, + { + "epoch": 4.898521773871354, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12261 + }, + { + "epoch": 4.898921294446664, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12262 + }, + { + "epoch": 4.899320815021974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12263 + }, + { + "epoch": 4.899720335597284, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12264 + }, + { + "epoch": 4.9001198561725925, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12265 + }, + { + "epoch": 4.900519376747902, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12266 + }, + { + "epoch": 4.900918897323212, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12267 + }, + { + "epoch": 4.901318417898522, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12268 + }, + { + "epoch": 4.901717938473832, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12269 + }, + { + "epoch": 4.9021174590491405, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12270 + }, + { + "epoch": 4.90251697962445, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 12271 + }, + { + "epoch": 4.90291650019976, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12272 + }, + { + "epoch": 4.90331602077507, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12273 + }, + { + "epoch": 4.90371554135038, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12274 + }, + { + "epoch": 4.9041150619256895, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 12275 + }, + { + "epoch": 4.904514582500999, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12276 + }, + { + "epoch": 4.904914103076308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 12277 + }, + { + "epoch": 4.905313623651618, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12278 + }, + { + "epoch": 4.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12279 + }, + { + "epoch": 4.9061126648022375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12280 + }, + { + "epoch": 4.906512185377547, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12281 + }, + { + "epoch": 4.906911705952856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12282 + }, + { + "epoch": 4.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 12283 + }, + { + "epoch": 4.907710747103476, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12284 + }, + { + "epoch": 4.9081102676787856, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12285 + }, + { + "epoch": 4.908509788254095, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12286 + }, + { + "epoch": 4.908909308829404, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12287 + }, + { + "epoch": 4.909308829404714, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12288 + }, + { + "epoch": 4.909708349980024, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12289 + }, + { + "epoch": 4.910107870555334, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12290 + }, + { + "epoch": 4.910507391130643, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12291 + }, + { + "epoch": 4.910906911705953, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12292 + }, + { + "epoch": 4.911306432281263, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12293 + }, + { + "epoch": 4.911705952856572, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 12294 + }, + { + "epoch": 4.912105473431882, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12295 + }, + { + "epoch": 4.912504994007191, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12296 + }, + { + "epoch": 4.912904514582501, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12297 + }, + { + "epoch": 4.913304035157811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12298 + }, + { + "epoch": 4.91370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12299 + }, + { + "epoch": 4.91410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12300 + }, + { + "epoch": 4.9145025968837395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12301 + }, + { + "epoch": 4.914902117459049, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12302 + }, + { + "epoch": 4.915301638034359, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12303 + }, + { + "epoch": 4.915701158609668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12304 + }, + { + "epoch": 4.916100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12305 + }, + { + "epoch": 4.9165001997602875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12306 + }, + { + "epoch": 4.916899720335597, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12307 + }, + { + "epoch": 4.917299240910907, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12308 + }, + { + "epoch": 4.917698761486217, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12309 + }, + { + "epoch": 4.918098282061527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12310 + }, + { + "epoch": 4.918497802636836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12311 + }, + { + "epoch": 4.918897323212145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12312 + }, + { + "epoch": 4.919296843787455, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12313 + }, + { + "epoch": 4.919696364362765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12314 + }, + { + "epoch": 4.920095884938075, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12315 + }, + { + "epoch": 4.920495405513384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12316 + }, + { + "epoch": 4.920894926088693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12317 + }, + { + "epoch": 4.921294446664003, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12318 + }, + { + "epoch": 4.921693967239313, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12319 + }, + { + "epoch": 4.922093487814623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12320 + }, + { + "epoch": 4.922493008389932, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12321 + }, + { + "epoch": 4.922892528965241, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12322 + }, + { + "epoch": 4.923292049540551, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12323 + }, + { + "epoch": 4.923691570115861, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12324 + }, + { + "epoch": 4.924091090691171, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12325 + }, + { + "epoch": 4.924490611266481, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12326 + }, + { + "epoch": 4.9248901318417895, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12327 + }, + { + "epoch": 4.925289652417099, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12328 + }, + { + "epoch": 4.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12329 + }, + { + "epoch": 4.926088693567719, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12330 + }, + { + "epoch": 4.926488214143029, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2973, + "step": 12331 + }, + { + "epoch": 4.926887734718338, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12332 + }, + { + "epoch": 4.927287255293647, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 12333 + }, + { + "epoch": 4.927686775868957, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12334 + }, + { + "epoch": 4.928086296444267, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12335 + }, + { + "epoch": 4.928485817019577, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12336 + }, + { + "epoch": 4.9288853375948865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12337 + }, + { + "epoch": 4.929284858170195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 12338 + }, + { + "epoch": 4.929684378745505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12339 + }, + { + "epoch": 4.930083899320815, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12340 + }, + { + "epoch": 4.930483419896125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12341 + }, + { + "epoch": 4.9308829404714345, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12342 + }, + { + "epoch": 4.931282461046744, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12343 + }, + { + "epoch": 4.931681981622053, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12344 + }, + { + "epoch": 4.932081502197363, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12345 + }, + { + "epoch": 4.932481022772673, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12346 + }, + { + "epoch": 4.9328805433479825, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12347 + }, + { + "epoch": 4.933280063923292, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12348 + }, + { + "epoch": 4.933679584498602, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12349 + }, + { + "epoch": 4.934079105073911, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12350 + }, + { + "epoch": 4.934478625649221, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12351 + }, + { + "epoch": 4.934878146224531, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12352 + }, + { + "epoch": 4.93527766679984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12353 + }, + { + "epoch": 4.93567718737515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12354 + }, + { + "epoch": 4.936076707950459, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12355 + }, + { + "epoch": 4.936476228525769, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12356 + }, + { + "epoch": 4.936875749101079, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12357 + }, + { + "epoch": 4.937275269676388, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12358 + }, + { + "epoch": 4.937674790251698, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12359 + }, + { + "epoch": 4.938074310827007, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12360 + }, + { + "epoch": 4.938473831402317, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12361 + }, + { + "epoch": 4.938873351977627, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12362 + }, + { + "epoch": 4.9392728725529365, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12363 + }, + { + "epoch": 4.939672393128246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12364 + }, + { + "epoch": 4.940071913703556, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12365 + }, + { + "epoch": 4.940471434278866, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12366 + }, + { + "epoch": 4.940870954854175, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12367 + }, + { + "epoch": 4.9412704754294845, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12368 + }, + { + "epoch": 4.941669996004794, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12369 + }, + { + "epoch": 4.942069516580104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12370 + }, + { + "epoch": 4.942469037155414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12371 + }, + { + "epoch": 4.942868557730723, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12372 + }, + { + "epoch": 4.9432680783060325, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12373 + }, + { + "epoch": 4.943667598881342, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12374 + }, + { + "epoch": 4.944067119456652, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 12375 + }, + { + "epoch": 4.944466640031962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12376 + }, + { + "epoch": 4.944866160607271, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12377 + }, + { + "epoch": 4.945265681182581, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12378 + }, + { + "epoch": 4.94566520175789, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12379 + }, + { + "epoch": 4.9460647223332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12380 + }, + { + "epoch": 4.94646424290851, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12381 + }, + { + "epoch": 4.94686376348382, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12382 + }, + { + "epoch": 4.9472632840591295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12383 + }, + { + "epoch": 4.947662804634438, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12384 + }, + { + "epoch": 4.948062325209748, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12385 + }, + { + "epoch": 4.948461845785058, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12386 + }, + { + "epoch": 4.948861366360368, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12387 + }, + { + "epoch": 4.949260886935678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12388 + }, + { + "epoch": 4.9496604075109865, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12389 + }, + { + "epoch": 4.950059928086296, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12390 + }, + { + "epoch": 4.950459448661606, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12391 + }, + { + "epoch": 4.950858969236916, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12392 + }, + { + "epoch": 4.951258489812226, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 12393 + }, + { + "epoch": 4.9516580103875345, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12394 + }, + { + "epoch": 4.952057530962844, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 12395 + }, + { + "epoch": 4.952457051538154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12396 + }, + { + "epoch": 4.952856572113464, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12397 + }, + { + "epoch": 4.953256092688774, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12398 + }, + { + "epoch": 4.953655613264083, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12399 + }, + { + "epoch": 4.954055133839393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12400 + }, + { + "epoch": 4.954454654414702, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12401 + }, + { + "epoch": 4.954854174990012, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12402 + }, + { + "epoch": 4.955253695565322, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12403 + }, + { + "epoch": 4.9556532161406315, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12404 + }, + { + "epoch": 4.956052736715941, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12405 + }, + { + "epoch": 4.95645225729125, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12406 + }, + { + "epoch": 4.95685177786656, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12407 + }, + { + "epoch": 4.95725129844187, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12408 + }, + { + "epoch": 4.9576508190171795, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 12409 + }, + { + "epoch": 4.958050339592489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12410 + }, + { + "epoch": 4.958449860167798, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12411 + }, + { + "epoch": 4.958849380743108, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 12412 + }, + { + "epoch": 4.959248901318418, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12413 + }, + { + "epoch": 4.959648421893728, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12414 + }, + { + "epoch": 4.960047942469037, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12415 + }, + { + "epoch": 4.960447463044347, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12416 + }, + { + "epoch": 4.960846983619656, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12417 + }, + { + "epoch": 4.961246504194966, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12418 + }, + { + "epoch": 4.961646024770276, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12419 + }, + { + "epoch": 4.962045545345585, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12420 + }, + { + "epoch": 4.962445065920895, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12421 + }, + { + "epoch": 4.962844586496205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12422 + }, + { + "epoch": 4.963244107071514, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12423 + }, + { + "epoch": 4.963643627646824, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12424 + }, + { + "epoch": 4.964043148222133, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12425 + }, + { + "epoch": 4.964442668797443, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12426 + }, + { + "epoch": 4.964842189372753, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12427 + }, + { + "epoch": 4.965241709948062, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12428 + }, + { + "epoch": 4.965641230523372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12429 + }, + { + "epoch": 4.9660407510986815, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 12430 + }, + { + "epoch": 4.966440271673991, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12431 + }, + { + "epoch": 4.966839792249301, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12432 + }, + { + "epoch": 4.967239312824611, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12433 + }, + { + "epoch": 4.96763883339992, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12434 + }, + { + "epoch": 4.9680383539752295, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12435 + }, + { + "epoch": 4.968437874550539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12436 + }, + { + "epoch": 4.968837395125849, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12437 + }, + { + "epoch": 4.969236915701159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12438 + }, + { + "epoch": 4.969636436276469, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12439 + }, + { + "epoch": 4.970035956851778, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12440 + }, + { + "epoch": 4.970435477427087, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12441 + }, + { + "epoch": 4.970834998002397, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12442 + }, + { + "epoch": 4.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12443 + }, + { + "epoch": 4.971634039153017, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12444 + }, + { + "epoch": 4.972033559728326, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12445 + }, + { + "epoch": 4.972433080303635, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12446 + }, + { + "epoch": 4.972832600878945, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12447 + }, + { + "epoch": 4.973232121454255, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12448 + }, + { + "epoch": 4.973631642029565, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12449 + }, + { + "epoch": 4.974031162604874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12450 + }, + { + "epoch": 4.974430683180183, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12451 + }, + { + "epoch": 4.974830203755493, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12452 + }, + { + "epoch": 4.975229724330803, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12453 + }, + { + "epoch": 4.975629244906113, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12454 + }, + { + "epoch": 4.976028765481423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12455 + }, + { + "epoch": 4.976428286056732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12456 + }, + { + "epoch": 4.976827806632041, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12457 + }, + { + "epoch": 4.977227327207351, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12458 + }, + { + "epoch": 4.977626847782661, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12459 + }, + { + "epoch": 4.978026368357971, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 12460 + }, + { + "epoch": 4.97842588893328, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12461 + }, + { + "epoch": 4.978825409508589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 12462 + }, + { + "epoch": 4.979224930083899, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12463 + }, + { + "epoch": 4.979624450659209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 12464 + }, + { + "epoch": 4.980023971234519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 12465 + }, + { + "epoch": 4.9804234918098285, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12466 + }, + { + "epoch": 4.980823012385137, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12467 + }, + { + "epoch": 4.981222532960447, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12468 + }, + { + "epoch": 4.981622053535757, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12469 + }, + { + "epoch": 4.982021574111067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12470 + }, + { + "epoch": 4.9824210946863765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12471 + }, + { + "epoch": 4.982820615261686, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12472 + }, + { + "epoch": 4.983220135836996, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12473 + }, + { + "epoch": 4.983619656412305, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12474 + }, + { + "epoch": 4.984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12475 + }, + { + "epoch": 4.9844186975629245, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12476 + }, + { + "epoch": 4.984818218138234, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12477 + }, + { + "epoch": 4.985217738713544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12478 + }, + { + "epoch": 4.985617259288853, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12479 + }, + { + "epoch": 4.986016779864163, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12480 + }, + { + "epoch": 4.986416300439473, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12481 + }, + { + "epoch": 4.986815821014782, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12482 + }, + { + "epoch": 4.987215341590092, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12483 + }, + { + "epoch": 4.987614862165401, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12484 + }, + { + "epoch": 4.988014382740711, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12485 + }, + { + "epoch": 4.988413903316021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12486 + }, + { + "epoch": 4.98881342389133, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12487 + }, + { + "epoch": 4.98921294446664, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12488 + }, + { + "epoch": 4.98961246504195, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12489 + }, + { + "epoch": 4.99001198561726, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12490 + }, + { + "epoch": 4.990411506192569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12491 + }, + { + "epoch": 4.9908110267678785, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12492 + }, + { + "epoch": 4.991210547343188, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12493 + }, + { + "epoch": 4.991610067918498, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12494 + }, + { + "epoch": 4.992009588493808, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12495 + }, + { + "epoch": 4.992409109069117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12496 + }, + { + "epoch": 4.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12497 + }, + { + "epoch": 4.993208150219736, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12498 + }, + { + "epoch": 4.993607670795046, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12499 + }, + { + "epoch": 4.994007191370356, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 12500 + }, + { + "epoch": 4.994406711945665, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12501 + }, + { + "epoch": 4.9948062325209746, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12502 + }, + { + "epoch": 4.995205753096284, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12503 + }, + { + "epoch": 4.995605273671594, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12504 + }, + { + "epoch": 4.996004794246904, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12505 + }, + { + "epoch": 4.996404314822214, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12506 + }, + { + "epoch": 4.996803835397523, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12507 + }, + { + "epoch": 4.997203355972832, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12508 + }, + { + "epoch": 4.997602876548142, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12509 + }, + { + "epoch": 4.998002397123452, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12510 + }, + { + "epoch": 4.998401917698762, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 12511 + }, + { + "epoch": 4.9988014382740715, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12512 + }, + { + "epoch": 4.99920095884938, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12513 + }, + { + "epoch": 4.99960047942469, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12514 + }, + { + "epoch": 5.0, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12515 + }, + { + "epoch": 5.00039952057531, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12516 + }, + { + "epoch": 5.00079904115062, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12517 + }, + { + "epoch": 5.0011985617259285, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12518 + }, + { + "epoch": 5.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12519 + }, + { + "epoch": 5.001997602876548, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12520 + }, + { + "epoch": 5.002397123451858, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12521 + }, + { + "epoch": 5.002796644027168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12522 + }, + { + "epoch": 5.003196164602477, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12523 + }, + { + "epoch": 5.003595685177786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12524 + }, + { + "epoch": 5.003995205753096, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12525 + }, + { + "epoch": 5.004394726328406, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12526 + }, + { + "epoch": 5.004794246903716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12527 + }, + { + "epoch": 5.0051937674790254, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12528 + }, + { + "epoch": 5.005593288054335, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12529 + }, + { + "epoch": 5.005992808629644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12530 + }, + { + "epoch": 5.006392329204954, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12531 + }, + { + "epoch": 5.006791849780264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12532 + }, + { + "epoch": 5.0071913703555735, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12533 + }, + { + "epoch": 5.007590890930883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12534 + }, + { + "epoch": 5.007990411506192, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12535 + }, + { + "epoch": 5.008389932081502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12536 + }, + { + "epoch": 5.008789452656812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12537 + }, + { + "epoch": 5.0091889732321215, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12538 + }, + { + "epoch": 5.009588493807431, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12539 + }, + { + "epoch": 5.009988014382741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12540 + }, + { + "epoch": 5.01038753495805, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12541 + }, + { + "epoch": 5.01078705553336, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12542 + }, + { + "epoch": 5.01118657610867, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12543 + }, + { + "epoch": 5.011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12544 + }, + { + "epoch": 5.011985617259289, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12545 + }, + { + "epoch": 5.012385137834598, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12546 + }, + { + "epoch": 5.012784658409908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12547 + }, + { + "epoch": 5.013184178985218, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12548 + }, + { + "epoch": 5.013583699560527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12549 + }, + { + "epoch": 5.013983220135837, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 12550 + }, + { + "epoch": 5.014382740711147, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12551 + }, + { + "epoch": 5.014782261286456, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12552 + }, + { + "epoch": 5.015181781861766, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 12553 + }, + { + "epoch": 5.0155813024370755, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12554 + }, + { + "epoch": 5.015980823012385, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12555 + }, + { + "epoch": 5.016380343587695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12556 + }, + { + "epoch": 5.016779864163005, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12557 + }, + { + "epoch": 5.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12558 + }, + { + "epoch": 5.0175789053136235, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12559 + }, + { + "epoch": 5.017978425888933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12560 + }, + { + "epoch": 5.018377946464243, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12561 + }, + { + "epoch": 5.018777467039553, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12562 + }, + { + "epoch": 5.019176987614862, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12563 + }, + { + "epoch": 5.0195765081901715, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12564 + }, + { + "epoch": 5.019976028765481, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12565 + }, + { + "epoch": 5.020375549340791, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12566 + }, + { + "epoch": 5.020775069916101, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12567 + }, + { + "epoch": 5.021174590491411, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12568 + }, + { + "epoch": 5.02157411106672, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12569 + }, + { + "epoch": 5.021973631642029, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 12570 + }, + { + "epoch": 5.022373152217339, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12571 + }, + { + "epoch": 5.022772672792649, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12572 + }, + { + "epoch": 5.023172193367959, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12573 + }, + { + "epoch": 5.0235717139432685, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12574 + }, + { + "epoch": 5.023971234518577, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12575 + }, + { + "epoch": 5.024370755093887, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12576 + }, + { + "epoch": 5.024770275669197, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12577 + }, + { + "epoch": 5.025169796244507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12578 + }, + { + "epoch": 5.025569316819817, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12579 + }, + { + "epoch": 5.0259688373951255, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12580 + }, + { + "epoch": 5.026368357970435, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12581 + }, + { + "epoch": 5.026767878545745, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12582 + }, + { + "epoch": 5.027167399121055, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12583 + }, + { + "epoch": 5.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12584 + }, + { + "epoch": 5.027966440271674, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12585 + }, + { + "epoch": 5.028365960846983, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12586 + }, + { + "epoch": 5.028765481422293, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12587 + }, + { + "epoch": 5.029165001997603, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12588 + }, + { + "epoch": 5.029564522572913, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12589 + }, + { + "epoch": 5.029964043148222, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12590 + }, + { + "epoch": 5.030363563723531, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12591 + }, + { + "epoch": 5.030763084298841, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12592 + }, + { + "epoch": 5.031162604874151, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12593 + }, + { + "epoch": 5.031562125449461, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12594 + }, + { + "epoch": 5.0319616460247705, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12595 + }, + { + "epoch": 5.03236116660008, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12596 + }, + { + "epoch": 5.032760687175389, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12597 + }, + { + "epoch": 5.033160207750699, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 12598 + }, + { + "epoch": 5.033559728326009, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12599 + }, + { + "epoch": 5.0339592489013185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12600 + }, + { + "epoch": 5.034358769476628, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12601 + }, + { + "epoch": 5.034758290051938, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12602 + }, + { + "epoch": 5.035157810627247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12603 + }, + { + "epoch": 5.035557331202557, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 12604 + }, + { + "epoch": 5.035956851777867, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12605 + }, + { + "epoch": 5.036356372353176, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12606 + }, + { + "epoch": 5.036755892928486, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12607 + }, + { + "epoch": 5.037155413503795, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12608 + }, + { + "epoch": 5.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12609 + }, + { + "epoch": 5.037954454654415, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12610 + }, + { + "epoch": 5.038353975229724, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12611 + }, + { + "epoch": 5.038753495805034, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12612 + }, + { + "epoch": 5.039153016380344, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12613 + }, + { + "epoch": 5.039552536955653, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12614 + }, + { + "epoch": 5.039952057530963, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12615 + }, + { + "epoch": 5.040351578106272, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 12616 + }, + { + "epoch": 5.040751098681582, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12617 + }, + { + "epoch": 5.041150619256892, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 12618 + }, + { + "epoch": 5.041550139832202, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12619 + }, + { + "epoch": 5.041949660407511, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12620 + }, + { + "epoch": 5.0423491809828205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12621 + }, + { + "epoch": 5.04274870155813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12622 + }, + { + "epoch": 5.04314822213344, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 12623 + }, + { + "epoch": 5.04354774270875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12624 + }, + { + "epoch": 5.043947263284059, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12625 + }, + { + "epoch": 5.0443467838593685, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12626 + }, + { + "epoch": 5.044746304434678, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12627 + }, + { + "epoch": 5.045145825009988, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12628 + }, + { + "epoch": 5.045545345585298, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12629 + }, + { + "epoch": 5.045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12630 + }, + { + "epoch": 5.046344386735917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12631 + }, + { + "epoch": 5.046743907311226, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12632 + }, + { + "epoch": 5.047143427886536, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12633 + }, + { + "epoch": 5.047542948461846, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12634 + }, + { + "epoch": 5.047942469037156, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12635 + }, + { + "epoch": 5.048341989612465, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12636 + }, + { + "epoch": 5.048741510187774, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12637 + }, + { + "epoch": 5.049141030763084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12638 + }, + { + "epoch": 5.049540551338394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12639 + }, + { + "epoch": 5.049940071913704, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12640 + }, + { + "epoch": 5.0503395924890135, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12641 + }, + { + "epoch": 5.050739113064322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12642 + }, + { + "epoch": 5.051138633639632, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12643 + }, + { + "epoch": 5.051538154214942, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12644 + }, + { + "epoch": 5.051937674790252, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 12645 + }, + { + "epoch": 5.052337195365562, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12646 + }, + { + "epoch": 5.052736715940871, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12647 + }, + { + "epoch": 5.05313623651618, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12648 + }, + { + "epoch": 5.05353575709149, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12649 + }, + { + "epoch": 5.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12650 + }, + { + "epoch": 5.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12651 + }, + { + "epoch": 5.054734318817419, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12652 + }, + { + "epoch": 5.055133839392728, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12653 + }, + { + "epoch": 5.055533359968038, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12654 + }, + { + "epoch": 5.055932880543348, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12655 + }, + { + "epoch": 5.056332401118658, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12656 + }, + { + "epoch": 5.0567319216939675, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12657 + }, + { + "epoch": 5.057131442269277, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12658 + }, + { + "epoch": 5.057530962844586, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12659 + }, + { + "epoch": 5.057930483419896, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12660 + }, + { + "epoch": 5.058330003995206, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12661 + }, + { + "epoch": 5.0587295245705155, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12662 + }, + { + "epoch": 5.059129045145825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12663 + }, + { + "epoch": 5.059528565721135, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12664 + }, + { + "epoch": 5.059928086296444, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12665 + }, + { + "epoch": 5.060327606871754, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12666 + }, + { + "epoch": 5.0607271274470635, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12667 + }, + { + "epoch": 5.061126648022373, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12668 + }, + { + "epoch": 5.061526168597683, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12669 + }, + { + "epoch": 5.061925689172992, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 12670 + }, + { + "epoch": 5.062325209748302, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12671 + }, + { + "epoch": 5.062724730323612, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12672 + }, + { + "epoch": 5.063124250898921, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12673 + }, + { + "epoch": 5.063523771474231, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12674 + }, + { + "epoch": 5.063923292049541, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12675 + }, + { + "epoch": 5.06432281262485, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12676 + }, + { + "epoch": 5.06472233320016, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12677 + }, + { + "epoch": 5.065121853775469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12678 + }, + { + "epoch": 5.065521374350779, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12679 + }, + { + "epoch": 5.065920894926089, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12680 + }, + { + "epoch": 5.066320415501398, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12681 + }, + { + "epoch": 5.066719936076708, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12682 + }, + { + "epoch": 5.0671194566520175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12683 + }, + { + "epoch": 5.067518977227327, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12684 + }, + { + "epoch": 5.067918497802637, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12685 + }, + { + "epoch": 5.068318018377947, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12686 + }, + { + "epoch": 5.068717538953256, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12687 + }, + { + "epoch": 5.0691170595285655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12688 + }, + { + "epoch": 5.069516580103875, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12689 + }, + { + "epoch": 5.069916100679185, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12690 + }, + { + "epoch": 5.070315621254495, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12691 + }, + { + "epoch": 5.070715141829805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12692 + }, + { + "epoch": 5.0711146624051135, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12693 + }, + { + "epoch": 5.071514182980423, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12694 + }, + { + "epoch": 5.071913703555733, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12695 + }, + { + "epoch": 5.072313224131043, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 12696 + }, + { + "epoch": 5.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12697 + }, + { + "epoch": 5.073112265281662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12698 + }, + { + "epoch": 5.073511785856971, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12699 + }, + { + "epoch": 5.073911306432281, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12700 + }, + { + "epoch": 5.074310827007591, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12701 + }, + { + "epoch": 5.074710347582901, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 12702 + }, + { + "epoch": 5.0751098681582105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12703 + }, + { + "epoch": 5.075509388733519, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12704 + }, + { + "epoch": 5.075908909308829, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12705 + }, + { + "epoch": 5.076308429884139, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12706 + }, + { + "epoch": 5.076707950459449, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12707 + }, + { + "epoch": 5.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12708 + }, + { + "epoch": 5.077506991610068, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12709 + }, + { + "epoch": 5.077906512185377, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 12710 + }, + { + "epoch": 5.078306032760687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12711 + }, + { + "epoch": 5.078705553335997, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12712 + }, + { + "epoch": 5.079105073911307, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12713 + }, + { + "epoch": 5.079504594486616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12714 + }, + { + "epoch": 5.079904115061925, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12715 + }, + { + "epoch": 5.080303635637235, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12716 + }, + { + "epoch": 5.080703156212545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12717 + }, + { + "epoch": 5.081102676787855, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12718 + }, + { + "epoch": 5.081502197363164, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12719 + }, + { + "epoch": 5.081901717938474, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12720 + }, + { + "epoch": 5.082301238513783, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12721 + }, + { + "epoch": 5.082700759089093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12722 + }, + { + "epoch": 5.083100279664403, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12723 + }, + { + "epoch": 5.0834998002397125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12724 + }, + { + "epoch": 5.083899320815022, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12725 + }, + { + "epoch": 5.084298841390331, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 12726 + }, + { + "epoch": 5.084698361965641, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12727 + }, + { + "epoch": 5.085097882540951, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12728 + }, + { + "epoch": 5.0854974031162605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12729 + }, + { + "epoch": 5.08589692369157, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12730 + }, + { + "epoch": 5.08629644426688, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12731 + }, + { + "epoch": 5.086695964842189, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12732 + }, + { + "epoch": 5.087095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12733 + }, + { + "epoch": 5.087495005992809, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 12734 + }, + { + "epoch": 5.087894526568118, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12735 + }, + { + "epoch": 5.088294047143428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12736 + }, + { + "epoch": 5.088693567718738, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12737 + }, + { + "epoch": 5.089093088294047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 12738 + }, + { + "epoch": 5.089492608869357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12739 + }, + { + "epoch": 5.089892129444666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12740 + }, + { + "epoch": 5.090291650019976, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12741 + }, + { + "epoch": 5.090691170595286, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12742 + }, + { + "epoch": 5.091090691170595, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12743 + }, + { + "epoch": 5.091490211745905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12744 + }, + { + "epoch": 5.0918897323212144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12745 + }, + { + "epoch": 5.092289252896524, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12746 + }, + { + "epoch": 5.092688773471834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12747 + }, + { + "epoch": 5.093088294047144, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 12748 + }, + { + "epoch": 5.093487814622453, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 12749 + }, + { + "epoch": 5.0938873351977625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12750 + }, + { + "epoch": 5.094286855773072, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12751 + }, + { + "epoch": 5.094686376348382, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12752 + }, + { + "epoch": 5.095085896923692, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12753 + }, + { + "epoch": 5.095485417499002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12754 + }, + { + "epoch": 5.0958849380743105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12755 + }, + { + "epoch": 5.09628445864962, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12756 + }, + { + "epoch": 5.09668397922493, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12757 + }, + { + "epoch": 5.09708349980024, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12758 + }, + { + "epoch": 5.09748302037555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12759 + }, + { + "epoch": 5.097882540950859, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12760 + }, + { + "epoch": 5.098282061526168, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 12761 + }, + { + "epoch": 5.098681582101478, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12762 + }, + { + "epoch": 5.099081102676788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12763 + }, + { + "epoch": 5.099480623252098, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12764 + }, + { + "epoch": 5.0998801438274075, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12765 + }, + { + "epoch": 5.100279664402716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 12766 + }, + { + "epoch": 5.100679184978026, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12767 + }, + { + "epoch": 5.101078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 12768 + }, + { + "epoch": 5.101478226128646, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12769 + }, + { + "epoch": 5.1018777467039556, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12770 + }, + { + "epoch": 5.1022772672792644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12771 + }, + { + "epoch": 5.102676787854574, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12772 + }, + { + "epoch": 5.103076308429884, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12773 + }, + { + "epoch": 5.103475829005194, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 12774 + }, + { + "epoch": 5.103875349580504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12775 + }, + { + "epoch": 5.104274870155813, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12776 + }, + { + "epoch": 5.104674390731122, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12777 + }, + { + "epoch": 5.105073911306432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12778 + }, + { + "epoch": 5.105473431881742, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12779 + }, + { + "epoch": 5.105872952457052, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12780 + }, + { + "epoch": 5.106272473032361, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12781 + }, + { + "epoch": 5.106671993607671, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12782 + }, + { + "epoch": 5.10707151418298, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12783 + }, + { + "epoch": 5.10747103475829, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12784 + }, + { + "epoch": 5.1078705553336, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12785 + }, + { + "epoch": 5.1082700759089095, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 12786 + }, + { + "epoch": 5.108669596484219, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12787 + }, + { + "epoch": 5.109069117059528, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12788 + }, + { + "epoch": 5.109468637634838, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12789 + }, + { + "epoch": 5.109868158210148, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12790 + }, + { + "epoch": 5.1102676787854575, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12791 + }, + { + "epoch": 5.110667199360767, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12792 + }, + { + "epoch": 5.111066719936077, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12793 + }, + { + "epoch": 5.111466240511386, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12794 + }, + { + "epoch": 5.111865761086696, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12795 + }, + { + "epoch": 5.112265281662006, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12796 + }, + { + "epoch": 5.112664802237315, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12797 + }, + { + "epoch": 5.113064322812625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12798 + }, + { + "epoch": 5.113463843387935, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12799 + }, + { + "epoch": 5.113863363963244, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12800 + }, + { + "epoch": 5.114262884538554, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12801 + }, + { + "epoch": 5.114662405113863, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12802 + }, + { + "epoch": 5.115061925689173, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12803 + }, + { + "epoch": 5.115461446264483, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12804 + }, + { + "epoch": 5.115860966839792, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12805 + }, + { + "epoch": 5.116260487415102, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12806 + }, + { + "epoch": 5.116660007990411, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12807 + }, + { + "epoch": 5.117059528565721, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12808 + }, + { + "epoch": 5.117459049141031, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12809 + }, + { + "epoch": 5.117858569716341, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12810 + }, + { + "epoch": 5.11825809029165, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12811 + }, + { + "epoch": 5.1186576108669595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12812 + }, + { + "epoch": 5.119057131442269, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12813 + }, + { + "epoch": 5.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12814 + }, + { + "epoch": 5.119856172592889, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12815 + }, + { + "epoch": 5.120255693168198, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12816 + }, + { + "epoch": 5.1206552137435075, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12817 + }, + { + "epoch": 5.121054734318817, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12818 + }, + { + "epoch": 5.121454254894127, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12819 + }, + { + "epoch": 5.121853775469437, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12820 + }, + { + "epoch": 5.122253296044747, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3011, + "step": 12821 + }, + { + "epoch": 5.122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12822 + }, + { + "epoch": 5.123052337195365, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12823 + }, + { + "epoch": 5.123451857770675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12824 + }, + { + "epoch": 5.123851378345985, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12825 + }, + { + "epoch": 5.124250898921295, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 12826 + }, + { + "epoch": 5.1246504194966045, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12827 + }, + { + "epoch": 5.125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12828 + }, + { + "epoch": 5.125449460647223, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 12829 + }, + { + "epoch": 5.125848981222533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12830 + }, + { + "epoch": 5.126248501797843, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12831 + }, + { + "epoch": 5.1266480223731525, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12832 + }, + { + "epoch": 5.127047542948461, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12833 + }, + { + "epoch": 5.127447063523771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12834 + }, + { + "epoch": 5.127846584099081, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12835 + }, + { + "epoch": 5.128246104674391, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12836 + }, + { + "epoch": 5.128645625249701, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12837 + }, + { + "epoch": 5.12904514582501, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12838 + }, + { + "epoch": 5.129444666400319, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12839 + }, + { + "epoch": 5.129844186975629, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12840 + }, + { + "epoch": 5.130243707550939, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12841 + }, + { + "epoch": 5.130643228126249, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12842 + }, + { + "epoch": 5.131042748701558, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 12843 + }, + { + "epoch": 5.131442269276867, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12844 + }, + { + "epoch": 5.131841789852177, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12845 + }, + { + "epoch": 5.132241310427487, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12846 + }, + { + "epoch": 5.132640831002797, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12847 + }, + { + "epoch": 5.1330403515781065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12848 + }, + { + "epoch": 5.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12849 + }, + { + "epoch": 5.133839392728725, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12850 + }, + { + "epoch": 5.134238913304035, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 12851 + }, + { + "epoch": 5.134638433879345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12852 + }, + { + "epoch": 5.1350379544546545, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12853 + }, + { + "epoch": 5.135437475029964, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12854 + }, + { + "epoch": 5.135836995605274, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12855 + }, + { + "epoch": 5.136236516180583, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12856 + }, + { + "epoch": 5.136636036755893, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12857 + }, + { + "epoch": 5.1370355573312025, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12858 + }, + { + "epoch": 5.137435077906512, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12859 + }, + { + "epoch": 5.137834598481822, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12860 + }, + { + "epoch": 5.138234119057131, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12861 + }, + { + "epoch": 5.138633639632441, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12862 + }, + { + "epoch": 5.139033160207751, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12863 + }, + { + "epoch": 5.13943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12864 + }, + { + "epoch": 5.13983220135837, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12865 + }, + { + "epoch": 5.14023172193368, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12866 + }, + { + "epoch": 5.140631242508989, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12867 + }, + { + "epoch": 5.141030763084299, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12868 + }, + { + "epoch": 5.141430283659608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12869 + }, + { + "epoch": 5.141829804234918, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12870 + }, + { + "epoch": 5.142229324810228, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12871 + }, + { + "epoch": 5.142628845385538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12872 + }, + { + "epoch": 5.143028365960847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12873 + }, + { + "epoch": 5.1434278865361565, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12874 + }, + { + "epoch": 5.143827407111466, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12875 + }, + { + "epoch": 5.144226927686776, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12876 + }, + { + "epoch": 5.144626448262086, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12877 + }, + { + "epoch": 5.145025968837395, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12878 + }, + { + "epoch": 5.1454254894127045, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12879 + }, + { + "epoch": 5.145825009988014, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12880 + }, + { + "epoch": 5.146224530563324, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12881 + }, + { + "epoch": 5.146624051138634, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12882 + }, + { + "epoch": 5.147023571713944, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12883 + }, + { + "epoch": 5.1474230922892525, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12884 + }, + { + "epoch": 5.147822612864562, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12885 + }, + { + "epoch": 5.148222133439872, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12886 + }, + { + "epoch": 5.148621654015182, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12887 + }, + { + "epoch": 5.149021174590492, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12888 + }, + { + "epoch": 5.1494206951658015, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12889 + }, + { + "epoch": 5.14982021574111, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 12890 + }, + { + "epoch": 5.15021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12891 + }, + { + "epoch": 5.15061925689173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12892 + }, + { + "epoch": 5.15101877746704, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12893 + }, + { + "epoch": 5.1514182980423495, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12894 + }, + { + "epoch": 5.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12895 + }, + { + "epoch": 5.152217339192968, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12896 + }, + { + "epoch": 5.152616859768278, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12897 + }, + { + "epoch": 5.153016380343588, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12898 + }, + { + "epoch": 5.153415900918898, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12899 + }, + { + "epoch": 5.153815421494207, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12900 + }, + { + "epoch": 5.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12901 + }, + { + "epoch": 5.154614462644826, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12902 + }, + { + "epoch": 5.155013983220136, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12903 + }, + { + "epoch": 5.155413503795446, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12904 + }, + { + "epoch": 5.155813024370755, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12905 + }, + { + "epoch": 5.156212544946065, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12906 + }, + { + "epoch": 5.156612065521374, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12907 + }, + { + "epoch": 5.157011586096684, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 12908 + }, + { + "epoch": 5.157411106671994, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12909 + }, + { + "epoch": 5.157810627247303, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12910 + }, + { + "epoch": 5.158210147822613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12911 + }, + { + "epoch": 5.158609668397922, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12912 + }, + { + "epoch": 5.159009188973232, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12913 + }, + { + "epoch": 5.159408709548542, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12914 + }, + { + "epoch": 5.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12915 + }, + { + "epoch": 5.160207750699161, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12916 + }, + { + "epoch": 5.160607271274471, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12917 + }, + { + "epoch": 5.16100679184978, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12918 + }, + { + "epoch": 5.16140631242509, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12919 + }, + { + "epoch": 5.1618058330003995, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12920 + }, + { + "epoch": 5.162205353575709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12921 + }, + { + "epoch": 5.162604874151019, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12922 + }, + { + "epoch": 5.163004394726328, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12923 + }, + { + "epoch": 5.163403915301638, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12924 + }, + { + "epoch": 5.163803435876948, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12925 + }, + { + "epoch": 5.164202956452257, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12926 + }, + { + "epoch": 5.164602477027567, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12927 + }, + { + "epoch": 5.165001997602877, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12928 + }, + { + "epoch": 5.165401518178186, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12929 + }, + { + "epoch": 5.165801038753496, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12930 + }, + { + "epoch": 5.166200559328805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12931 + }, + { + "epoch": 5.166600079904115, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12932 + }, + { + "epoch": 5.166999600479425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12933 + }, + { + "epoch": 5.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12934 + }, + { + "epoch": 5.167798641630044, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12935 + }, + { + "epoch": 5.168198162205353, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12936 + }, + { + "epoch": 5.168597682780663, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12937 + }, + { + "epoch": 5.168997203355973, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12938 + }, + { + "epoch": 5.169396723931283, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12939 + }, + { + "epoch": 5.169796244506592, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12940 + }, + { + "epoch": 5.1701957650819015, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12941 + }, + { + "epoch": 5.170595285657211, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12942 + }, + { + "epoch": 5.170994806232521, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12943 + }, + { + "epoch": 5.171394326807831, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 12944 + }, + { + "epoch": 5.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12945 + }, + { + "epoch": 5.1721933679584495, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12946 + }, + { + "epoch": 5.172592888533759, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12947 + }, + { + "epoch": 5.172992409109069, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12948 + }, + { + "epoch": 5.173391929684379, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 12949 + }, + { + "epoch": 5.173791450259689, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12950 + }, + { + "epoch": 5.174190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12951 + }, + { + "epoch": 5.174590491410307, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12952 + }, + { + "epoch": 5.174990011985617, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12953 + }, + { + "epoch": 5.175389532560927, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12954 + }, + { + "epoch": 5.175789053136237, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12955 + }, + { + "epoch": 5.1761885737115465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12956 + }, + { + "epoch": 5.176588094286855, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12957 + }, + { + "epoch": 5.176987614862165, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12958 + }, + { + "epoch": 5.177387135437475, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12959 + }, + { + "epoch": 5.177786656012785, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12960 + }, + { + "epoch": 5.1781861765880945, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12961 + }, + { + "epoch": 5.178585697163404, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12962 + }, + { + "epoch": 5.178985217738713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12963 + }, + { + "epoch": 5.179384738314023, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12964 + }, + { + "epoch": 5.179784258889333, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12965 + }, + { + "epoch": 5.180183779464643, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12966 + }, + { + "epoch": 5.180583300039952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12967 + }, + { + "epoch": 5.180982820615261, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12968 + }, + { + "epoch": 5.181382341190571, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12969 + }, + { + "epoch": 5.181781861765881, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12970 + }, + { + "epoch": 5.182181382341191, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12971 + }, + { + "epoch": 5.1825809029165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12972 + }, + { + "epoch": 5.18298042349181, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12973 + }, + { + "epoch": 5.183379944067119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12974 + }, + { + "epoch": 5.183779464642429, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12975 + }, + { + "epoch": 5.184178985217739, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12976 + }, + { + "epoch": 5.1845785057930485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12977 + }, + { + "epoch": 5.184978026368358, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12978 + }, + { + "epoch": 5.185377546943668, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 12979 + }, + { + "epoch": 5.185777067518977, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12980 + }, + { + "epoch": 5.186176588094287, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12981 + }, + { + "epoch": 5.1865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12982 + }, + { + "epoch": 5.186975629244906, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12983 + }, + { + "epoch": 5.187375149820216, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12984 + }, + { + "epoch": 5.187774670395525, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12985 + }, + { + "epoch": 5.188174190970835, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12986 + }, + { + "epoch": 5.1885737115461446, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12987 + }, + { + "epoch": 5.188973232121454, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12988 + }, + { + "epoch": 5.189372752696764, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12989 + }, + { + "epoch": 5.189772273272074, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12990 + }, + { + "epoch": 5.190171793847383, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12991 + }, + { + "epoch": 5.190571314422693, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12992 + }, + { + "epoch": 5.190970834998002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12993 + }, + { + "epoch": 5.191370355573312, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 12994 + }, + { + "epoch": 5.191769876148622, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12995 + }, + { + "epoch": 5.192169396723932, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12996 + }, + { + "epoch": 5.192568917299241, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12997 + }, + { + "epoch": 5.19296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12998 + }, + { + "epoch": 5.19336795844986, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12999 + }, + { + "epoch": 5.19376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13000 + }, + { + "epoch": 5.19416699960048, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13001 + }, + { + "epoch": 5.194566520175789, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13002 + }, + { + "epoch": 5.1949660407510985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13003 + }, + { + "epoch": 5.195365561326408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13004 + }, + { + "epoch": 5.195765081901718, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13005 + }, + { + "epoch": 5.196164602477028, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13006 + }, + { + "epoch": 5.196564123052338, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13007 + }, + { + "epoch": 5.1969636436276465, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13008 + }, + { + "epoch": 5.197363164202956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13009 + }, + { + "epoch": 5.197762684778266, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13010 + }, + { + "epoch": 5.198162205353576, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13011 + }, + { + "epoch": 5.198561725928886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13012 + }, + { + "epoch": 5.1989612465041946, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13013 + }, + { + "epoch": 5.199360767079504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13014 + }, + { + "epoch": 5.199760287654814, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13015 + }, + { + "epoch": 5.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13016 + }, + { + "epoch": 5.200559328805434, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13017 + }, + { + "epoch": 5.2009588493807435, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13018 + }, + { + "epoch": 5.201358369956052, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13019 + }, + { + "epoch": 5.201757890531362, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13020 + }, + { + "epoch": 5.202157411106672, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13021 + }, + { + "epoch": 5.202556931681982, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13022 + }, + { + "epoch": 5.2029564522572915, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 13023 + }, + { + "epoch": 5.2033559728326, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13024 + }, + { + "epoch": 5.20375549340791, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13025 + }, + { + "epoch": 5.20415501398322, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13026 + }, + { + "epoch": 5.20455453455853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 13027 + }, + { + "epoch": 5.20495405513384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13028 + }, + { + "epoch": 5.205353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13029 + }, + { + "epoch": 5.205753096284458, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13030 + }, + { + "epoch": 5.206152616859768, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13031 + }, + { + "epoch": 5.206552137435078, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13032 + }, + { + "epoch": 5.206951658010388, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13033 + }, + { + "epoch": 5.207351178585697, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13034 + }, + { + "epoch": 5.207750699161007, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13035 + }, + { + "epoch": 5.208150219736316, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13036 + }, + { + "epoch": 5.208549740311626, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13037 + }, + { + "epoch": 5.208949260886936, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 13038 + }, + { + "epoch": 5.2093487814622454, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13039 + }, + { + "epoch": 5.209748302037555, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13040 + }, + { + "epoch": 5.210147822612864, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13041 + }, + { + "epoch": 5.210547343188174, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13042 + }, + { + "epoch": 5.210946863763484, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13043 + }, + { + "epoch": 5.2113463843387935, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13044 + }, + { + "epoch": 5.211745904914103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13045 + }, + { + "epoch": 5.212145425489413, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13046 + }, + { + "epoch": 5.212544946064722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13047 + }, + { + "epoch": 5.212944466640032, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13048 + }, + { + "epoch": 5.2133439872153415, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 13049 + }, + { + "epoch": 5.213743507790651, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13050 + }, + { + "epoch": 5.214143028365961, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13051 + }, + { + "epoch": 5.214542548941271, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13052 + }, + { + "epoch": 5.21494206951658, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13053 + }, + { + "epoch": 5.21534159009189, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13054 + }, + { + "epoch": 5.215741110667199, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13055 + }, + { + "epoch": 5.216140631242509, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 13056 + }, + { + "epoch": 5.216540151817819, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13057 + }, + { + "epoch": 5.216939672393128, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13058 + }, + { + "epoch": 5.217339192968438, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13059 + }, + { + "epoch": 5.217738713543747, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13060 + }, + { + "epoch": 5.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13061 + }, + { + "epoch": 5.218537754694367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 13062 + }, + { + "epoch": 5.218937275269677, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13063 + }, + { + "epoch": 5.219336795844986, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13064 + }, + { + "epoch": 5.2197363164202955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13065 + }, + { + "epoch": 5.220135836995605, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13066 + }, + { + "epoch": 5.220535357570915, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13067 + }, + { + "epoch": 5.220934878146225, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13068 + }, + { + "epoch": 5.221334398721535, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13069 + }, + { + "epoch": 5.2217339192968435, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13070 + }, + { + "epoch": 5.222133439872153, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13071 + }, + { + "epoch": 5.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13072 + }, + { + "epoch": 5.222932481022773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13073 + }, + { + "epoch": 5.223332001598083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13074 + }, + { + "epoch": 5.2237315221733915, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13075 + }, + { + "epoch": 5.224131042748701, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13076 + }, + { + "epoch": 5.224530563324011, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13077 + }, + { + "epoch": 5.224930083899321, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13078 + }, + { + "epoch": 5.225329604474631, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13079 + }, + { + "epoch": 5.2257291250499405, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13080 + }, + { + "epoch": 5.226128645625249, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13081 + }, + { + "epoch": 5.226528166200559, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13082 + }, + { + "epoch": 5.226927686775869, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13083 + }, + { + "epoch": 5.227327207351179, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13084 + }, + { + "epoch": 5.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13085 + }, + { + "epoch": 5.228126248501797, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13086 + }, + { + "epoch": 5.228525769077107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13087 + }, + { + "epoch": 5.228925289652417, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13088 + }, + { + "epoch": 5.229324810227727, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13089 + }, + { + "epoch": 5.229724330803037, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13090 + }, + { + "epoch": 5.230123851378346, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13091 + }, + { + "epoch": 5.230523371953655, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13092 + }, + { + "epoch": 5.230922892528965, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13093 + }, + { + "epoch": 5.231322413104275, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13094 + }, + { + "epoch": 5.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13095 + }, + { + "epoch": 5.232121454254894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13096 + }, + { + "epoch": 5.232520974830204, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13097 + }, + { + "epoch": 5.232920495405513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13098 + }, + { + "epoch": 5.233320015980823, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13099 + }, + { + "epoch": 5.233719536556133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13100 + }, + { + "epoch": 5.234119057131442, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13101 + }, + { + "epoch": 5.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13102 + }, + { + "epoch": 5.234918098282061, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13103 + }, + { + "epoch": 5.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13104 + }, + { + "epoch": 5.235717139432681, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13105 + }, + { + "epoch": 5.2361166600079905, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13106 + }, + { + "epoch": 5.2365161805833, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13107 + }, + { + "epoch": 5.23691570115861, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13108 + }, + { + "epoch": 5.237315221733919, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13109 + }, + { + "epoch": 5.237714742309229, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13110 + }, + { + "epoch": 5.2381142628845385, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13111 + }, + { + "epoch": 5.238513783459848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13112 + }, + { + "epoch": 5.238913304035158, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13113 + }, + { + "epoch": 5.239312824610467, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13114 + }, + { + "epoch": 5.239712345185777, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13115 + }, + { + "epoch": 5.240111865761087, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13116 + }, + { + "epoch": 5.240511386336396, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13117 + }, + { + "epoch": 5.240910906911706, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13118 + }, + { + "epoch": 5.241310427487016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13119 + }, + { + "epoch": 5.241709948062325, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13120 + }, + { + "epoch": 5.242109468637635, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13121 + }, + { + "epoch": 5.242508989212944, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13122 + }, + { + "epoch": 5.242908509788254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13123 + }, + { + "epoch": 5.243308030363564, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13124 + }, + { + "epoch": 5.243707550938874, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13125 + }, + { + "epoch": 5.244107071514183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13126 + }, + { + "epoch": 5.244506592089492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13127 + }, + { + "epoch": 5.244906112664802, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13128 + }, + { + "epoch": 5.245305633240112, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 13129 + }, + { + "epoch": 5.245705153815422, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13130 + }, + { + "epoch": 5.246104674390731, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13131 + }, + { + "epoch": 5.2465041949660405, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13132 + }, + { + "epoch": 5.24690371554135, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13133 + }, + { + "epoch": 5.24730323611666, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13134 + }, + { + "epoch": 5.24770275669197, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13135 + }, + { + "epoch": 5.24810227726728, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13136 + }, + { + "epoch": 5.2485017978425885, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13137 + }, + { + "epoch": 5.248901318417898, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13138 + }, + { + "epoch": 5.249300838993208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13139 + }, + { + "epoch": 5.249700359568518, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13140 + }, + { + "epoch": 5.250099880143828, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13141 + }, + { + "epoch": 5.2504994007191375, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13142 + }, + { + "epoch": 5.250898921294446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13143 + }, + { + "epoch": 5.251298441869756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13144 + }, + { + "epoch": 5.251697962445066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13145 + }, + { + "epoch": 5.252097483020376, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13146 + }, + { + "epoch": 5.2524970035956855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13147 + }, + { + "epoch": 5.252896524170994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13148 + }, + { + "epoch": 5.253296044746304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13149 + }, + { + "epoch": 5.253695565321614, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13150 + }, + { + "epoch": 5.254095085896924, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13151 + }, + { + "epoch": 5.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13152 + }, + { + "epoch": 5.254894127047543, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13153 + }, + { + "epoch": 5.255293647622852, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13154 + }, + { + "epoch": 5.255693168198162, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13155 + }, + { + "epoch": 5.256092688773472, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13156 + }, + { + "epoch": 5.256492209348782, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13157 + }, + { + "epoch": 5.256891729924091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13158 + }, + { + "epoch": 5.257291250499401, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13159 + }, + { + "epoch": 5.25769077107471, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13160 + }, + { + "epoch": 5.25809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13161 + }, + { + "epoch": 5.25848981222533, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13162 + }, + { + "epoch": 5.258889332800639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13163 + }, + { + "epoch": 5.259288853375949, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13164 + }, + { + "epoch": 5.259688373951258, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13165 + }, + { + "epoch": 5.260087894526568, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13166 + }, + { + "epoch": 5.260487415101878, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13167 + }, + { + "epoch": 5.2608869356771875, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13168 + }, + { + "epoch": 5.261286456252497, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13169 + }, + { + "epoch": 5.261685976827807, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13170 + }, + { + "epoch": 5.262085497403116, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13171 + }, + { + "epoch": 5.262485017978426, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13172 + }, + { + "epoch": 5.2628845385537355, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13173 + }, + { + "epoch": 5.263284059129045, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13174 + }, + { + "epoch": 5.263683579704355, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13175 + }, + { + "epoch": 5.264083100279665, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13176 + }, + { + "epoch": 5.264482620854974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13177 + }, + { + "epoch": 5.2648821414302835, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13178 + }, + { + "epoch": 5.265281662005593, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13179 + }, + { + "epoch": 5.265681182580903, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13180 + }, + { + "epoch": 5.266080703156213, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13181 + }, + { + "epoch": 5.266480223731522, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13182 + }, + { + "epoch": 5.266879744306832, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13183 + }, + { + "epoch": 5.267279264882141, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13184 + }, + { + "epoch": 5.267678785457451, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13185 + }, + { + "epoch": 5.268078306032761, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13186 + }, + { + "epoch": 5.26847782660807, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13187 + }, + { + "epoch": 5.26887734718338, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13188 + }, + { + "epoch": 5.269276867758689, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13189 + }, + { + "epoch": 5.269676388333999, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13190 + }, + { + "epoch": 5.270075908909309, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13191 + }, + { + "epoch": 5.270475429484619, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13192 + }, + { + "epoch": 5.270874950059928, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13193 + }, + { + "epoch": 5.2712744706352375, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13194 + }, + { + "epoch": 5.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13195 + }, + { + "epoch": 5.272073511785857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13196 + }, + { + "epoch": 5.272473032361167, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13197 + }, + { + "epoch": 5.272872552936477, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13198 + }, + { + "epoch": 5.2732720735117855, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13199 + }, + { + "epoch": 5.273671594087095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13200 + }, + { + "epoch": 5.274071114662405, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13201 + }, + { + "epoch": 5.274470635237715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13202 + }, + { + "epoch": 5.274870155813025, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13203 + }, + { + "epoch": 5.2752696763883336, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 13204 + }, + { + "epoch": 5.275669196963643, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13205 + }, + { + "epoch": 5.276068717538953, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13206 + }, + { + "epoch": 5.276468238114263, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13207 + }, + { + "epoch": 5.276867758689573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13208 + }, + { + "epoch": 5.2772672792648825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13209 + }, + { + "epoch": 5.277666799840191, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13210 + }, + { + "epoch": 5.278066320415501, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13211 + }, + { + "epoch": 5.278465840990811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 13212 + }, + { + "epoch": 5.278865361566121, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13213 + }, + { + "epoch": 5.2792648821414305, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13214 + }, + { + "epoch": 5.27966440271674, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13215 + }, + { + "epoch": 5.280063923292049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13216 + }, + { + "epoch": 5.280463443867359, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13217 + }, + { + "epoch": 5.280862964442669, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13218 + }, + { + "epoch": 5.281262485017979, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13219 + }, + { + "epoch": 5.281662005593288, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13220 + }, + { + "epoch": 5.282061526168597, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13221 + }, + { + "epoch": 5.282461046743907, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13222 + }, + { + "epoch": 5.282860567319217, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13223 + }, + { + "epoch": 5.283260087894527, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13224 + }, + { + "epoch": 5.283659608469836, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13225 + }, + { + "epoch": 5.284059129045146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13226 + }, + { + "epoch": 5.284458649620455, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13227 + }, + { + "epoch": 5.284858170195765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 13228 + }, + { + "epoch": 5.285257690771075, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 13229 + }, + { + "epoch": 5.285657211346384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13230 + }, + { + "epoch": 5.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13231 + }, + { + "epoch": 5.286456252497004, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13232 + }, + { + "epoch": 5.286855773072313, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 13233 + }, + { + "epoch": 5.287255293647623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13234 + }, + { + "epoch": 5.2876548142229325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13235 + }, + { + "epoch": 5.288054334798242, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13236 + }, + { + "epoch": 5.288453855373552, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13237 + }, + { + "epoch": 5.288853375948861, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13238 + }, + { + "epoch": 5.289252896524171, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13239 + }, + { + "epoch": 5.2896524170994805, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13240 + }, + { + "epoch": 5.29005193767479, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13241 + }, + { + "epoch": 5.2904514582501, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13242 + }, + { + "epoch": 5.29085097882541, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13243 + }, + { + "epoch": 5.291250499400719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 13244 + }, + { + "epoch": 5.291650019976029, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13245 + }, + { + "epoch": 5.292049540551338, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13246 + }, + { + "epoch": 5.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13247 + }, + { + "epoch": 5.292848581701958, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13248 + }, + { + "epoch": 5.293248102277268, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13249 + }, + { + "epoch": 5.293647622852577, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13250 + }, + { + "epoch": 5.294047143427886, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13251 + }, + { + "epoch": 5.294446664003196, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13252 + }, + { + "epoch": 5.294846184578506, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13253 + }, + { + "epoch": 5.295245705153816, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13254 + }, + { + "epoch": 5.295645225729125, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13255 + }, + { + "epoch": 5.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13256 + }, + { + "epoch": 5.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.2965, + "step": 13257 + }, + { + "epoch": 5.296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13258 + }, + { + "epoch": 5.297243308030364, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13259 + }, + { + "epoch": 5.297642828605674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13260 + }, + { + "epoch": 5.2980423491809825, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13261 + }, + { + "epoch": 5.298441869756292, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13262 + }, + { + "epoch": 5.298841390331602, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13263 + }, + { + "epoch": 5.299240910906912, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13264 + }, + { + "epoch": 5.299640431482222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13265 + }, + { + "epoch": 5.300039952057531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13266 + }, + { + "epoch": 5.30043947263284, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13267 + }, + { + "epoch": 5.30083899320815, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13268 + }, + { + "epoch": 5.30123851378346, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13269 + }, + { + "epoch": 5.30163803435877, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13270 + }, + { + "epoch": 5.3020375549340795, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13271 + }, + { + "epoch": 5.302437075509388, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13272 + }, + { + "epoch": 5.302836596084698, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13273 + }, + { + "epoch": 5.303236116660008, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13274 + }, + { + "epoch": 5.303635637235318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13275 + }, + { + "epoch": 5.3040351578106275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13276 + }, + { + "epoch": 5.304434678385936, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13277 + }, + { + "epoch": 5.304834198961246, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13278 + }, + { + "epoch": 5.305233719536556, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13279 + }, + { + "epoch": 5.305633240111866, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13280 + }, + { + "epoch": 5.3060327606871756, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13281 + }, + { + "epoch": 5.306432281262485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13282 + }, + { + "epoch": 5.306831801837794, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13283 + }, + { + "epoch": 5.307231322413104, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13284 + }, + { + "epoch": 5.307630842988414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13285 + }, + { + "epoch": 5.308030363563724, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13286 + }, + { + "epoch": 5.308429884139033, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13287 + }, + { + "epoch": 5.308829404714343, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13288 + }, + { + "epoch": 5.309228925289652, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13289 + }, + { + "epoch": 5.309628445864962, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13290 + }, + { + "epoch": 5.310027966440272, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13291 + }, + { + "epoch": 5.310427487015581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 13292 + }, + { + "epoch": 5.310827007590891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13293 + }, + { + "epoch": 5.3112265281662, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13294 + }, + { + "epoch": 5.31162604874151, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13295 + }, + { + "epoch": 5.31202556931682, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13296 + }, + { + "epoch": 5.3124250898921295, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13297 + }, + { + "epoch": 5.312824610467439, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13298 + }, + { + "epoch": 5.313224131042749, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13299 + }, + { + "epoch": 5.313623651618058, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13300 + }, + { + "epoch": 5.314023172193368, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13301 + }, + { + "epoch": 5.3144226927686775, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13302 + }, + { + "epoch": 5.314822213343987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13303 + }, + { + "epoch": 5.315221733919297, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13304 + }, + { + "epoch": 5.315621254494607, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13305 + }, + { + "epoch": 5.316020775069916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13306 + }, + { + "epoch": 5.316420295645226, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13307 + }, + { + "epoch": 5.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13308 + }, + { + "epoch": 5.317219336795845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13309 + }, + { + "epoch": 5.317618857371155, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13310 + }, + { + "epoch": 5.318018377946464, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13311 + }, + { + "epoch": 5.318417898521774, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13312 + }, + { + "epoch": 5.318817419097083, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13313 + }, + { + "epoch": 5.319216939672393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 13314 + }, + { + "epoch": 5.319616460247703, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13315 + }, + { + "epoch": 5.320015980823013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 13316 + }, + { + "epoch": 5.320415501398322, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13317 + }, + { + "epoch": 5.320815021973631, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13318 + }, + { + "epoch": 5.321214542548941, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13319 + }, + { + "epoch": 5.321614063124251, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13320 + }, + { + "epoch": 5.322013583699561, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13321 + }, + { + "epoch": 5.322413104274871, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13322 + }, + { + "epoch": 5.3228126248501795, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13323 + }, + { + "epoch": 5.323212145425489, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13324 + }, + { + "epoch": 5.323611666000799, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13325 + }, + { + "epoch": 5.324011186576109, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13326 + }, + { + "epoch": 5.324410707151419, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13327 + }, + { + "epoch": 5.3248102277267275, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13328 + }, + { + "epoch": 5.325209748302037, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13329 + }, + { + "epoch": 5.325609268877347, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13330 + }, + { + "epoch": 5.326008789452657, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13331 + }, + { + "epoch": 5.326408310027967, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13332 + }, + { + "epoch": 5.3268078306032765, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13333 + }, + { + "epoch": 5.327207351178585, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13334 + }, + { + "epoch": 5.327606871753895, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13335 + }, + { + "epoch": 5.328006392329205, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 13336 + }, + { + "epoch": 5.328405912904515, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13337 + }, + { + "epoch": 5.3288054334798245, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13338 + }, + { + "epoch": 5.329204954055134, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13339 + }, + { + "epoch": 5.329604474630443, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13340 + }, + { + "epoch": 5.330003995205753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13341 + }, + { + "epoch": 5.330403515781063, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13342 + }, + { + "epoch": 5.3308030363563725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13343 + }, + { + "epoch": 5.331202556931682, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13344 + }, + { + "epoch": 5.331602077506991, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13345 + }, + { + "epoch": 5.332001598082301, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13346 + }, + { + "epoch": 5.332401118657611, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13347 + }, + { + "epoch": 5.332800639232921, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13348 + }, + { + "epoch": 5.33320015980823, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13349 + }, + { + "epoch": 5.33359968038354, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13350 + }, + { + "epoch": 5.333999200958849, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13351 + }, + { + "epoch": 5.334398721534159, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13352 + }, + { + "epoch": 5.334798242109469, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13353 + }, + { + "epoch": 5.335197762684778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13354 + }, + { + "epoch": 5.335597283260088, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13355 + }, + { + "epoch": 5.335996803835398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13356 + }, + { + "epoch": 5.336396324410707, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13357 + }, + { + "epoch": 5.336795844986017, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13358 + }, + { + "epoch": 5.3371953655613265, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13359 + }, + { + "epoch": 5.337594886136636, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13360 + }, + { + "epoch": 5.337994406711946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13361 + }, + { + "epoch": 5.338393927287255, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13362 + }, + { + "epoch": 5.338793447862565, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13363 + }, + { + "epoch": 5.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13364 + }, + { + "epoch": 5.339592489013184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13365 + }, + { + "epoch": 5.339992009588494, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13366 + }, + { + "epoch": 5.340391530163803, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13367 + }, + { + "epoch": 5.340791050739113, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13368 + }, + { + "epoch": 5.3411905713144225, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13369 + }, + { + "epoch": 5.341590091889732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13370 + }, + { + "epoch": 5.341989612465042, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13371 + }, + { + "epoch": 5.342389133040352, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13372 + }, + { + "epoch": 5.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13373 + }, + { + "epoch": 5.343188174190971, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 13374 + }, + { + "epoch": 5.34358769476628, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13375 + }, + { + "epoch": 5.34398721534159, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13376 + }, + { + "epoch": 5.3443867359169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13377 + }, + { + "epoch": 5.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13378 + }, + { + "epoch": 5.345185777067519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13379 + }, + { + "epoch": 5.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13380 + }, + { + "epoch": 5.345984818218138, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13381 + }, + { + "epoch": 5.346384338793448, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13382 + }, + { + "epoch": 5.346783859368758, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13383 + }, + { + "epoch": 5.347183379944067, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13384 + }, + { + "epoch": 5.3475829005193765, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13385 + }, + { + "epoch": 5.347982421094686, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13386 + }, + { + "epoch": 5.348381941669996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13387 + }, + { + "epoch": 5.348781462245306, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13388 + }, + { + "epoch": 5.349180982820616, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13389 + }, + { + "epoch": 5.3495805033959245, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13390 + }, + { + "epoch": 5.349980023971234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13391 + }, + { + "epoch": 5.350379544546544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13392 + }, + { + "epoch": 5.350779065121854, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13393 + }, + { + "epoch": 5.351178585697164, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13394 + }, + { + "epoch": 5.351578106272473, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13395 + }, + { + "epoch": 5.351977626847782, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13396 + }, + { + "epoch": 5.352377147423092, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13397 + }, + { + "epoch": 5.352776667998402, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13398 + }, + { + "epoch": 5.353176188573712, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13399 + }, + { + "epoch": 5.3535757091490215, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13400 + }, + { + "epoch": 5.35397522972433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13401 + }, + { + "epoch": 5.35437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13402 + }, + { + "epoch": 5.35477427087495, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13403 + }, + { + "epoch": 5.35517379145026, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13404 + }, + { + "epoch": 5.3555733120255695, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13405 + }, + { + "epoch": 5.355972832600879, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13406 + }, + { + "epoch": 5.356372353176188, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13407 + }, + { + "epoch": 5.356771873751498, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 13408 + }, + { + "epoch": 5.357171394326808, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13409 + }, + { + "epoch": 5.357570914902118, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13410 + }, + { + "epoch": 5.357970435477427, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13411 + }, + { + "epoch": 5.358369956052737, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13412 + }, + { + "epoch": 5.358769476628046, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13413 + }, + { + "epoch": 5.359168997203356, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13414 + }, + { + "epoch": 5.359568517778666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13415 + }, + { + "epoch": 5.359968038353975, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 13416 + }, + { + "epoch": 5.360367558929285, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13417 + }, + { + "epoch": 5.360767079504594, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13418 + }, + { + "epoch": 5.361166600079904, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13419 + }, + { + "epoch": 5.361566120655214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13420 + }, + { + "epoch": 5.361965641230523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13421 + }, + { + "epoch": 5.362365161805833, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13422 + }, + { + "epoch": 5.362764682381143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13423 + }, + { + "epoch": 5.363164202956452, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13424 + }, + { + "epoch": 5.363563723531762, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 13425 + }, + { + "epoch": 5.3639632441070715, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13426 + }, + { + "epoch": 5.364362764682381, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13427 + }, + { + "epoch": 5.364762285257691, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13428 + }, + { + "epoch": 5.365161805833001, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13429 + }, + { + "epoch": 5.36556132640831, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 13430 + }, + { + "epoch": 5.3659608469836195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13431 + }, + { + "epoch": 5.366360367558929, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13432 + }, + { + "epoch": 5.366759888134239, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13433 + }, + { + "epoch": 5.367159408709549, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13434 + }, + { + "epoch": 5.367558929284858, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 13435 + }, + { + "epoch": 5.367958449860168, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13436 + }, + { + "epoch": 5.368357970435477, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 13437 + }, + { + "epoch": 5.368757491010787, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13438 + }, + { + "epoch": 5.369157011586097, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13439 + }, + { + "epoch": 5.369556532161407, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13440 + }, + { + "epoch": 5.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13441 + }, + { + "epoch": 5.370355573312025, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13442 + }, + { + "epoch": 5.370755093887335, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 13443 + }, + { + "epoch": 5.371154614462645, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13444 + }, + { + "epoch": 5.371554135037955, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13445 + }, + { + "epoch": 5.3719536556132645, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13446 + }, + { + "epoch": 5.372353176188573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13447 + }, + { + "epoch": 5.372752696763883, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13448 + }, + { + "epoch": 5.373152217339193, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13449 + }, + { + "epoch": 5.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13450 + }, + { + "epoch": 5.373951258489813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13451 + }, + { + "epoch": 5.3743507790651215, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13452 + }, + { + "epoch": 5.374750299640431, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13453 + }, + { + "epoch": 5.375149820215741, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13454 + }, + { + "epoch": 5.375549340791051, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13455 + }, + { + "epoch": 5.375948861366361, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13456 + }, + { + "epoch": 5.3763483819416695, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13457 + }, + { + "epoch": 5.376747902516979, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13458 + }, + { + "epoch": 5.377147423092289, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13459 + }, + { + "epoch": 5.377546943667599, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13460 + }, + { + "epoch": 5.377946464242909, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13461 + }, + { + "epoch": 5.3783459848182185, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13462 + }, + { + "epoch": 5.378745505393527, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13463 + }, + { + "epoch": 5.379145025968837, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13464 + }, + { + "epoch": 5.379544546544147, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13465 + }, + { + "epoch": 5.379944067119457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13466 + }, + { + "epoch": 5.3803435876947665, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13467 + }, + { + "epoch": 5.380743108270076, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13468 + }, + { + "epoch": 5.381142628845385, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13469 + }, + { + "epoch": 5.381542149420695, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13470 + }, + { + "epoch": 5.381941669996005, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13471 + }, + { + "epoch": 5.3823411905713145, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13472 + }, + { + "epoch": 5.382740711146624, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13473 + }, + { + "epoch": 5.383140231721933, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13474 + }, + { + "epoch": 5.383539752297243, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13475 + }, + { + "epoch": 5.383939272872553, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13476 + }, + { + "epoch": 5.384338793447863, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13477 + }, + { + "epoch": 5.384738314023172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13478 + }, + { + "epoch": 5.385137834598482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13479 + }, + { + "epoch": 5.385537355173791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13480 + }, + { + "epoch": 5.385936875749101, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13481 + }, + { + "epoch": 5.386336396324411, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13482 + }, + { + "epoch": 5.38673591689972, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13483 + }, + { + "epoch": 5.38713543747503, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13484 + }, + { + "epoch": 5.38753495805034, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13485 + }, + { + "epoch": 5.387934478625649, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13486 + }, + { + "epoch": 5.388333999200959, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 13487 + }, + { + "epoch": 5.3887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13488 + }, + { + "epoch": 5.389133040351578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13489 + }, + { + "epoch": 5.389532560926888, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13490 + }, + { + "epoch": 5.389932081502197, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13491 + }, + { + "epoch": 5.390331602077507, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13492 + }, + { + "epoch": 5.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13493 + }, + { + "epoch": 5.391130643228126, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13494 + }, + { + "epoch": 5.391530163803436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13495 + }, + { + "epoch": 5.391929684378746, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13496 + }, + { + "epoch": 5.392329204954055, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13497 + }, + { + "epoch": 5.3927287255293646, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 13498 + }, + { + "epoch": 5.393128246104674, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13499 + }, + { + "epoch": 5.393527766679984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13500 + }, + { + "epoch": 5.393927287255294, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13501 + }, + { + "epoch": 5.394326807830604, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13502 + }, + { + "epoch": 5.394726328405913, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13503 + }, + { + "epoch": 5.395125848981222, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13504 + }, + { + "epoch": 5.395525369556532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13505 + }, + { + "epoch": 5.395924890131842, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13506 + }, + { + "epoch": 5.396324410707152, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13507 + }, + { + "epoch": 5.396723931282461, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13508 + }, + { + "epoch": 5.39712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13509 + }, + { + "epoch": 5.39752297243308, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13510 + }, + { + "epoch": 5.39792249300839, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13511 + }, + { + "epoch": 5.3983220135837, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13512 + }, + { + "epoch": 5.39872153415901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13513 + }, + { + "epoch": 5.3991210547343185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13514 + }, + { + "epoch": 5.399520575309628, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13515 + }, + { + "epoch": 5.399920095884938, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13516 + }, + { + "epoch": 5.400319616460248, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13517 + }, + { + "epoch": 5.400719137035558, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13518 + }, + { + "epoch": 5.401118657610867, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13519 + }, + { + "epoch": 5.401518178186176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13520 + }, + { + "epoch": 5.401917698761486, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13521 + }, + { + "epoch": 5.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13522 + }, + { + "epoch": 5.402716739912106, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13523 + }, + { + "epoch": 5.4031162604874154, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13524 + }, + { + "epoch": 5.403515781062724, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13525 + }, + { + "epoch": 5.403915301638034, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13526 + }, + { + "epoch": 5.404314822213344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13527 + }, + { + "epoch": 5.404714342788654, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13528 + }, + { + "epoch": 5.4051138633639635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13529 + }, + { + "epoch": 5.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.2988, + "step": 13530 + }, + { + "epoch": 5.405912904514582, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13531 + }, + { + "epoch": 5.406312425089892, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13532 + }, + { + "epoch": 5.406711945665202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13533 + }, + { + "epoch": 5.4071114662405115, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13534 + }, + { + "epoch": 5.407510986815821, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13535 + }, + { + "epoch": 5.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13536 + }, + { + "epoch": 5.40831002796644, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 13537 + }, + { + "epoch": 5.40870954854175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13538 + }, + { + "epoch": 5.40910906911706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13539 + }, + { + "epoch": 5.409508589692369, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13540 + }, + { + "epoch": 5.409908110267679, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13541 + }, + { + "epoch": 5.410307630842988, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13542 + }, + { + "epoch": 5.410707151418298, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13543 + }, + { + "epoch": 5.411106671993608, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13544 + }, + { + "epoch": 5.411506192568917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13545 + }, + { + "epoch": 5.411905713144227, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13546 + }, + { + "epoch": 5.412305233719536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13547 + }, + { + "epoch": 5.412704754294846, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13548 + }, + { + "epoch": 5.413104274870156, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13549 + }, + { + "epoch": 5.4135037954454654, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13550 + }, + { + "epoch": 5.413903316020775, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13551 + }, + { + "epoch": 5.414302836596085, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13552 + }, + { + "epoch": 5.414702357171394, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 13553 + }, + { + "epoch": 5.415101877746704, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13554 + }, + { + "epoch": 5.4155013983220135, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13555 + }, + { + "epoch": 5.415900918897323, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 13556 + }, + { + "epoch": 5.416300439472633, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13557 + }, + { + "epoch": 5.416699960047943, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13558 + }, + { + "epoch": 5.417099480623252, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13559 + }, + { + "epoch": 5.4174990011985615, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 13560 + }, + { + "epoch": 5.417898521773871, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13561 + }, + { + "epoch": 5.418298042349181, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13562 + }, + { + "epoch": 5.418697562924491, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13563 + }, + { + "epoch": 5.4190970834998, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13564 + }, + { + "epoch": 5.41949660407511, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13565 + }, + { + "epoch": 5.419896124650419, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13566 + }, + { + "epoch": 5.420295645225729, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13567 + }, + { + "epoch": 5.420695165801039, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13568 + }, + { + "epoch": 5.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13569 + }, + { + "epoch": 5.421494206951658, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13570 + }, + { + "epoch": 5.421893727526967, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13571 + }, + { + "epoch": 5.422293248102277, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13572 + }, + { + "epoch": 5.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13573 + }, + { + "epoch": 5.423092289252897, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13574 + }, + { + "epoch": 5.4234918098282066, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13575 + }, + { + "epoch": 5.4238913304035155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13576 + }, + { + "epoch": 5.424290850978825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13577 + }, + { + "epoch": 5.424690371554135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13578 + }, + { + "epoch": 5.425089892129445, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13579 + }, + { + "epoch": 5.425489412704755, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13580 + }, + { + "epoch": 5.4258889332800635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13581 + }, + { + "epoch": 5.426288453855373, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13582 + }, + { + "epoch": 5.426687974430683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13583 + }, + { + "epoch": 5.427087495005993, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13584 + }, + { + "epoch": 5.427487015581303, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13585 + }, + { + "epoch": 5.427886536156612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13586 + }, + { + "epoch": 5.428286056731921, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13587 + }, + { + "epoch": 5.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 13588 + }, + { + "epoch": 5.429085097882541, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13589 + }, + { + "epoch": 5.429484618457851, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13590 + }, + { + "epoch": 5.4298841390331605, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13591 + }, + { + "epoch": 5.43028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13592 + }, + { + "epoch": 5.430683180183779, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13593 + }, + { + "epoch": 5.431082700759089, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13594 + }, + { + "epoch": 5.431482221334399, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13595 + }, + { + "epoch": 5.4318817419097085, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13596 + }, + { + "epoch": 5.432281262485018, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13597 + }, + { + "epoch": 5.432680783060327, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13598 + }, + { + "epoch": 5.433080303635637, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13599 + }, + { + "epoch": 5.433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13600 + }, + { + "epoch": 5.433879344786257, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13601 + }, + { + "epoch": 5.434278865361566, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13602 + }, + { + "epoch": 5.434678385936876, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 13603 + }, + { + "epoch": 5.435077906512185, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13604 + }, + { + "epoch": 5.435477427087495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13605 + }, + { + "epoch": 5.435876947662805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13606 + }, + { + "epoch": 5.436276468238114, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13607 + }, + { + "epoch": 5.436675988813424, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13608 + }, + { + "epoch": 5.437075509388734, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13609 + }, + { + "epoch": 5.437475029964043, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13610 + }, + { + "epoch": 5.437874550539353, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13611 + }, + { + "epoch": 5.438274071114662, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13612 + }, + { + "epoch": 5.438673591689972, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13613 + }, + { + "epoch": 5.439073112265282, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13614 + }, + { + "epoch": 5.439472632840591, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13615 + }, + { + "epoch": 5.439872153415901, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13616 + }, + { + "epoch": 5.4402716739912105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 13617 + }, + { + "epoch": 5.44067119456652, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13618 + }, + { + "epoch": 5.44107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 13619 + }, + { + "epoch": 5.44147023571714, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13620 + }, + { + "epoch": 5.441869756292449, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13621 + }, + { + "epoch": 5.4422692768677585, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13622 + }, + { + "epoch": 5.442668797443068, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13623 + }, + { + "epoch": 5.443068318018378, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13624 + }, + { + "epoch": 5.443467838593688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13625 + }, + { + "epoch": 5.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13626 + }, + { + "epoch": 5.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13627 + }, + { + "epoch": 5.444666400319616, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13628 + }, + { + "epoch": 5.445065920894926, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13629 + }, + { + "epoch": 5.445465441470236, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13630 + }, + { + "epoch": 5.445864962045546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13631 + }, + { + "epoch": 5.446264482620855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13632 + }, + { + "epoch": 5.446664003196164, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13633 + }, + { + "epoch": 5.447063523771474, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13634 + }, + { + "epoch": 5.447463044346784, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13635 + }, + { + "epoch": 5.447862564922094, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13636 + }, + { + "epoch": 5.448262085497403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13637 + }, + { + "epoch": 5.448661606072712, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13638 + }, + { + "epoch": 5.449061126648022, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13639 + }, + { + "epoch": 5.449460647223332, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13640 + }, + { + "epoch": 5.449860167798642, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13641 + }, + { + "epoch": 5.450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13642 + }, + { + "epoch": 5.4506592089492605, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13643 + }, + { + "epoch": 5.45105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13644 + }, + { + "epoch": 5.45145825009988, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 13645 + }, + { + "epoch": 5.45185777067519, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13646 + }, + { + "epoch": 5.4522572912505, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13647 + }, + { + "epoch": 5.452656811825809, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13648 + }, + { + "epoch": 5.453056332401118, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13649 + }, + { + "epoch": 5.453455852976428, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13650 + }, + { + "epoch": 5.453855373551738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13651 + }, + { + "epoch": 5.454254894127048, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 13652 + }, + { + "epoch": 5.4546544147023575, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13653 + }, + { + "epoch": 5.455053935277666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13654 + }, + { + "epoch": 5.455453455852976, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13655 + }, + { + "epoch": 5.455852976428286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13656 + }, + { + "epoch": 5.456252497003596, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13657 + }, + { + "epoch": 5.4566520175789055, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13658 + }, + { + "epoch": 5.457051538154215, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13659 + }, + { + "epoch": 5.457451058729524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13660 + }, + { + "epoch": 5.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13661 + }, + { + "epoch": 5.458250099880144, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13662 + }, + { + "epoch": 5.4586496204554535, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13663 + }, + { + "epoch": 5.459049141030763, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 13664 + }, + { + "epoch": 5.459448661606073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13665 + }, + { + "epoch": 5.459848182181382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13666 + }, + { + "epoch": 5.460247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13667 + }, + { + "epoch": 5.460647223332002, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 13668 + }, + { + "epoch": 5.461046743907311, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13669 + }, + { + "epoch": 5.461446264482621, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13670 + }, + { + "epoch": 5.46184578505793, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13671 + }, + { + "epoch": 5.46224530563324, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13672 + }, + { + "epoch": 5.46264482620855, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13673 + }, + { + "epoch": 5.463044346783859, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13674 + }, + { + "epoch": 5.463443867359169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13675 + }, + { + "epoch": 5.463843387934479, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13676 + }, + { + "epoch": 5.464242908509788, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13677 + }, + { + "epoch": 5.464642429085098, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13678 + }, + { + "epoch": 5.4650419496604075, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13679 + }, + { + "epoch": 5.465441470235717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13680 + }, + { + "epoch": 5.465840990811027, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13681 + }, + { + "epoch": 5.466240511386337, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13682 + }, + { + "epoch": 5.466640031961646, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13683 + }, + { + "epoch": 5.4670395525369555, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13684 + }, + { + "epoch": 5.467439073112265, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13685 + }, + { + "epoch": 5.467838593687575, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13686 + }, + { + "epoch": 5.468238114262885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13687 + }, + { + "epoch": 5.468637634838194, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13688 + }, + { + "epoch": 5.4690371554135035, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13689 + }, + { + "epoch": 5.469436675988813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13690 + }, + { + "epoch": 5.469836196564123, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13691 + }, + { + "epoch": 5.470235717139433, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13692 + }, + { + "epoch": 5.470635237714743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13693 + }, + { + "epoch": 5.471034758290052, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13694 + }, + { + "epoch": 5.471434278865361, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13695 + }, + { + "epoch": 5.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13696 + }, + { + "epoch": 5.472233320015981, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13697 + }, + { + "epoch": 5.472632840591291, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13698 + }, + { + "epoch": 5.4730323611666005, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13699 + }, + { + "epoch": 5.473431881741909, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13700 + }, + { + "epoch": 5.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13701 + }, + { + "epoch": 5.474230922892529, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13702 + }, + { + "epoch": 5.474630443467839, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13703 + }, + { + "epoch": 5.475029964043149, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 13704 + }, + { + "epoch": 5.4754294846184575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13705 + }, + { + "epoch": 5.475829005193767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13706 + }, + { + "epoch": 5.476228525769077, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13707 + }, + { + "epoch": 5.476628046344387, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13708 + }, + { + "epoch": 5.477027566919697, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13709 + }, + { + "epoch": 5.477427087495006, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13710 + }, + { + "epoch": 5.477826608070315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13711 + }, + { + "epoch": 5.478226128645625, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13712 + }, + { + "epoch": 5.478625649220935, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13713 + }, + { + "epoch": 5.479025169796245, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 13714 + }, + { + "epoch": 5.479424690371554, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13715 + }, + { + "epoch": 5.479824210946864, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13716 + }, + { + "epoch": 5.480223731522173, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 13717 + }, + { + "epoch": 5.480623252097483, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13718 + }, + { + "epoch": 5.481022772672793, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13719 + }, + { + "epoch": 5.4814222932481025, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13720 + }, + { + "epoch": 5.481821813823412, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13721 + }, + { + "epoch": 5.482221334398721, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13722 + }, + { + "epoch": 5.482620854974031, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13723 + }, + { + "epoch": 5.483020375549341, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13724 + }, + { + "epoch": 5.4834198961246505, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13725 + }, + { + "epoch": 5.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 13726 + }, + { + "epoch": 5.484218937275269, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13727 + }, + { + "epoch": 5.484618457850579, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13728 + }, + { + "epoch": 5.485017978425889, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13729 + }, + { + "epoch": 5.485417499001199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13730 + }, + { + "epoch": 5.485817019576508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13731 + }, + { + "epoch": 5.486216540151818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13732 + }, + { + "epoch": 5.486616060727127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13733 + }, + { + "epoch": 5.487015581302437, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13734 + }, + { + "epoch": 5.487415101877747, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13735 + }, + { + "epoch": 5.487814622453056, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13736 + }, + { + "epoch": 5.488214143028366, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13737 + }, + { + "epoch": 5.488613663603676, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13738 + }, + { + "epoch": 5.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13739 + }, + { + "epoch": 5.489412704754295, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13740 + }, + { + "epoch": 5.4898122253296044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13741 + }, + { + "epoch": 5.490211745904914, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13742 + }, + { + "epoch": 5.490611266480224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13743 + }, + { + "epoch": 5.491010787055533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 13744 + }, + { + "epoch": 5.491410307630843, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13745 + }, + { + "epoch": 5.4918098282061525, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13746 + }, + { + "epoch": 5.492209348781462, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13747 + }, + { + "epoch": 5.492608869356772, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13748 + }, + { + "epoch": 5.493008389932082, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13749 + }, + { + "epoch": 5.493407910507391, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13750 + }, + { + "epoch": 5.4938074310827005, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13751 + }, + { + "epoch": 5.49420695165801, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13752 + }, + { + "epoch": 5.49460647223332, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13753 + }, + { + "epoch": 5.49500599280863, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13754 + }, + { + "epoch": 5.49540551338394, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13755 + }, + { + "epoch": 5.495805033959249, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13756 + }, + { + "epoch": 5.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13757 + }, + { + "epoch": 5.496604075109868, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13758 + }, + { + "epoch": 5.497003595685178, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13759 + }, + { + "epoch": 5.497403116260488, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13760 + }, + { + "epoch": 5.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13761 + }, + { + "epoch": 5.498202157411106, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13762 + }, + { + "epoch": 5.498601677986416, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13763 + }, + { + "epoch": 5.499001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13764 + }, + { + "epoch": 5.499400719137036, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13765 + }, + { + "epoch": 5.4998002397123456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13766 + }, + { + "epoch": 5.5001997602876544, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13767 + }, + { + "epoch": 5.500599280862964, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13768 + }, + { + "epoch": 5.500998801438274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13769 + }, + { + "epoch": 5.501398322013584, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13770 + }, + { + "epoch": 5.501797842588894, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13771 + }, + { + "epoch": 5.502197363164203, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13772 + }, + { + "epoch": 5.502596883739512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13773 + }, + { + "epoch": 5.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13774 + }, + { + "epoch": 5.503395924890132, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13775 + }, + { + "epoch": 5.503795445465442, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13776 + }, + { + "epoch": 5.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13777 + }, + { + "epoch": 5.50459448661606, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13778 + }, + { + "epoch": 5.50499400719137, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13779 + }, + { + "epoch": 5.50539352776668, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13780 + }, + { + "epoch": 5.50579304834199, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13781 + }, + { + "epoch": 5.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13782 + }, + { + "epoch": 5.506592089492609, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 13783 + }, + { + "epoch": 5.506991610067918, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13784 + }, + { + "epoch": 5.507391130643228, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13785 + }, + { + "epoch": 5.507790651218538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13786 + }, + { + "epoch": 5.5081901717938475, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13787 + }, + { + "epoch": 5.508589692369157, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13788 + }, + { + "epoch": 5.508989212944467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13789 + }, + { + "epoch": 5.509388733519776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13790 + }, + { + "epoch": 5.509788254095086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13791 + }, + { + "epoch": 5.5101877746703956, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13792 + }, + { + "epoch": 5.510587295245705, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 13793 + }, + { + "epoch": 5.510986815821015, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13794 + }, + { + "epoch": 5.511386336396324, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13795 + }, + { + "epoch": 5.511785856971634, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13796 + }, + { + "epoch": 5.512185377546944, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13797 + }, + { + "epoch": 5.512584898122253, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13798 + }, + { + "epoch": 5.512984418697563, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13799 + }, + { + "epoch": 5.513383939272872, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13800 + }, + { + "epoch": 5.513783459848182, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13801 + }, + { + "epoch": 5.514182980423492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13802 + }, + { + "epoch": 5.514582500998801, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13803 + }, + { + "epoch": 5.514982021574111, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13804 + }, + { + "epoch": 5.515381542149421, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13805 + }, + { + "epoch": 5.515781062724731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13806 + }, + { + "epoch": 5.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13807 + }, + { + "epoch": 5.5165801038753495, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13808 + }, + { + "epoch": 5.516979624450659, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13809 + }, + { + "epoch": 5.517379145025969, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13810 + }, + { + "epoch": 5.517778665601279, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13811 + }, + { + "epoch": 5.518178186176588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13812 + }, + { + "epoch": 5.5185777067518975, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13813 + }, + { + "epoch": 5.518977227327207, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13814 + }, + { + "epoch": 5.519376747902517, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13815 + }, + { + "epoch": 5.519776268477827, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13816 + }, + { + "epoch": 5.520175789053136, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13817 + }, + { + "epoch": 5.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13818 + }, + { + "epoch": 5.520974830203755, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13819 + }, + { + "epoch": 5.521374350779065, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13820 + }, + { + "epoch": 5.521773871354375, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13821 + }, + { + "epoch": 5.522173391929685, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13822 + }, + { + "epoch": 5.5225729125049945, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13823 + }, + { + "epoch": 5.522972433080303, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13824 + }, + { + "epoch": 5.523371953655613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13825 + }, + { + "epoch": 5.523771474230923, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13826 + }, + { + "epoch": 5.524170994806233, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13827 + }, + { + "epoch": 5.5245705153815425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13828 + }, + { + "epoch": 5.524970035956851, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13829 + }, + { + "epoch": 5.525369556532161, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13830 + }, + { + "epoch": 5.525769077107471, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13831 + }, + { + "epoch": 5.526168597682781, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13832 + }, + { + "epoch": 5.526568118258091, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13833 + }, + { + "epoch": 5.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13834 + }, + { + "epoch": 5.527367159408709, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13835 + }, + { + "epoch": 5.527766679984019, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13836 + }, + { + "epoch": 5.528166200559329, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13837 + }, + { + "epoch": 5.528565721134639, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13838 + }, + { + "epoch": 5.528965241709948, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13839 + }, + { + "epoch": 5.529364762285257, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13840 + }, + { + "epoch": 5.529764282860567, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13841 + }, + { + "epoch": 5.530163803435877, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13842 + }, + { + "epoch": 5.530563324011187, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13843 + }, + { + "epoch": 5.5309628445864965, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13844 + }, + { + "epoch": 5.531362365161806, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13845 + }, + { + "epoch": 5.531761885737115, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13846 + }, + { + "epoch": 5.532161406312425, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13847 + }, + { + "epoch": 5.532560926887735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13848 + }, + { + "epoch": 5.5329604474630445, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13849 + }, + { + "epoch": 5.533359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13850 + }, + { + "epoch": 5.533759488613663, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13851 + }, + { + "epoch": 5.534159009188973, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 13852 + }, + { + "epoch": 5.534558529764283, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13853 + }, + { + "epoch": 5.5349580503395925, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13854 + }, + { + "epoch": 5.535357570914902, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13855 + }, + { + "epoch": 5.535757091490212, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13856 + }, + { + "epoch": 5.536156612065521, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13857 + }, + { + "epoch": 5.536556132640831, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13858 + }, + { + "epoch": 5.536955653216141, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13859 + }, + { + "epoch": 5.53735517379145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 13860 + }, + { + "epoch": 5.53775469436676, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 13861 + }, + { + "epoch": 5.53815421494207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13862 + }, + { + "epoch": 5.538553735517379, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 13863 + }, + { + "epoch": 5.538953256092689, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13864 + }, + { + "epoch": 5.539352776667998, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13865 + }, + { + "epoch": 5.539752297243308, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13866 + }, + { + "epoch": 5.540151817818618, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 13867 + }, + { + "epoch": 5.540551338393927, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13868 + }, + { + "epoch": 5.540950858969237, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13869 + }, + { + "epoch": 5.5413503795445465, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13870 + }, + { + "epoch": 5.541749900119856, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13871 + }, + { + "epoch": 5.542149420695166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13872 + }, + { + "epoch": 5.542548941270476, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13873 + }, + { + "epoch": 5.542948461845785, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13874 + }, + { + "epoch": 5.5433479824210945, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13875 + }, + { + "epoch": 5.543747502996404, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 13876 + }, + { + "epoch": 5.544147023571714, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13877 + }, + { + "epoch": 5.544546544147024, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13878 + }, + { + "epoch": 5.544946064722334, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13879 + }, + { + "epoch": 5.5453455852976425, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13880 + }, + { + "epoch": 5.545745105872952, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13881 + }, + { + "epoch": 5.546144626448262, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13882 + }, + { + "epoch": 5.546544147023572, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13883 + }, + { + "epoch": 5.546943667598882, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13884 + }, + { + "epoch": 5.547343188174191, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13885 + }, + { + "epoch": 5.5477427087495, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13886 + }, + { + "epoch": 5.54814222932481, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 13887 + }, + { + "epoch": 5.54854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13888 + }, + { + "epoch": 5.54894127047543, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13889 + }, + { + "epoch": 5.549340791050739, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13890 + }, + { + "epoch": 5.549740311626048, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13891 + }, + { + "epoch": 5.550139832201358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13892 + }, + { + "epoch": 5.550539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13893 + }, + { + "epoch": 5.550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13894 + }, + { + "epoch": 5.551338393927288, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13895 + }, + { + "epoch": 5.551737914502597, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13896 + }, + { + "epoch": 5.552137435077906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13897 + }, + { + "epoch": 5.552536955653216, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13898 + }, + { + "epoch": 5.552936476228526, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13899 + }, + { + "epoch": 5.553335996803836, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13900 + }, + { + "epoch": 5.553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13901 + }, + { + "epoch": 5.554135037954454, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13902 + }, + { + "epoch": 5.554534558529764, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13903 + }, + { + "epoch": 5.554934079105074, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13904 + }, + { + "epoch": 5.555333599680384, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13905 + }, + { + "epoch": 5.555733120255693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13906 + }, + { + "epoch": 5.556132640831002, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13907 + }, + { + "epoch": 5.556532161406312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13908 + }, + { + "epoch": 5.556931681981622, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13909 + }, + { + "epoch": 5.557331202556932, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13910 + }, + { + "epoch": 5.5577307231322415, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13911 + }, + { + "epoch": 5.558130243707551, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13912 + }, + { + "epoch": 5.558529764282861, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13913 + }, + { + "epoch": 5.55892928485817, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13914 + }, + { + "epoch": 5.55932880543348, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13915 + }, + { + "epoch": 5.5597283260087895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13916 + }, + { + "epoch": 5.560127846584099, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13917 + }, + { + "epoch": 5.560527367159409, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13918 + }, + { + "epoch": 5.560926887734718, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13919 + }, + { + "epoch": 5.561326408310028, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13920 + }, + { + "epoch": 5.561725928885338, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13921 + }, + { + "epoch": 5.562125449460647, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13922 + }, + { + "epoch": 5.562524970035957, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13923 + }, + { + "epoch": 5.562924490611266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13924 + }, + { + "epoch": 5.563324011186576, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13925 + }, + { + "epoch": 5.563723531761886, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13926 + }, + { + "epoch": 5.564123052337195, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 13927 + }, + { + "epoch": 5.564522572912505, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13928 + }, + { + "epoch": 5.564922093487815, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13929 + }, + { + "epoch": 5.565321614063124, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13930 + }, + { + "epoch": 5.565721134638434, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13931 + }, + { + "epoch": 5.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13932 + }, + { + "epoch": 5.566520175789053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13933 + }, + { + "epoch": 5.566919696364363, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13934 + }, + { + "epoch": 5.567319216939673, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13935 + }, + { + "epoch": 5.567718737514982, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13936 + }, + { + "epoch": 5.5681182580902915, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13937 + }, + { + "epoch": 5.568517778665601, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13938 + }, + { + "epoch": 5.568917299240911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13939 + }, + { + "epoch": 5.569316819816221, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13940 + }, + { + "epoch": 5.56971634039153, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13941 + }, + { + "epoch": 5.5701158609668395, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13942 + }, + { + "epoch": 5.570515381542149, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13943 + }, + { + "epoch": 5.570914902117459, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13944 + }, + { + "epoch": 5.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13945 + }, + { + "epoch": 5.571713943268079, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13946 + }, + { + "epoch": 5.572113463843388, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13947 + }, + { + "epoch": 5.572512984418697, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13948 + }, + { + "epoch": 5.572912504994007, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13949 + }, + { + "epoch": 5.573312025569317, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13950 + }, + { + "epoch": 5.573711546144627, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13951 + }, + { + "epoch": 5.5741110667199365, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13952 + }, + { + "epoch": 5.574510587295245, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13953 + }, + { + "epoch": 5.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13954 + }, + { + "epoch": 5.575309628445865, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13955 + }, + { + "epoch": 5.575709149021175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13956 + }, + { + "epoch": 5.5761086695964845, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13957 + }, + { + "epoch": 5.5765081901717934, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13958 + }, + { + "epoch": 5.576907710747103, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 13959 + }, + { + "epoch": 5.577307231322413, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13960 + }, + { + "epoch": 5.577706751897723, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13961 + }, + { + "epoch": 5.578106272473033, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13962 + }, + { + "epoch": 5.578505793048342, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13963 + }, + { + "epoch": 5.578905313623651, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13964 + }, + { + "epoch": 5.579304834198961, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13965 + }, + { + "epoch": 5.579704354774271, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13966 + }, + { + "epoch": 5.580103875349581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13967 + }, + { + "epoch": 5.58050339592489, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13968 + }, + { + "epoch": 5.5809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 13969 + }, + { + "epoch": 5.581302437075509, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13970 + }, + { + "epoch": 5.581701957650819, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13971 + }, + { + "epoch": 5.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13972 + }, + { + "epoch": 5.5825009988014385, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13973 + }, + { + "epoch": 5.582900519376748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13974 + }, + { + "epoch": 5.583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13975 + }, + { + "epoch": 5.583699560527367, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13976 + }, + { + "epoch": 5.584099081102677, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13977 + }, + { + "epoch": 5.5844986016779865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13978 + }, + { + "epoch": 5.584898122253296, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13979 + }, + { + "epoch": 5.585297642828605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13980 + }, + { + "epoch": 5.585697163403915, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13981 + }, + { + "epoch": 5.586096683979225, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13982 + }, + { + "epoch": 5.5864962045545346, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13983 + }, + { + "epoch": 5.586895725129844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13984 + }, + { + "epoch": 5.587295245705154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13985 + }, + { + "epoch": 5.587694766280464, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13986 + }, + { + "epoch": 5.588094286855773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13987 + }, + { + "epoch": 5.588493807431083, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 13988 + }, + { + "epoch": 5.588893328006392, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13989 + }, + { + "epoch": 5.589292848581702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13990 + }, + { + "epoch": 5.589692369157012, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13991 + }, + { + "epoch": 5.590091889732321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13992 + }, + { + "epoch": 5.590491410307631, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13993 + }, + { + "epoch": 5.59089093088294, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 13994 + }, + { + "epoch": 5.59129045145825, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13995 + }, + { + "epoch": 5.59168997203356, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13996 + }, + { + "epoch": 5.592089492608869, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13997 + }, + { + "epoch": 5.592489013184179, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13998 + }, + { + "epoch": 5.5928885337594885, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13999 + }, + { + "epoch": 5.593288054334798, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14000 + }, + { + "epoch": 5.593687574910108, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14001 + }, + { + "epoch": 5.594087095485418, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14002 + }, + { + "epoch": 5.594486616060728, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 14003 + }, + { + "epoch": 5.5948861366360365, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14004 + }, + { + "epoch": 5.595285657211346, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14005 + }, + { + "epoch": 5.595685177786656, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14006 + }, + { + "epoch": 5.596084698361966, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14007 + }, + { + "epoch": 5.596484218937276, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14008 + }, + { + "epoch": 5.5968837395125846, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14009 + }, + { + "epoch": 5.597283260087894, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.2923, + "step": 14010 + }, + { + "epoch": 5.597682780663204, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14011 + }, + { + "epoch": 5.598082301238514, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14012 + }, + { + "epoch": 5.598481821813824, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14013 + }, + { + "epoch": 5.598881342389133, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14014 + }, + { + "epoch": 5.599280862964442, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14015 + }, + { + "epoch": 5.599680383539752, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14016 + }, + { + "epoch": 5.600079904115062, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 14017 + }, + { + "epoch": 5.600479424690372, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14018 + }, + { + "epoch": 5.6008789452656815, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14019 + }, + { + "epoch": 5.60127846584099, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14020 + }, + { + "epoch": 5.6016779864163, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14021 + }, + { + "epoch": 5.60207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14022 + }, + { + "epoch": 5.60247702756692, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14023 + }, + { + "epoch": 5.60287654814223, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 14024 + }, + { + "epoch": 5.603276068717539, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14025 + }, + { + "epoch": 5.603675589292848, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14026 + }, + { + "epoch": 5.604075109868158, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14027 + }, + { + "epoch": 5.604474630443468, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14028 + }, + { + "epoch": 5.604874151018778, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14029 + }, + { + "epoch": 5.605273671594087, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 14030 + }, + { + "epoch": 5.605673192169396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14031 + }, + { + "epoch": 5.606072712744706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14032 + }, + { + "epoch": 5.606472233320016, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14033 + }, + { + "epoch": 5.606871753895326, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14034 + }, + { + "epoch": 5.6072712744706354, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14035 + }, + { + "epoch": 5.607670795045945, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14036 + }, + { + "epoch": 5.608070315621254, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14037 + }, + { + "epoch": 5.608469836196564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14038 + }, + { + "epoch": 5.608869356771874, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14039 + }, + { + "epoch": 5.6092688773471835, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14040 + }, + { + "epoch": 5.609668397922493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14041 + }, + { + "epoch": 5.610067918497803, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14042 + }, + { + "epoch": 5.610467439073112, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14043 + }, + { + "epoch": 5.610866959648422, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14044 + }, + { + "epoch": 5.6112664802237315, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14045 + }, + { + "epoch": 5.611666000799041, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14046 + }, + { + "epoch": 5.612065521374351, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14047 + }, + { + "epoch": 5.61246504194966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14048 + }, + { + "epoch": 5.61286456252497, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14049 + }, + { + "epoch": 5.61326408310028, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14050 + }, + { + "epoch": 5.613663603675589, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14051 + }, + { + "epoch": 5.614063124250899, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14052 + }, + { + "epoch": 5.614462644826209, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14053 + }, + { + "epoch": 5.614862165401518, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14054 + }, + { + "epoch": 5.615261685976828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14055 + }, + { + "epoch": 5.615661206552137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14056 + }, + { + "epoch": 5.616060727127447, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14057 + }, + { + "epoch": 5.616460247702757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14058 + }, + { + "epoch": 5.616859768278067, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14059 + }, + { + "epoch": 5.617259288853376, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14060 + }, + { + "epoch": 5.6176588094286855, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14061 + }, + { + "epoch": 5.618058330003995, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14062 + }, + { + "epoch": 5.618457850579305, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14063 + }, + { + "epoch": 5.618857371154615, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14064 + }, + { + "epoch": 5.619256891729924, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14065 + }, + { + "epoch": 5.6196564123052335, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14066 + }, + { + "epoch": 5.620055932880543, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14067 + }, + { + "epoch": 5.620455453455853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14068 + }, + { + "epoch": 5.620854974031163, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14069 + }, + { + "epoch": 5.621254494606472, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14070 + }, + { + "epoch": 5.6216540151817815, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14071 + }, + { + "epoch": 5.622053535757091, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14072 + }, + { + "epoch": 5.622453056332401, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14073 + }, + { + "epoch": 5.622852576907711, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14074 + }, + { + "epoch": 5.623252097483021, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14075 + }, + { + "epoch": 5.6236516180583305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14076 + }, + { + "epoch": 5.624051138633639, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14077 + }, + { + "epoch": 5.624450659208949, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14078 + }, + { + "epoch": 5.624850179784259, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14079 + }, + { + "epoch": 5.625249700359569, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.2954, + "step": 14080 + }, + { + "epoch": 5.6256492209348785, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14081 + }, + { + "epoch": 5.626048741510187, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14082 + }, + { + "epoch": 5.626448262085497, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14083 + }, + { + "epoch": 5.626847782660807, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14084 + }, + { + "epoch": 5.627247303236117, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 14085 + }, + { + "epoch": 5.627646823811427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14086 + }, + { + "epoch": 5.6280463443867355, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14087 + }, + { + "epoch": 5.628445864962045, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14088 + }, + { + "epoch": 5.628845385537355, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14089 + }, + { + "epoch": 5.629244906112665, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14090 + }, + { + "epoch": 5.629644426687975, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14091 + }, + { + "epoch": 5.630043947263284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14092 + }, + { + "epoch": 5.630443467838594, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14093 + }, + { + "epoch": 5.630842988413903, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14094 + }, + { + "epoch": 5.631242508989213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14095 + }, + { + "epoch": 5.631642029564523, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14096 + }, + { + "epoch": 5.632041550139832, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14097 + }, + { + "epoch": 5.632441070715142, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14098 + }, + { + "epoch": 5.632840591290451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14099 + }, + { + "epoch": 5.633240111865761, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14100 + }, + { + "epoch": 5.633639632441071, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14101 + }, + { + "epoch": 5.6340391530163805, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14102 + }, + { + "epoch": 5.63443867359169, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14103 + }, + { + "epoch": 5.634838194166999, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14104 + }, + { + "epoch": 5.635237714742309, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14105 + }, + { + "epoch": 5.635637235317619, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14106 + }, + { + "epoch": 5.6360367558929285, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14107 + }, + { + "epoch": 5.636436276468238, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14108 + }, + { + "epoch": 5.636835797043548, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14109 + }, + { + "epoch": 5.637235317618857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 14110 + }, + { + "epoch": 5.637634838194167, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14111 + }, + { + "epoch": 5.638034358769477, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14112 + }, + { + "epoch": 5.638433879344786, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14113 + }, + { + "epoch": 5.638833399920096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14114 + }, + { + "epoch": 5.639232920495406, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14115 + }, + { + "epoch": 5.639632441070715, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14116 + }, + { + "epoch": 5.640031961646025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14117 + }, + { + "epoch": 5.640431482221334, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14118 + }, + { + "epoch": 5.640831002796644, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14119 + }, + { + "epoch": 5.641230523371954, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14120 + }, + { + "epoch": 5.641630043947263, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14121 + }, + { + "epoch": 5.642029564522573, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14122 + }, + { + "epoch": 5.642429085097882, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 14123 + }, + { + "epoch": 5.642828605673192, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14124 + }, + { + "epoch": 5.643228126248502, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14125 + }, + { + "epoch": 5.643627646823812, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14126 + }, + { + "epoch": 5.644027167399121, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14127 + }, + { + "epoch": 5.6444266879744305, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14128 + }, + { + "epoch": 5.64482620854974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14129 + }, + { + "epoch": 5.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14130 + }, + { + "epoch": 5.64562524970036, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14131 + }, + { + "epoch": 5.64602477027567, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14132 + }, + { + "epoch": 5.6464242908509785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14133 + }, + { + "epoch": 5.646823811426288, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14134 + }, + { + "epoch": 5.647223332001598, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14135 + }, + { + "epoch": 5.647622852576908, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14136 + }, + { + "epoch": 5.648022373152218, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14137 + }, + { + "epoch": 5.648421893727527, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14138 + }, + { + "epoch": 5.648821414302836, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14139 + }, + { + "epoch": 5.649220934878146, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14140 + }, + { + "epoch": 5.649620455453456, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14141 + }, + { + "epoch": 5.650019976028766, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14142 + }, + { + "epoch": 5.6504194966040755, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14143 + }, + { + "epoch": 5.650819017179384, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14144 + }, + { + "epoch": 5.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14145 + }, + { + "epoch": 5.651618058330004, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14146 + }, + { + "epoch": 5.652017578905314, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14147 + }, + { + "epoch": 5.6524170994806235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14148 + }, + { + "epoch": 5.652816620055933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14149 + }, + { + "epoch": 5.653216140631242, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 14150 + }, + { + "epoch": 5.653615661206552, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14151 + }, + { + "epoch": 5.654015181781862, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14152 + }, + { + "epoch": 5.654414702357172, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14153 + }, + { + "epoch": 5.654814222932481, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14154 + }, + { + "epoch": 5.65521374350779, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14155 + }, + { + "epoch": 5.6556132640831, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14156 + }, + { + "epoch": 5.65601278465841, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14157 + }, + { + "epoch": 5.65641230523372, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14158 + }, + { + "epoch": 5.656811825809029, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14159 + }, + { + "epoch": 5.657211346384338, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14160 + }, + { + "epoch": 5.657610866959648, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 14161 + }, + { + "epoch": 5.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14162 + }, + { + "epoch": 5.658409908110268, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14163 + }, + { + "epoch": 5.6588094286855775, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14164 + }, + { + "epoch": 5.659208949260887, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14165 + }, + { + "epoch": 5.659608469836197, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14166 + }, + { + "epoch": 5.660007990411506, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14167 + }, + { + "epoch": 5.660407510986816, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14168 + }, + { + "epoch": 5.6608070315621255, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14169 + }, + { + "epoch": 5.661206552137435, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14170 + }, + { + "epoch": 5.661606072712745, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14171 + }, + { + "epoch": 5.662005593288054, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14172 + }, + { + "epoch": 5.662405113863364, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14173 + }, + { + "epoch": 5.6628046344386735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14174 + }, + { + "epoch": 5.663204155013983, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14175 + }, + { + "epoch": 5.663603675589293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 14176 + }, + { + "epoch": 5.664003196164602, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 14177 + }, + { + "epoch": 5.664402716739912, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14178 + }, + { + "epoch": 5.664802237315222, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14179 + }, + { + "epoch": 5.665201757890531, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14180 + }, + { + "epoch": 5.665601278465841, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14181 + }, + { + "epoch": 5.666000799041151, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14182 + }, + { + "epoch": 5.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14183 + }, + { + "epoch": 5.66679984019177, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14184 + }, + { + "epoch": 5.667199360767079, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14185 + }, + { + "epoch": 5.667598881342389, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14186 + }, + { + "epoch": 5.667998401917699, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14187 + }, + { + "epoch": 5.668397922493009, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14188 + }, + { + "epoch": 5.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14189 + }, + { + "epoch": 5.6691969636436275, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14190 + }, + { + "epoch": 5.669596484218937, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14191 + }, + { + "epoch": 5.669996004794247, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14192 + }, + { + "epoch": 5.670395525369557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14193 + }, + { + "epoch": 5.670795045944866, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14194 + }, + { + "epoch": 5.6711945665201755, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14195 + }, + { + "epoch": 5.671594087095485, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14196 + }, + { + "epoch": 5.671993607670795, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14197 + }, + { + "epoch": 5.672393128246105, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14198 + }, + { + "epoch": 5.672792648821415, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14199 + }, + { + "epoch": 5.6731921693967235, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14200 + }, + { + "epoch": 5.673591689972033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14201 + }, + { + "epoch": 5.673991210547343, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14202 + }, + { + "epoch": 5.674390731122653, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14203 + }, + { + "epoch": 5.674790251697963, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14204 + }, + { + "epoch": 5.6751897722732725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 14205 + }, + { + "epoch": 5.675589292848581, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14206 + }, + { + "epoch": 5.675988813423891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14207 + }, + { + "epoch": 5.676388333999201, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14208 + }, + { + "epoch": 5.676787854574511, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14209 + }, + { + "epoch": 5.6771873751498205, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14210 + }, + { + "epoch": 5.677586895725129, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14211 + }, + { + "epoch": 5.677986416300439, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14212 + }, + { + "epoch": 5.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14213 + }, + { + "epoch": 5.678785457451059, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14214 + }, + { + "epoch": 5.679184978026369, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14215 + }, + { + "epoch": 5.679584498601678, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14216 + }, + { + "epoch": 5.679984019176987, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14217 + }, + { + "epoch": 5.680383539752297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14218 + }, + { + "epoch": 5.680783060327607, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14219 + }, + { + "epoch": 5.681182580902917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14220 + }, + { + "epoch": 5.681582101478226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14221 + }, + { + "epoch": 5.681981622053536, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14222 + }, + { + "epoch": 5.682381142628845, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14223 + }, + { + "epoch": 5.682780663204155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14224 + }, + { + "epoch": 5.683180183779465, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14225 + }, + { + "epoch": 5.683579704354774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 14226 + }, + { + "epoch": 5.683979224930084, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14227 + }, + { + "epoch": 5.684378745505393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14228 + }, + { + "epoch": 5.684778266080703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14229 + }, + { + "epoch": 5.685177786656013, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14230 + }, + { + "epoch": 5.6855773072313225, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14231 + }, + { + "epoch": 5.685976827806632, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14232 + }, + { + "epoch": 5.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14233 + }, + { + "epoch": 5.686775868957251, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14234 + }, + { + "epoch": 5.687175389532561, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14235 + }, + { + "epoch": 5.6875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14236 + }, + { + "epoch": 5.68797443068318, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14237 + }, + { + "epoch": 5.68837395125849, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14238 + }, + { + "epoch": 5.6887734718338, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14239 + }, + { + "epoch": 5.689172992409109, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14240 + }, + { + "epoch": 5.689572512984419, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14241 + }, + { + "epoch": 5.689972033559728, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 14242 + }, + { + "epoch": 5.690371554135038, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14243 + }, + { + "epoch": 5.690771074710348, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14244 + }, + { + "epoch": 5.691170595285657, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14245 + }, + { + "epoch": 5.691570115860967, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14246 + }, + { + "epoch": 5.691969636436276, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14247 + }, + { + "epoch": 5.692369157011586, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14248 + }, + { + "epoch": 5.692768677586896, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14249 + }, + { + "epoch": 5.693168198162205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14250 + }, + { + "epoch": 5.693567718737515, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 14251 + }, + { + "epoch": 5.6939672393128244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14252 + }, + { + "epoch": 5.694366759888134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14253 + }, + { + "epoch": 5.694766280463444, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14254 + }, + { + "epoch": 5.695165801038754, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14255 + }, + { + "epoch": 5.695565321614064, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14256 + }, + { + "epoch": 5.6959648421893725, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14257 + }, + { + "epoch": 5.696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14258 + }, + { + "epoch": 5.696763883339992, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14259 + }, + { + "epoch": 5.697163403915302, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14260 + }, + { + "epoch": 5.697562924490612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14261 + }, + { + "epoch": 5.6979624450659205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14262 + }, + { + "epoch": 5.69836196564123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14263 + }, + { + "epoch": 5.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14264 + }, + { + "epoch": 5.69916100679185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14265 + }, + { + "epoch": 5.69956052736716, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14266 + }, + { + "epoch": 5.699960047942469, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14267 + }, + { + "epoch": 5.700359568517778, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 14268 + }, + { + "epoch": 5.700759089093088, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14269 + }, + { + "epoch": 5.701158609668398, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14270 + }, + { + "epoch": 5.701558130243708, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14271 + }, + { + "epoch": 5.7019576508190175, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 14272 + }, + { + "epoch": 5.702357171394327, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14273 + }, + { + "epoch": 5.702756691969636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14274 + }, + { + "epoch": 5.703156212544946, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14275 + }, + { + "epoch": 5.703555733120256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14276 + }, + { + "epoch": 5.7039552536955656, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14277 + }, + { + "epoch": 5.704354774270875, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14278 + }, + { + "epoch": 5.704754294846184, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14279 + }, + { + "epoch": 5.705153815421494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14280 + }, + { + "epoch": 5.705553335996804, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14281 + }, + { + "epoch": 5.705952856572114, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14282 + }, + { + "epoch": 5.706352377147423, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14283 + }, + { + "epoch": 5.706751897722732, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14284 + }, + { + "epoch": 5.707151418298042, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14285 + }, + { + "epoch": 5.707550938873352, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14286 + }, + { + "epoch": 5.707950459448662, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14287 + }, + { + "epoch": 5.708349980023971, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14288 + }, + { + "epoch": 5.708749500599281, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14289 + }, + { + "epoch": 5.70914902117459, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 14290 + }, + { + "epoch": 5.7095485417499, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14291 + }, + { + "epoch": 5.70994806232521, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14292 + }, + { + "epoch": 5.7103475829005195, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14293 + }, + { + "epoch": 5.710747103475829, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14294 + }, + { + "epoch": 5.711146624051139, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14295 + }, + { + "epoch": 5.711546144626448, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14296 + }, + { + "epoch": 5.711945665201758, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 14297 + }, + { + "epoch": 5.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14298 + }, + { + "epoch": 5.712744706352377, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 14299 + }, + { + "epoch": 5.713144226927687, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14300 + }, + { + "epoch": 5.713543747502996, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14301 + }, + { + "epoch": 5.713943268078306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14302 + }, + { + "epoch": 5.714342788653616, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14303 + }, + { + "epoch": 5.714742309228925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14304 + }, + { + "epoch": 5.715141829804235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14305 + }, + { + "epoch": 5.715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14306 + }, + { + "epoch": 5.715940870954854, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14307 + }, + { + "epoch": 5.716340391530164, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14308 + }, + { + "epoch": 5.716739912105473, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14309 + }, + { + "epoch": 5.717139432680783, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14310 + }, + { + "epoch": 5.717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14311 + }, + { + "epoch": 5.717938473831403, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14312 + }, + { + "epoch": 5.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14313 + }, + { + "epoch": 5.718737514982021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14314 + }, + { + "epoch": 5.719137035557331, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14315 + }, + { + "epoch": 5.719536556132641, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14316 + }, + { + "epoch": 5.719936076707951, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 14317 + }, + { + "epoch": 5.72033559728326, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14318 + }, + { + "epoch": 5.7207351178585695, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14319 + }, + { + "epoch": 5.721134638433879, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14320 + }, + { + "epoch": 5.721534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14321 + }, + { + "epoch": 5.721933679584499, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14322 + }, + { + "epoch": 5.722333200159808, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14323 + }, + { + "epoch": 5.7227327207351175, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 14324 + }, + { + "epoch": 5.723132241310427, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14325 + }, + { + "epoch": 5.723531761885737, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14326 + }, + { + "epoch": 5.723931282461047, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14327 + }, + { + "epoch": 5.724330803036357, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14328 + }, + { + "epoch": 5.7247303236116664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14329 + }, + { + "epoch": 5.725129844186975, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14330 + }, + { + "epoch": 5.725529364762285, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14331 + }, + { + "epoch": 5.725928885337595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.301, + "step": 14332 + }, + { + "epoch": 5.726328405912905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14333 + }, + { + "epoch": 5.7267279264882145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14334 + }, + { + "epoch": 5.727127447063523, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14335 + }, + { + "epoch": 5.727526967638833, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14336 + }, + { + "epoch": 5.727926488214143, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14337 + }, + { + "epoch": 5.728326008789453, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14338 + }, + { + "epoch": 5.7287255293647625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14339 + }, + { + "epoch": 5.729125049940071, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14340 + }, + { + "epoch": 5.729524570515381, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14341 + }, + { + "epoch": 5.729924091090691, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14342 + }, + { + "epoch": 5.730323611666001, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14343 + }, + { + "epoch": 5.730723132241311, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14344 + }, + { + "epoch": 5.73112265281662, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14345 + }, + { + "epoch": 5.73152217339193, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14346 + }, + { + "epoch": 5.731921693967239, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14347 + }, + { + "epoch": 5.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14348 + }, + { + "epoch": 5.732720735117859, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14349 + }, + { + "epoch": 5.733120255693168, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14350 + }, + { + "epoch": 5.733519776268478, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14351 + }, + { + "epoch": 5.733919296843787, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 14352 + }, + { + "epoch": 5.734318817419097, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14353 + }, + { + "epoch": 5.734718337994407, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14354 + }, + { + "epoch": 5.7351178585697165, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14355 + }, + { + "epoch": 5.735517379145026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14356 + }, + { + "epoch": 5.735916899720335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14357 + }, + { + "epoch": 5.736316420295645, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14358 + }, + { + "epoch": 5.736715940870955, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14359 + }, + { + "epoch": 5.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14360 + }, + { + "epoch": 5.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14361 + }, + { + "epoch": 5.737914502596884, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14362 + }, + { + "epoch": 5.738314023172194, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14363 + }, + { + "epoch": 5.738713543747503, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14364 + }, + { + "epoch": 5.7391130643228125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14365 + }, + { + "epoch": 5.739512584898122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14366 + }, + { + "epoch": 5.739912105473432, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 14367 + }, + { + "epoch": 5.740311626048742, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14368 + }, + { + "epoch": 5.740711146624051, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14369 + }, + { + "epoch": 5.741110667199361, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14370 + }, + { + "epoch": 5.74151018777467, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14371 + }, + { + "epoch": 5.74190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14372 + }, + { + "epoch": 5.74230922892529, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14373 + }, + { + "epoch": 5.742708749500599, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14374 + }, + { + "epoch": 5.743108270075909, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 14375 + }, + { + "epoch": 5.743507790651218, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14376 + }, + { + "epoch": 5.743907311226528, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14377 + }, + { + "epoch": 5.744306831801838, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14378 + }, + { + "epoch": 5.744706352377148, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14379 + }, + { + "epoch": 5.745105872952457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14380 + }, + { + "epoch": 5.7455053935277665, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14381 + }, + { + "epoch": 5.745904914103076, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14382 + }, + { + "epoch": 5.746304434678386, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 14383 + }, + { + "epoch": 5.746703955253696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14384 + }, + { + "epoch": 5.747103475829006, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14385 + }, + { + "epoch": 5.7475029964043145, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14386 + }, + { + "epoch": 5.747902516979624, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14387 + }, + { + "epoch": 5.748302037554934, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14388 + }, + { + "epoch": 5.748701558130244, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14389 + }, + { + "epoch": 5.749101078705554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14390 + }, + { + "epoch": 5.7495005992808625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14391 + }, + { + "epoch": 5.749900119856172, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14392 + }, + { + "epoch": 5.750299640431482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14393 + }, + { + "epoch": 5.750699161006792, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14394 + }, + { + "epoch": 5.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14395 + }, + { + "epoch": 5.7514982021574115, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14396 + }, + { + "epoch": 5.75189772273272, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14397 + }, + { + "epoch": 5.75229724330803, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14398 + }, + { + "epoch": 5.75269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14399 + }, + { + "epoch": 5.75309628445865, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14400 + }, + { + "epoch": 5.7534958050339595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14401 + }, + { + "epoch": 5.753895325609269, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14402 + }, + { + "epoch": 5.754294846184578, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14403 + }, + { + "epoch": 5.754694366759888, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14404 + }, + { + "epoch": 5.755093887335198, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14405 + }, + { + "epoch": 5.755493407910508, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14406 + }, + { + "epoch": 5.755892928485817, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14407 + }, + { + "epoch": 5.756292449061126, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14408 + }, + { + "epoch": 5.756691969636436, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14409 + }, + { + "epoch": 5.757091490211746, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14410 + }, + { + "epoch": 5.757491010787056, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14411 + }, + { + "epoch": 5.757890531362365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14412 + }, + { + "epoch": 5.758290051937674, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14413 + }, + { + "epoch": 5.758689572512984, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14414 + }, + { + "epoch": 5.759089093088294, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14415 + }, + { + "epoch": 5.759488613663604, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14416 + }, + { + "epoch": 5.759888134238913, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14417 + }, + { + "epoch": 5.760287654814223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14418 + }, + { + "epoch": 5.760687175389533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14419 + }, + { + "epoch": 5.761086695964842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 14420 + }, + { + "epoch": 5.761486216540152, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14421 + }, + { + "epoch": 5.7618857371154615, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14422 + }, + { + "epoch": 5.762285257690771, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14423 + }, + { + "epoch": 5.762684778266081, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14424 + }, + { + "epoch": 5.76308429884139, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14425 + }, + { + "epoch": 5.7634838194167, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14426 + }, + { + "epoch": 5.7638833399920095, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14427 + }, + { + "epoch": 5.764282860567319, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14428 + }, + { + "epoch": 5.764682381142629, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14429 + }, + { + "epoch": 5.765081901717938, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14430 + }, + { + "epoch": 5.765481422293248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14431 + }, + { + "epoch": 5.765880942868558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14432 + }, + { + "epoch": 5.766280463443867, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14433 + }, + { + "epoch": 5.766679984019177, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14434 + }, + { + "epoch": 5.767079504594487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14435 + }, + { + "epoch": 5.767479025169797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14436 + }, + { + "epoch": 5.767878545745106, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14437 + }, + { + "epoch": 5.768278066320415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14438 + }, + { + "epoch": 5.768677586895725, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14439 + }, + { + "epoch": 5.769077107471035, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14440 + }, + { + "epoch": 5.769476628046345, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14441 + }, + { + "epoch": 5.769876148621654, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14442 + }, + { + "epoch": 5.770275669196963, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14443 + }, + { + "epoch": 5.770675189772273, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14444 + }, + { + "epoch": 5.771074710347583, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14445 + }, + { + "epoch": 5.771474230922893, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14446 + }, + { + "epoch": 5.771873751498202, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14447 + }, + { + "epoch": 5.7722732720735115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14448 + }, + { + "epoch": 5.772672792648821, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14449 + }, + { + "epoch": 5.773072313224131, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14450 + }, + { + "epoch": 5.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14451 + }, + { + "epoch": 5.773871354374751, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14452 + }, + { + "epoch": 5.77427087495006, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14453 + }, + { + "epoch": 5.774670395525369, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14454 + }, + { + "epoch": 5.775069916100679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14455 + }, + { + "epoch": 5.775469436675989, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14456 + }, + { + "epoch": 5.775868957251299, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14457 + }, + { + "epoch": 5.7762684778266085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14458 + }, + { + "epoch": 5.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14459 + }, + { + "epoch": 5.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14460 + }, + { + "epoch": 5.777467039552537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 14461 + }, + { + "epoch": 5.777866560127847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14462 + }, + { + "epoch": 5.7782660807031565, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14463 + }, + { + "epoch": 5.778665601278465, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14464 + }, + { + "epoch": 5.779065121853775, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 14465 + }, + { + "epoch": 5.779464642429085, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14466 + }, + { + "epoch": 5.779864163004395, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14467 + }, + { + "epoch": 5.7802636835797045, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14468 + }, + { + "epoch": 5.780663204155014, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14469 + }, + { + "epoch": 5.781062724730323, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14470 + }, + { + "epoch": 5.781462245305633, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14471 + }, + { + "epoch": 5.781861765880943, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14472 + }, + { + "epoch": 5.782261286456253, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14473 + }, + { + "epoch": 5.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14474 + }, + { + "epoch": 5.783060327606872, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14475 + }, + { + "epoch": 5.783459848182181, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14476 + }, + { + "epoch": 5.783859368757491, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14477 + }, + { + "epoch": 5.784258889332801, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14478 + }, + { + "epoch": 5.78465840990811, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14479 + }, + { + "epoch": 5.78505793048342, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14480 + }, + { + "epoch": 5.785457451058729, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14481 + }, + { + "epoch": 5.785856971634039, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 14482 + }, + { + "epoch": 5.786256492209349, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14483 + }, + { + "epoch": 5.7866560127846585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14484 + }, + { + "epoch": 5.787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 14485 + }, + { + "epoch": 5.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 14486 + }, + { + "epoch": 5.787854574510587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14487 + }, + { + "epoch": 5.788254095085897, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14488 + }, + { + "epoch": 5.7886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14489 + }, + { + "epoch": 5.789053136236516, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14490 + }, + { + "epoch": 5.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14491 + }, + { + "epoch": 5.789852177387136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14492 + }, + { + "epoch": 5.790251697962445, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14493 + }, + { + "epoch": 5.7906512185377546, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14494 + }, + { + "epoch": 5.791050739113064, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14495 + }, + { + "epoch": 5.791450259688374, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 14496 + }, + { + "epoch": 5.791849780263684, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14497 + }, + { + "epoch": 5.792249300838993, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14498 + }, + { + "epoch": 5.792648821414303, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14499 + }, + { + "epoch": 5.793048341989612, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 14500 + }, + { + "epoch": 5.793447862564922, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14501 + }, + { + "epoch": 5.793847383140232, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14502 + }, + { + "epoch": 5.794246903715541, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14503 + }, + { + "epoch": 5.794646424290851, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.2985, + "step": 14504 + }, + { + "epoch": 5.79504594486616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14505 + }, + { + "epoch": 5.79544546544147, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14506 + }, + { + "epoch": 5.79584498601678, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14507 + }, + { + "epoch": 5.79624450659209, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14508 + }, + { + "epoch": 5.7966440271674, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14509 + }, + { + "epoch": 5.7970435477427085, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14510 + }, + { + "epoch": 5.797443068318018, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14511 + }, + { + "epoch": 5.797842588893328, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14512 + }, + { + "epoch": 5.798242109468638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14513 + }, + { + "epoch": 5.798641630043948, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14514 + }, + { + "epoch": 5.7990411506192565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14515 + }, + { + "epoch": 5.799440671194566, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14516 + }, + { + "epoch": 5.799840191769876, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14517 + }, + { + "epoch": 5.800239712345186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14518 + }, + { + "epoch": 5.800639232920496, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14519 + }, + { + "epoch": 5.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14520 + }, + { + "epoch": 5.801438274071114, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14521 + }, + { + "epoch": 5.801837794646424, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14522 + }, + { + "epoch": 5.802237315221734, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14523 + }, + { + "epoch": 5.802636835797044, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14524 + }, + { + "epoch": 5.8030363563723535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14525 + }, + { + "epoch": 5.803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14526 + }, + { + "epoch": 5.803835397522972, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14527 + }, + { + "epoch": 5.804234918098282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14528 + }, + { + "epoch": 5.804634438673592, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14529 + }, + { + "epoch": 5.8050339592489015, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14530 + }, + { + "epoch": 5.805433479824211, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14531 + }, + { + "epoch": 5.80583300039952, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 14532 + }, + { + "epoch": 5.80623252097483, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14533 + }, + { + "epoch": 5.80663204155014, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14534 + }, + { + "epoch": 5.80703156212545, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14535 + }, + { + "epoch": 5.807431082700759, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 14536 + }, + { + "epoch": 5.807830603276068, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14537 + }, + { + "epoch": 5.808230123851378, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14538 + }, + { + "epoch": 5.808629644426688, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14539 + }, + { + "epoch": 5.809029165001998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14540 + }, + { + "epoch": 5.809428685577307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14541 + }, + { + "epoch": 5.809828206152617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14542 + }, + { + "epoch": 5.810227726727927, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14543 + }, + { + "epoch": 5.810627247303236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14544 + }, + { + "epoch": 5.811026767878546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14545 + }, + { + "epoch": 5.8114262884538554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14546 + }, + { + "epoch": 5.811825809029165, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14547 + }, + { + "epoch": 5.812225329604475, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14548 + }, + { + "epoch": 5.812624850179784, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14549 + }, + { + "epoch": 5.813024370755094, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14550 + }, + { + "epoch": 5.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 14551 + }, + { + "epoch": 5.813823411905713, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14552 + }, + { + "epoch": 5.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14553 + }, + { + "epoch": 5.814622453056332, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14554 + }, + { + "epoch": 5.815021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14555 + }, + { + "epoch": 5.8154214942069515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14556 + }, + { + "epoch": 5.815821014782261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14557 + }, + { + "epoch": 5.816220535357571, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14558 + }, + { + "epoch": 5.816620055932881, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14559 + }, + { + "epoch": 5.81701957650819, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14560 + }, + { + "epoch": 5.8174190970835, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14561 + }, + { + "epoch": 5.817818617658809, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14562 + }, + { + "epoch": 5.818218138234119, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14563 + }, + { + "epoch": 5.818617658809429, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 14564 + }, + { + "epoch": 5.819017179384739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14565 + }, + { + "epoch": 5.819416699960048, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14566 + }, + { + "epoch": 5.819816220535357, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14567 + }, + { + "epoch": 5.820215741110667, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14568 + }, + { + "epoch": 5.820615261685977, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14569 + }, + { + "epoch": 5.821014782261287, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14570 + }, + { + "epoch": 5.821414302836596, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14571 + }, + { + "epoch": 5.8218138234119055, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14572 + }, + { + "epoch": 5.822213343987215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14573 + }, + { + "epoch": 5.822612864562525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14574 + }, + { + "epoch": 5.823012385137835, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14575 + }, + { + "epoch": 5.823411905713145, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14576 + }, + { + "epoch": 5.8238114262884535, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14577 + }, + { + "epoch": 5.824210946863763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14578 + }, + { + "epoch": 5.824610467439073, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14579 + }, + { + "epoch": 5.825009988014383, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14580 + }, + { + "epoch": 5.825409508589693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14581 + }, + { + "epoch": 5.825809029165002, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14582 + }, + { + "epoch": 5.826208549740311, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14583 + }, + { + "epoch": 5.826608070315621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14584 + }, + { + "epoch": 5.827007590890931, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14585 + }, + { + "epoch": 5.827407111466241, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14586 + }, + { + "epoch": 5.8278066320415505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14587 + }, + { + "epoch": 5.828206152616859, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 14588 + }, + { + "epoch": 5.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14589 + }, + { + "epoch": 5.829005193767479, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14590 + }, + { + "epoch": 5.829404714342789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14591 + }, + { + "epoch": 5.8298042349180985, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14592 + }, + { + "epoch": 5.830203755493407, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14593 + }, + { + "epoch": 5.830603276068717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14594 + }, + { + "epoch": 5.831002796644027, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14595 + }, + { + "epoch": 5.831402317219337, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14596 + }, + { + "epoch": 5.831801837794647, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 14597 + }, + { + "epoch": 5.832201358369956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14598 + }, + { + "epoch": 5.832600878945266, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14599 + }, + { + "epoch": 5.833000399520575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 14600 + }, + { + "epoch": 5.833399920095885, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14601 + }, + { + "epoch": 5.833799440671195, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14602 + }, + { + "epoch": 5.834198961246504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14603 + }, + { + "epoch": 5.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14604 + }, + { + "epoch": 5.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14605 + }, + { + "epoch": 5.835397522972433, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 14606 + }, + { + "epoch": 5.835797043547743, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14607 + }, + { + "epoch": 5.836196564123052, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14608 + }, + { + "epoch": 5.836596084698362, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14609 + }, + { + "epoch": 5.836995605273671, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14610 + }, + { + "epoch": 5.837395125848981, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14611 + }, + { + "epoch": 5.837794646424291, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14612 + }, + { + "epoch": 5.8381941669996005, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14613 + }, + { + "epoch": 5.83859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14614 + }, + { + "epoch": 5.83899320815022, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14615 + }, + { + "epoch": 5.83939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14616 + }, + { + "epoch": 5.839792249300839, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14617 + }, + { + "epoch": 5.8401917698761485, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 14618 + }, + { + "epoch": 5.840591290451458, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14619 + }, + { + "epoch": 5.840990811026768, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14620 + }, + { + "epoch": 5.841390331602078, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14621 + }, + { + "epoch": 5.841789852177387, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14622 + }, + { + "epoch": 5.842189372752697, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14623 + }, + { + "epoch": 5.842588893328006, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14624 + }, + { + "epoch": 5.842988413903316, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14625 + }, + { + "epoch": 5.843387934478626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14626 + }, + { + "epoch": 5.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14627 + }, + { + "epoch": 5.844186975629245, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14628 + }, + { + "epoch": 5.844586496204554, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14629 + }, + { + "epoch": 5.844986016779864, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14630 + }, + { + "epoch": 5.845385537355174, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14631 + }, + { + "epoch": 5.845785057930484, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14632 + }, + { + "epoch": 5.8461845785057935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14633 + }, + { + "epoch": 5.846584099081102, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14634 + }, + { + "epoch": 5.846983619656412, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14635 + }, + { + "epoch": 5.847383140231722, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14636 + }, + { + "epoch": 5.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14637 + }, + { + "epoch": 5.848182181382342, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14638 + }, + { + "epoch": 5.8485817019576505, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14639 + }, + { + "epoch": 5.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14640 + }, + { + "epoch": 5.84938074310827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14641 + }, + { + "epoch": 5.84978026368358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 14642 + }, + { + "epoch": 5.85017978425889, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14643 + }, + { + "epoch": 5.8505793048341985, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14644 + }, + { + "epoch": 5.850978825409508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14645 + }, + { + "epoch": 5.851378345984818, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14646 + }, + { + "epoch": 5.851777866560128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14647 + }, + { + "epoch": 5.852177387135438, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14648 + }, + { + "epoch": 5.8525769077107475, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 14649 + }, + { + "epoch": 5.852976428286056, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14650 + }, + { + "epoch": 5.853375948861366, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 14651 + }, + { + "epoch": 5.853775469436676, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14652 + }, + { + "epoch": 5.854174990011986, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14653 + }, + { + "epoch": 5.8545745105872955, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14654 + }, + { + "epoch": 5.854974031162605, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14655 + }, + { + "epoch": 5.855373551737914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14656 + }, + { + "epoch": 5.855773072313224, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 14657 + }, + { + "epoch": 5.856172592888534, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14658 + }, + { + "epoch": 5.8565721134638435, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14659 + }, + { + "epoch": 5.856971634039153, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14660 + }, + { + "epoch": 5.857371154614462, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14661 + }, + { + "epoch": 5.857770675189772, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14662 + }, + { + "epoch": 5.858170195765082, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14663 + }, + { + "epoch": 5.858569716340392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14664 + }, + { + "epoch": 5.858969236915701, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14665 + }, + { + "epoch": 5.859368757491011, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14666 + }, + { + "epoch": 5.85976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14667 + }, + { + "epoch": 5.86016779864163, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14668 + }, + { + "epoch": 5.86056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14669 + }, + { + "epoch": 5.860966839792249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14670 + }, + { + "epoch": 5.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14671 + }, + { + "epoch": 5.861765880942869, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14672 + }, + { + "epoch": 5.862165401518178, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14673 + }, + { + "epoch": 5.862564922093488, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14674 + }, + { + "epoch": 5.8629644426687975, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14675 + }, + { + "epoch": 5.863363963244107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14676 + }, + { + "epoch": 5.863763483819417, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 14677 + }, + { + "epoch": 5.864163004394726, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14678 + }, + { + "epoch": 5.864562524970036, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14679 + }, + { + "epoch": 5.8649620455453455, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 14680 + }, + { + "epoch": 5.865361566120655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14681 + }, + { + "epoch": 5.865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14682 + }, + { + "epoch": 5.866160607271274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14683 + }, + { + "epoch": 5.866560127846584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14684 + }, + { + "epoch": 5.8669596484218935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14685 + }, + { + "epoch": 5.867359168997203, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14686 + }, + { + "epoch": 5.867758689572513, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14687 + }, + { + "epoch": 5.868158210147823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14688 + }, + { + "epoch": 5.868557730723133, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14689 + }, + { + "epoch": 5.868957251298442, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 14690 + }, + { + "epoch": 5.869356771873751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14691 + }, + { + "epoch": 5.869756292449061, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14692 + }, + { + "epoch": 5.870155813024371, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14693 + }, + { + "epoch": 5.870555333599681, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14694 + }, + { + "epoch": 5.87095485417499, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14695 + }, + { + "epoch": 5.871354374750299, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14696 + }, + { + "epoch": 5.871753895325609, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14697 + }, + { + "epoch": 5.872153415900919, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14698 + }, + { + "epoch": 5.872552936476229, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14699 + }, + { + "epoch": 5.872952457051538, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14700 + }, + { + "epoch": 5.8733519776268475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14701 + }, + { + "epoch": 5.873751498202157, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 14702 + }, + { + "epoch": 5.874151018777467, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 14703 + }, + { + "epoch": 5.874550539352777, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14704 + }, + { + "epoch": 5.874950059928087, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14705 + }, + { + "epoch": 5.875349580503396, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14706 + }, + { + "epoch": 5.875749101078705, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14707 + }, + { + "epoch": 5.876148621654015, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14708 + }, + { + "epoch": 5.876548142229325, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 14709 + }, + { + "epoch": 5.876947662804635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 14710 + }, + { + "epoch": 5.877347183379944, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14711 + }, + { + "epoch": 5.877746703955253, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14712 + }, + { + "epoch": 5.878146224530563, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14713 + }, + { + "epoch": 5.878545745105873, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14714 + }, + { + "epoch": 5.878945265681183, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14715 + }, + { + "epoch": 5.8793447862564925, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14716 + }, + { + "epoch": 5.879744306831801, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14717 + }, + { + "epoch": 5.880143827407111, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14718 + }, + { + "epoch": 5.880543347982421, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14719 + }, + { + "epoch": 5.880942868557731, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14720 + }, + { + "epoch": 5.8813423891330405, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14721 + }, + { + "epoch": 5.88174190970835, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14722 + }, + { + "epoch": 5.88214143028366, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 14723 + }, + { + "epoch": 5.882540950858969, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14724 + }, + { + "epoch": 5.882940471434279, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14725 + }, + { + "epoch": 5.883339992009589, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 14726 + }, + { + "epoch": 5.883739512584898, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14727 + }, + { + "epoch": 5.884139033160208, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14728 + }, + { + "epoch": 5.884538553735517, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14729 + }, + { + "epoch": 5.884938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14730 + }, + { + "epoch": 5.885337594886137, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14731 + }, + { + "epoch": 5.885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14732 + }, + { + "epoch": 5.886136636036756, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14733 + }, + { + "epoch": 5.886536156612065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 14734 + }, + { + "epoch": 5.886935677187375, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14735 + }, + { + "epoch": 5.887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14736 + }, + { + "epoch": 5.887734718337994, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14737 + }, + { + "epoch": 5.888134238913304, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14738 + }, + { + "epoch": 5.888533759488614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 14739 + }, + { + "epoch": 5.888933280063923, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14740 + }, + { + "epoch": 5.889332800639233, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 14741 + }, + { + "epoch": 5.8897323212145425, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14742 + }, + { + "epoch": 5.890131841789852, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 14743 + }, + { + "epoch": 5.890531362365162, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14744 + }, + { + "epoch": 5.890930882940472, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14745 + }, + { + "epoch": 5.891330403515781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14746 + }, + { + "epoch": 5.8917299240910905, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14747 + }, + { + "epoch": 5.8921294446664, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14748 + }, + { + "epoch": 5.89252896524171, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 14749 + }, + { + "epoch": 5.89292848581702, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14750 + }, + { + "epoch": 5.893328006392329, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14751 + }, + { + "epoch": 5.893727526967639, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14752 + }, + { + "epoch": 5.894127047542948, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14753 + }, + { + "epoch": 5.894526568118258, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14754 + }, + { + "epoch": 5.894926088693568, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14755 + }, + { + "epoch": 5.895325609268878, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14756 + }, + { + "epoch": 5.895725129844187, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14757 + }, + { + "epoch": 5.896124650419496, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 14758 + }, + { + "epoch": 5.896524170994806, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14759 + }, + { + "epoch": 5.896923691570116, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14760 + }, + { + "epoch": 5.897323212145426, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14761 + }, + { + "epoch": 5.8977227327207356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14762 + }, + { + "epoch": 5.8981222532960444, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14763 + }, + { + "epoch": 5.898521773871354, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14764 + }, + { + "epoch": 5.898921294446664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14765 + }, + { + "epoch": 5.899320815021974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14766 + }, + { + "epoch": 5.899720335597284, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14767 + }, + { + "epoch": 5.9001198561725925, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14768 + }, + { + "epoch": 5.900519376747902, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14769 + }, + { + "epoch": 5.900918897323212, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14770 + }, + { + "epoch": 5.901318417898522, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 14771 + }, + { + "epoch": 5.901717938473832, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14772 + }, + { + "epoch": 5.9021174590491405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14773 + }, + { + "epoch": 5.90251697962445, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14774 + }, + { + "epoch": 5.90291650019976, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14775 + }, + { + "epoch": 5.90331602077507, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 14776 + }, + { + "epoch": 5.90371554135038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14777 + }, + { + "epoch": 5.9041150619256895, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14778 + }, + { + "epoch": 5.904514582500999, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14779 + }, + { + "epoch": 5.904914103076308, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14780 + }, + { + "epoch": 5.905313623651618, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14781 + }, + { + "epoch": 5.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 14782 + }, + { + "epoch": 5.9061126648022375, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14783 + }, + { + "epoch": 5.906512185377547, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14784 + }, + { + "epoch": 5.906911705952856, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14785 + }, + { + "epoch": 5.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14786 + }, + { + "epoch": 5.907710747103476, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14787 + }, + { + "epoch": 5.9081102676787856, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14788 + }, + { + "epoch": 5.908509788254095, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14789 + }, + { + "epoch": 5.908909308829404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14790 + }, + { + "epoch": 5.909308829404714, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14791 + }, + { + "epoch": 5.909708349980024, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14792 + }, + { + "epoch": 5.910107870555334, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14793 + }, + { + "epoch": 5.910507391130643, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14794 + }, + { + "epoch": 5.910906911705953, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14795 + }, + { + "epoch": 5.911306432281263, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 14796 + }, + { + "epoch": 5.911705952856572, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14797 + }, + { + "epoch": 5.912105473431882, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14798 + }, + { + "epoch": 5.912504994007191, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14799 + }, + { + "epoch": 5.912904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 14800 + }, + { + "epoch": 5.913304035157811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14801 + }, + { + "epoch": 5.91370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14802 + }, + { + "epoch": 5.91410307630843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14803 + }, + { + "epoch": 5.9145025968837395, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14804 + }, + { + "epoch": 5.914902117459049, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 14805 + }, + { + "epoch": 5.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14806 + }, + { + "epoch": 5.915701158609668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14807 + }, + { + "epoch": 5.916100679184978, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14808 + }, + { + "epoch": 5.9165001997602875, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14809 + }, + { + "epoch": 5.916899720335597, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14810 + }, + { + "epoch": 5.917299240910907, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14811 + }, + { + "epoch": 5.917698761486217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14812 + }, + { + "epoch": 5.918098282061527, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14813 + }, + { + "epoch": 5.918497802636836, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14814 + }, + { + "epoch": 5.918897323212145, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14815 + }, + { + "epoch": 5.919296843787455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14816 + }, + { + "epoch": 5.919696364362765, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14817 + }, + { + "epoch": 5.920095884938075, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14818 + }, + { + "epoch": 5.920495405513384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14819 + }, + { + "epoch": 5.920894926088693, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14820 + }, + { + "epoch": 5.921294446664003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14821 + }, + { + "epoch": 5.921693967239313, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14822 + }, + { + "epoch": 5.922093487814623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14823 + }, + { + "epoch": 5.922493008389932, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14824 + }, + { + "epoch": 5.922892528965241, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14825 + }, + { + "epoch": 5.923292049540551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14826 + }, + { + "epoch": 5.923691570115861, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14827 + }, + { + "epoch": 5.924091090691171, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14828 + }, + { + "epoch": 5.924490611266481, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14829 + }, + { + "epoch": 5.9248901318417895, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14830 + }, + { + "epoch": 5.925289652417099, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14831 + }, + { + "epoch": 5.925689172992409, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14832 + }, + { + "epoch": 5.926088693567719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14833 + }, + { + "epoch": 5.926488214143029, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14834 + }, + { + "epoch": 5.926887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14835 + }, + { + "epoch": 5.927287255293647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14836 + }, + { + "epoch": 5.927686775868957, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 14837 + }, + { + "epoch": 5.928086296444267, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14838 + }, + { + "epoch": 5.928485817019577, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14839 + }, + { + "epoch": 5.9288853375948865, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14840 + }, + { + "epoch": 5.929284858170195, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 14841 + }, + { + "epoch": 5.929684378745505, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14842 + }, + { + "epoch": 5.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14843 + }, + { + "epoch": 5.930483419896125, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 14844 + }, + { + "epoch": 5.9308829404714345, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 14845 + }, + { + "epoch": 5.931282461046744, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14846 + }, + { + "epoch": 5.931681981622053, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14847 + }, + { + "epoch": 5.932081502197363, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14848 + }, + { + "epoch": 5.932481022772673, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14849 + }, + { + "epoch": 5.9328805433479825, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14850 + }, + { + "epoch": 5.933280063923292, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14851 + }, + { + "epoch": 5.933679584498602, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14852 + }, + { + "epoch": 5.934079105073911, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14853 + }, + { + "epoch": 5.934478625649221, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14854 + }, + { + "epoch": 5.934878146224531, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14855 + }, + { + "epoch": 5.93527766679984, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14856 + }, + { + "epoch": 5.93567718737515, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14857 + }, + { + "epoch": 5.936076707950459, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14858 + }, + { + "epoch": 5.936476228525769, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14859 + }, + { + "epoch": 5.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 14860 + }, + { + "epoch": 5.937275269676388, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 14861 + }, + { + "epoch": 5.937674790251698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14862 + }, + { + "epoch": 5.938074310827007, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14863 + }, + { + "epoch": 5.938473831402317, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14864 + }, + { + "epoch": 5.938873351977627, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14865 + }, + { + "epoch": 5.9392728725529365, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 14866 + }, + { + "epoch": 5.939672393128246, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14867 + }, + { + "epoch": 5.940071913703556, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14868 + }, + { + "epoch": 5.940471434278866, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14869 + }, + { + "epoch": 5.940870954854175, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14870 + }, + { + "epoch": 5.9412704754294845, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14871 + }, + { + "epoch": 5.941669996004794, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14872 + }, + { + "epoch": 5.942069516580104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14873 + }, + { + "epoch": 5.942469037155414, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14874 + }, + { + "epoch": 5.942868557730723, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14875 + }, + { + "epoch": 5.9432680783060325, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14876 + }, + { + "epoch": 5.943667598881342, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14877 + }, + { + "epoch": 5.944067119456652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14878 + }, + { + "epoch": 5.944466640031962, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14879 + }, + { + "epoch": 5.944866160607271, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 14880 + }, + { + "epoch": 5.945265681182581, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14881 + }, + { + "epoch": 5.94566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14882 + }, + { + "epoch": 5.9460647223332, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14883 + }, + { + "epoch": 5.94646424290851, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14884 + }, + { + "epoch": 5.94686376348382, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14885 + }, + { + "epoch": 5.9472632840591295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14886 + }, + { + "epoch": 5.947662804634438, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 14887 + }, + { + "epoch": 5.948062325209748, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14888 + }, + { + "epoch": 5.948461845785058, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14889 + }, + { + "epoch": 5.948861366360368, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14890 + }, + { + "epoch": 5.949260886935678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14891 + }, + { + "epoch": 5.9496604075109865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14892 + }, + { + "epoch": 5.950059928086296, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14893 + }, + { + "epoch": 5.950459448661606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14894 + }, + { + "epoch": 5.950858969236916, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14895 + }, + { + "epoch": 5.951258489812226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 14896 + }, + { + "epoch": 5.9516580103875345, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 14897 + }, + { + "epoch": 5.952057530962844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14898 + }, + { + "epoch": 5.952457051538154, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14899 + }, + { + "epoch": 5.952856572113464, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14900 + }, + { + "epoch": 5.953256092688774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14901 + }, + { + "epoch": 5.953655613264083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14902 + }, + { + "epoch": 5.954055133839393, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14903 + }, + { + "epoch": 5.954454654414702, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14904 + }, + { + "epoch": 5.954854174990012, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14905 + }, + { + "epoch": 5.955253695565322, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14906 + }, + { + "epoch": 5.9556532161406315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14907 + }, + { + "epoch": 5.956052736715941, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14908 + }, + { + "epoch": 5.95645225729125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14909 + }, + { + "epoch": 5.95685177786656, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14910 + }, + { + "epoch": 5.95725129844187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14911 + }, + { + "epoch": 5.9576508190171795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14912 + }, + { + "epoch": 5.958050339592489, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14913 + }, + { + "epoch": 5.958449860167798, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14914 + }, + { + "epoch": 5.958849380743108, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14915 + }, + { + "epoch": 5.959248901318418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14916 + }, + { + "epoch": 5.959648421893728, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14917 + }, + { + "epoch": 5.960047942469037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14918 + }, + { + "epoch": 5.960447463044347, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14919 + }, + { + "epoch": 5.960846983619656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14920 + }, + { + "epoch": 5.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14921 + }, + { + "epoch": 5.961646024770276, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14922 + }, + { + "epoch": 5.962045545345585, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14923 + }, + { + "epoch": 5.962445065920895, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14924 + }, + { + "epoch": 5.962844586496205, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14925 + }, + { + "epoch": 5.963244107071514, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14926 + }, + { + "epoch": 5.963643627646824, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14927 + }, + { + "epoch": 5.964043148222133, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14928 + }, + { + "epoch": 5.964442668797443, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14929 + }, + { + "epoch": 5.964842189372753, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14930 + }, + { + "epoch": 5.965241709948062, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14931 + }, + { + "epoch": 5.965641230523372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14932 + }, + { + "epoch": 5.9660407510986815, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14933 + }, + { + "epoch": 5.966440271673991, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14934 + }, + { + "epoch": 5.966839792249301, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14935 + }, + { + "epoch": 5.967239312824611, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14936 + }, + { + "epoch": 5.96763883339992, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14937 + }, + { + "epoch": 5.9680383539752295, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14938 + }, + { + "epoch": 5.968437874550539, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14939 + }, + { + "epoch": 5.968837395125849, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14940 + }, + { + "epoch": 5.969236915701159, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14941 + }, + { + "epoch": 5.969636436276469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14942 + }, + { + "epoch": 5.970035956851778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14943 + }, + { + "epoch": 5.970435477427087, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14944 + }, + { + "epoch": 5.970834998002397, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 14945 + }, + { + "epoch": 5.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14946 + }, + { + "epoch": 5.971634039153017, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14947 + }, + { + "epoch": 5.972033559728326, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14948 + }, + { + "epoch": 5.972433080303635, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14949 + }, + { + "epoch": 5.972832600878945, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14950 + }, + { + "epoch": 5.973232121454255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14951 + }, + { + "epoch": 5.973631642029565, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14952 + }, + { + "epoch": 5.974031162604874, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14953 + }, + { + "epoch": 5.974430683180183, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14954 + }, + { + "epoch": 5.974830203755493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14955 + }, + { + "epoch": 5.975229724330803, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14956 + }, + { + "epoch": 5.975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14957 + }, + { + "epoch": 5.976028765481423, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14958 + }, + { + "epoch": 5.976428286056732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14959 + }, + { + "epoch": 5.976827806632041, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 14960 + }, + { + "epoch": 5.977227327207351, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14961 + }, + { + "epoch": 5.977626847782661, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14962 + }, + { + "epoch": 5.978026368357971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14963 + }, + { + "epoch": 5.97842588893328, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14964 + }, + { + "epoch": 5.978825409508589, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14965 + }, + { + "epoch": 5.979224930083899, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14966 + }, + { + "epoch": 5.979624450659209, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14967 + }, + { + "epoch": 5.980023971234519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14968 + }, + { + "epoch": 5.9804234918098285, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14969 + }, + { + "epoch": 5.980823012385137, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14970 + }, + { + "epoch": 5.981222532960447, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14971 + }, + { + "epoch": 5.981622053535757, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 14972 + }, + { + "epoch": 5.982021574111067, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14973 + }, + { + "epoch": 5.9824210946863765, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14974 + }, + { + "epoch": 5.982820615261686, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14975 + }, + { + "epoch": 5.983220135836996, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14976 + }, + { + "epoch": 5.983619656412305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14977 + }, + { + "epoch": 5.984019176987615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14978 + }, + { + "epoch": 5.9844186975629245, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14979 + }, + { + "epoch": 5.984818218138234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14980 + }, + { + "epoch": 5.985217738713544, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14981 + }, + { + "epoch": 5.985617259288853, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14982 + }, + { + "epoch": 5.986016779864163, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14983 + }, + { + "epoch": 5.986416300439473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14984 + }, + { + "epoch": 5.986815821014782, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14985 + }, + { + "epoch": 5.987215341590092, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14986 + }, + { + "epoch": 5.987614862165401, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14987 + }, + { + "epoch": 5.988014382740711, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14988 + }, + { + "epoch": 5.988413903316021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14989 + }, + { + "epoch": 5.98881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14990 + }, + { + "epoch": 5.98921294446664, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14991 + }, + { + "epoch": 5.98961246504195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14992 + }, + { + "epoch": 5.99001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14993 + }, + { + "epoch": 5.990411506192569, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14994 + }, + { + "epoch": 5.9908110267678785, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14995 + }, + { + "epoch": 5.991210547343188, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14996 + }, + { + "epoch": 5.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14997 + }, + { + "epoch": 5.992009588493808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14998 + }, + { + "epoch": 5.992409109069117, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14999 + }, + { + "epoch": 5.9928086296444265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15000 + }, + { + "epoch": 5.993208150219736, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15001 + }, + { + "epoch": 5.993607670795046, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 15002 + }, + { + "epoch": 5.994007191370356, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15003 + }, + { + "epoch": 5.994406711945665, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15004 + }, + { + "epoch": 5.9948062325209746, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15005 + }, + { + "epoch": 5.995205753096284, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 15006 + }, + { + "epoch": 5.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15007 + }, + { + "epoch": 5.996004794246904, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 15008 + }, + { + "epoch": 5.996404314822214, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15009 + }, + { + "epoch": 5.996803835397523, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15010 + }, + { + "epoch": 5.997203355972832, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15011 + }, + { + "epoch": 5.997602876548142, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15012 + }, + { + "epoch": 5.998002397123452, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15013 + }, + { + "epoch": 5.998401917698762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15014 + }, + { + "epoch": 5.9988014382740715, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15015 + }, + { + "epoch": 5.99920095884938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15016 + }, + { + "epoch": 5.99960047942469, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 15017 + }, + { + "epoch": 6.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15018 + }, + { + "epoch": 6.00039952057531, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15019 + }, + { + "epoch": 6.00079904115062, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15020 + }, + { + "epoch": 6.0011985617259285, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15021 + }, + { + "epoch": 6.001598082301238, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15022 + }, + { + "epoch": 6.001997602876548, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15023 + }, + { + "epoch": 6.002397123451858, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15024 + }, + { + "epoch": 6.002796644027168, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 15025 + }, + { + "epoch": 6.003196164602477, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15026 + }, + { + "epoch": 6.003595685177786, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15027 + }, + { + "epoch": 6.003995205753096, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15028 + }, + { + "epoch": 6.004394726328406, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15029 + }, + { + "epoch": 6.004794246903716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15030 + }, + { + "epoch": 6.0051937674790254, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15031 + }, + { + "epoch": 6.005593288054335, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15032 + }, + { + "epoch": 6.005992808629644, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15033 + }, + { + "epoch": 6.006392329204954, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15034 + }, + { + "epoch": 6.006791849780264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15035 + }, + { + "epoch": 6.0071913703555735, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 15036 + }, + { + "epoch": 6.007590890930883, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15037 + }, + { + "epoch": 6.007990411506192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15038 + }, + { + "epoch": 6.008389932081502, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15039 + }, + { + "epoch": 6.008789452656812, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15040 + }, + { + "epoch": 6.0091889732321215, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15041 + }, + { + "epoch": 6.009588493807431, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15042 + }, + { + "epoch": 6.009988014382741, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15043 + }, + { + "epoch": 6.01038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15044 + }, + { + "epoch": 6.01078705553336, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15045 + }, + { + "epoch": 6.01118657610867, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15046 + }, + { + "epoch": 6.011586096683979, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15047 + }, + { + "epoch": 6.011985617259289, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15048 + }, + { + "epoch": 6.012385137834598, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15049 + }, + { + "epoch": 6.012784658409908, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15050 + }, + { + "epoch": 6.013184178985218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15051 + }, + { + "epoch": 6.013583699560527, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15052 + }, + { + "epoch": 6.013983220135837, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15053 + }, + { + "epoch": 6.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15054 + }, + { + "epoch": 6.014782261286456, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15055 + }, + { + "epoch": 6.015181781861766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15056 + }, + { + "epoch": 6.0155813024370755, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15057 + }, + { + "epoch": 6.015980823012385, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 15058 + }, + { + "epoch": 6.016380343587695, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15059 + }, + { + "epoch": 6.016779864163005, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 15060 + }, + { + "epoch": 6.017179384738314, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15061 + }, + { + "epoch": 6.0175789053136235, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15062 + }, + { + "epoch": 6.017978425888933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15063 + }, + { + "epoch": 6.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15064 + }, + { + "epoch": 6.018777467039553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15065 + }, + { + "epoch": 6.019176987614862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15066 + }, + { + "epoch": 6.0195765081901715, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15067 + }, + { + "epoch": 6.019976028765481, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15068 + }, + { + "epoch": 6.020375549340791, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15069 + }, + { + "epoch": 6.020775069916101, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 15070 + }, + { + "epoch": 6.021174590491411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15071 + }, + { + "epoch": 6.02157411106672, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15072 + }, + { + "epoch": 6.021973631642029, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15073 + }, + { + "epoch": 6.022373152217339, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15074 + }, + { + "epoch": 6.022772672792649, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 15075 + }, + { + "epoch": 6.023172193367959, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15076 + }, + { + "epoch": 6.0235717139432685, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15077 + }, + { + "epoch": 6.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15078 + }, + { + "epoch": 6.024370755093887, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15079 + }, + { + "epoch": 6.024770275669197, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15080 + }, + { + "epoch": 6.025169796244507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15081 + }, + { + "epoch": 6.025569316819817, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15082 + }, + { + "epoch": 6.0259688373951255, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15083 + }, + { + "epoch": 6.026368357970435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15084 + }, + { + "epoch": 6.026767878545745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15085 + }, + { + "epoch": 6.027167399121055, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15086 + }, + { + "epoch": 6.027566919696365, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15087 + }, + { + "epoch": 6.027966440271674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15088 + }, + { + "epoch": 6.028365960846983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15089 + }, + { + "epoch": 6.028765481422293, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15090 + }, + { + "epoch": 6.029165001997603, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15091 + }, + { + "epoch": 6.029564522572913, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15092 + }, + { + "epoch": 6.029964043148222, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15093 + }, + { + "epoch": 6.030363563723531, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15094 + }, + { + "epoch": 6.030763084298841, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15095 + }, + { + "epoch": 6.031162604874151, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15096 + }, + { + "epoch": 6.031562125449461, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15097 + }, + { + "epoch": 6.0319616460247705, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15098 + }, + { + "epoch": 6.03236116660008, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15099 + }, + { + "epoch": 6.032760687175389, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15100 + }, + { + "epoch": 6.033160207750699, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15101 + }, + { + "epoch": 6.033559728326009, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15102 + }, + { + "epoch": 6.0339592489013185, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15103 + }, + { + "epoch": 6.034358769476628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15104 + }, + { + "epoch": 6.034758290051938, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15105 + }, + { + "epoch": 6.035157810627247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15106 + }, + { + "epoch": 6.035557331202557, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15107 + }, + { + "epoch": 6.035956851777867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 15108 + }, + { + "epoch": 6.036356372353176, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15109 + }, + { + "epoch": 6.036755892928486, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15110 + }, + { + "epoch": 6.037155413503795, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15111 + }, + { + "epoch": 6.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15112 + }, + { + "epoch": 6.037954454654415, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15113 + }, + { + "epoch": 6.038353975229724, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15114 + }, + { + "epoch": 6.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15115 + }, + { + "epoch": 6.039153016380344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15116 + }, + { + "epoch": 6.039552536955653, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15117 + }, + { + "epoch": 6.039952057530963, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15118 + }, + { + "epoch": 6.040351578106272, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15119 + }, + { + "epoch": 6.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15120 + }, + { + "epoch": 6.041150619256892, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15121 + }, + { + "epoch": 6.041550139832202, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15122 + }, + { + "epoch": 6.041949660407511, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15123 + }, + { + "epoch": 6.0423491809828205, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 15124 + }, + { + "epoch": 6.04274870155813, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15125 + }, + { + "epoch": 6.04314822213344, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15126 + }, + { + "epoch": 6.04354774270875, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15127 + }, + { + "epoch": 6.043947263284059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15128 + }, + { + "epoch": 6.0443467838593685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15129 + }, + { + "epoch": 6.044746304434678, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15130 + }, + { + "epoch": 6.045145825009988, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15131 + }, + { + "epoch": 6.045545345585298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15132 + }, + { + "epoch": 6.045944866160608, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15133 + }, + { + "epoch": 6.046344386735917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15134 + }, + { + "epoch": 6.046743907311226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15135 + }, + { + "epoch": 6.047143427886536, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15136 + }, + { + "epoch": 6.047542948461846, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15137 + }, + { + "epoch": 6.047942469037156, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15138 + }, + { + "epoch": 6.048341989612465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15139 + }, + { + "epoch": 6.048741510187774, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15140 + }, + { + "epoch": 6.049141030763084, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15141 + }, + { + "epoch": 6.049540551338394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15142 + }, + { + "epoch": 6.049940071913704, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15143 + }, + { + "epoch": 6.0503395924890135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15144 + }, + { + "epoch": 6.050739113064322, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15145 + }, + { + "epoch": 6.051138633639632, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15146 + }, + { + "epoch": 6.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15147 + }, + { + "epoch": 6.051937674790252, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15148 + }, + { + "epoch": 6.052337195365562, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15149 + }, + { + "epoch": 6.052736715940871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15150 + }, + { + "epoch": 6.05313623651618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15151 + }, + { + "epoch": 6.05353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15152 + }, + { + "epoch": 6.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 15153 + }, + { + "epoch": 6.05433479824211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 15154 + }, + { + "epoch": 6.054734318817419, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15155 + }, + { + "epoch": 6.055133839392728, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15156 + }, + { + "epoch": 6.055533359968038, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15157 + }, + { + "epoch": 6.055932880543348, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15158 + }, + { + "epoch": 6.056332401118658, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15159 + }, + { + "epoch": 6.0567319216939675, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15160 + }, + { + "epoch": 6.057131442269277, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15161 + }, + { + "epoch": 6.057530962844586, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15162 + }, + { + "epoch": 6.057930483419896, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15163 + }, + { + "epoch": 6.058330003995206, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15164 + }, + { + "epoch": 6.0587295245705155, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15165 + }, + { + "epoch": 6.059129045145825, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15166 + }, + { + "epoch": 6.059528565721135, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15167 + }, + { + "epoch": 6.059928086296444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15168 + }, + { + "epoch": 6.060327606871754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15169 + }, + { + "epoch": 6.0607271274470635, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15170 + }, + { + "epoch": 6.061126648022373, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15171 + }, + { + "epoch": 6.061526168597683, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15172 + }, + { + "epoch": 6.061925689172992, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15173 + }, + { + "epoch": 6.062325209748302, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15174 + }, + { + "epoch": 6.062724730323612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 15175 + }, + { + "epoch": 6.063124250898921, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15176 + }, + { + "epoch": 6.063523771474231, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15177 + }, + { + "epoch": 6.063923292049541, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15178 + }, + { + "epoch": 6.06432281262485, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 15179 + }, + { + "epoch": 6.06472233320016, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15180 + }, + { + "epoch": 6.065121853775469, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15181 + }, + { + "epoch": 6.065521374350779, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15182 + }, + { + "epoch": 6.065920894926089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15183 + }, + { + "epoch": 6.066320415501398, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15184 + }, + { + "epoch": 6.066719936076708, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15185 + }, + { + "epoch": 6.0671194566520175, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15186 + }, + { + "epoch": 6.067518977227327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15187 + }, + { + "epoch": 6.067918497802637, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15188 + }, + { + "epoch": 6.068318018377947, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15189 + }, + { + "epoch": 6.068717538953256, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15190 + }, + { + "epoch": 6.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15191 + }, + { + "epoch": 6.069516580103875, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15192 + }, + { + "epoch": 6.069916100679185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15193 + }, + { + "epoch": 6.070315621254495, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15194 + }, + { + "epoch": 6.070715141829805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15195 + }, + { + "epoch": 6.0711146624051135, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15196 + }, + { + "epoch": 6.071514182980423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15197 + }, + { + "epoch": 6.071913703555733, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15198 + }, + { + "epoch": 6.072313224131043, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15199 + }, + { + "epoch": 6.072712744706353, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15200 + }, + { + "epoch": 6.073112265281662, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15201 + }, + { + "epoch": 6.073511785856971, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15202 + }, + { + "epoch": 6.073911306432281, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15203 + }, + { + "epoch": 6.074310827007591, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15204 + }, + { + "epoch": 6.074710347582901, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15205 + }, + { + "epoch": 6.0751098681582105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15206 + }, + { + "epoch": 6.075509388733519, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15207 + }, + { + "epoch": 6.075908909308829, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15208 + }, + { + "epoch": 6.076308429884139, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15209 + }, + { + "epoch": 6.076707950459449, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15210 + }, + { + "epoch": 6.077107471034759, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15211 + }, + { + "epoch": 6.077506991610068, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15212 + }, + { + "epoch": 6.077906512185377, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15213 + }, + { + "epoch": 6.078306032760687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15214 + }, + { + "epoch": 6.078705553335997, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15215 + }, + { + "epoch": 6.079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15216 + }, + { + "epoch": 6.079504594486616, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15217 + }, + { + "epoch": 6.079904115061925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15218 + }, + { + "epoch": 6.080303635637235, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15219 + }, + { + "epoch": 6.080703156212545, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 15220 + }, + { + "epoch": 6.081102676787855, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15221 + }, + { + "epoch": 6.081502197363164, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15222 + }, + { + "epoch": 6.081901717938474, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15223 + }, + { + "epoch": 6.082301238513783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15224 + }, + { + "epoch": 6.082700759089093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15225 + }, + { + "epoch": 6.083100279664403, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15226 + }, + { + "epoch": 6.0834998002397125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15227 + }, + { + "epoch": 6.083899320815022, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15228 + }, + { + "epoch": 6.084298841390331, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15229 + }, + { + "epoch": 6.084698361965641, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15230 + }, + { + "epoch": 6.085097882540951, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15231 + }, + { + "epoch": 6.0854974031162605, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15232 + }, + { + "epoch": 6.08589692369157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15233 + }, + { + "epoch": 6.08629644426688, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15234 + }, + { + "epoch": 6.086695964842189, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15235 + }, + { + "epoch": 6.087095485417499, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15236 + }, + { + "epoch": 6.087495005992809, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15237 + }, + { + "epoch": 6.087894526568118, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15238 + }, + { + "epoch": 6.088294047143428, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15239 + }, + { + "epoch": 6.088693567718738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15240 + }, + { + "epoch": 6.089093088294047, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15241 + }, + { + "epoch": 6.089492608869357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15242 + }, + { + "epoch": 6.089892129444666, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15243 + }, + { + "epoch": 6.090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15244 + }, + { + "epoch": 6.090691170595286, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15245 + }, + { + "epoch": 6.091090691170595, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15246 + }, + { + "epoch": 6.091490211745905, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15247 + }, + { + "epoch": 6.0918897323212144, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15248 + }, + { + "epoch": 6.092289252896524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15249 + }, + { + "epoch": 6.092688773471834, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15250 + }, + { + "epoch": 6.093088294047144, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15251 + }, + { + "epoch": 6.093487814622453, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15252 + }, + { + "epoch": 6.0938873351977625, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15253 + }, + { + "epoch": 6.094286855773072, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15254 + }, + { + "epoch": 6.094686376348382, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15255 + }, + { + "epoch": 6.095085896923692, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15256 + }, + { + "epoch": 6.095485417499002, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15257 + }, + { + "epoch": 6.0958849380743105, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15258 + }, + { + "epoch": 6.09628445864962, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15259 + }, + { + "epoch": 6.09668397922493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15260 + }, + { + "epoch": 6.09708349980024, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15261 + }, + { + "epoch": 6.09748302037555, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15262 + }, + { + "epoch": 6.097882540950859, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15263 + }, + { + "epoch": 6.098282061526168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15264 + }, + { + "epoch": 6.098681582101478, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15265 + }, + { + "epoch": 6.099081102676788, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15266 + }, + { + "epoch": 6.099480623252098, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15267 + }, + { + "epoch": 6.0998801438274075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15268 + }, + { + "epoch": 6.100279664402716, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15269 + }, + { + "epoch": 6.100679184978026, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15270 + }, + { + "epoch": 6.101078705553336, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15271 + }, + { + "epoch": 6.101478226128646, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15272 + }, + { + "epoch": 6.1018777467039556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15273 + }, + { + "epoch": 6.1022772672792644, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15274 + }, + { + "epoch": 6.102676787854574, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15275 + }, + { + "epoch": 6.103076308429884, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15276 + }, + { + "epoch": 6.103475829005194, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15277 + }, + { + "epoch": 6.103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15278 + }, + { + "epoch": 6.104274870155813, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15279 + }, + { + "epoch": 6.104674390731122, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15280 + }, + { + "epoch": 6.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15281 + }, + { + "epoch": 6.105473431881742, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15282 + }, + { + "epoch": 6.105872952457052, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15283 + }, + { + "epoch": 6.106272473032361, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15284 + }, + { + "epoch": 6.106671993607671, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15285 + }, + { + "epoch": 6.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15286 + }, + { + "epoch": 6.10747103475829, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15287 + }, + { + "epoch": 6.1078705553336, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15288 + }, + { + "epoch": 6.1082700759089095, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15289 + }, + { + "epoch": 6.108669596484219, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15290 + }, + { + "epoch": 6.109069117059528, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15291 + }, + { + "epoch": 6.109468637634838, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15292 + }, + { + "epoch": 6.109868158210148, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15293 + }, + { + "epoch": 6.1102676787854575, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15294 + }, + { + "epoch": 6.110667199360767, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15295 + }, + { + "epoch": 6.111066719936077, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15296 + }, + { + "epoch": 6.111466240511386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15297 + }, + { + "epoch": 6.111865761086696, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15298 + }, + { + "epoch": 6.112265281662006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15299 + }, + { + "epoch": 6.112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15300 + }, + { + "epoch": 6.113064322812625, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15301 + }, + { + "epoch": 6.113463843387935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15302 + }, + { + "epoch": 6.113863363963244, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15303 + }, + { + "epoch": 6.114262884538554, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15304 + }, + { + "epoch": 6.114662405113863, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15305 + }, + { + "epoch": 6.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15306 + }, + { + "epoch": 6.115461446264483, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15307 + }, + { + "epoch": 6.115860966839792, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15308 + }, + { + "epoch": 6.116260487415102, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15309 + }, + { + "epoch": 6.116660007990411, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15310 + }, + { + "epoch": 6.117059528565721, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15311 + }, + { + "epoch": 6.117459049141031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 15312 + }, + { + "epoch": 6.117858569716341, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15313 + }, + { + "epoch": 6.11825809029165, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15314 + }, + { + "epoch": 6.1186576108669595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15315 + }, + { + "epoch": 6.119057131442269, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 15316 + }, + { + "epoch": 6.119456652017579, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15317 + }, + { + "epoch": 6.119856172592889, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15318 + }, + { + "epoch": 6.120255693168198, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15319 + }, + { + "epoch": 6.1206552137435075, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15320 + }, + { + "epoch": 6.121054734318817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15321 + }, + { + "epoch": 6.121454254894127, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15322 + }, + { + "epoch": 6.121853775469437, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15323 + }, + { + "epoch": 6.122253296044747, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15324 + }, + { + "epoch": 6.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15325 + }, + { + "epoch": 6.123052337195365, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15326 + }, + { + "epoch": 6.123451857770675, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15327 + }, + { + "epoch": 6.123851378345985, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15328 + }, + { + "epoch": 6.124250898921295, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15329 + }, + { + "epoch": 6.1246504194966045, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15330 + }, + { + "epoch": 6.125049940071913, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 15331 + }, + { + "epoch": 6.125449460647223, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15332 + }, + { + "epoch": 6.125848981222533, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15333 + }, + { + "epoch": 6.126248501797843, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15334 + }, + { + "epoch": 6.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15335 + }, + { + "epoch": 6.127047542948461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15336 + }, + { + "epoch": 6.127447063523771, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15337 + }, + { + "epoch": 6.127846584099081, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15338 + }, + { + "epoch": 6.128246104674391, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 15339 + }, + { + "epoch": 6.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15340 + }, + { + "epoch": 6.12904514582501, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15341 + }, + { + "epoch": 6.129444666400319, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15342 + }, + { + "epoch": 6.129844186975629, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15343 + }, + { + "epoch": 6.130243707550939, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15344 + }, + { + "epoch": 6.130643228126249, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15345 + }, + { + "epoch": 6.131042748701558, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15346 + }, + { + "epoch": 6.131442269276867, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15347 + }, + { + "epoch": 6.131841789852177, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15348 + }, + { + "epoch": 6.132241310427487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15349 + }, + { + "epoch": 6.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15350 + }, + { + "epoch": 6.1330403515781065, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15351 + }, + { + "epoch": 6.133439872153416, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15352 + }, + { + "epoch": 6.133839392728725, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15353 + }, + { + "epoch": 6.134238913304035, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15354 + }, + { + "epoch": 6.134638433879345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15355 + }, + { + "epoch": 6.1350379544546545, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15356 + }, + { + "epoch": 6.135437475029964, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15357 + }, + { + "epoch": 6.135836995605274, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15358 + }, + { + "epoch": 6.136236516180583, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15359 + }, + { + "epoch": 6.136636036755893, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15360 + }, + { + "epoch": 6.1370355573312025, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15361 + }, + { + "epoch": 6.137435077906512, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15362 + }, + { + "epoch": 6.137834598481822, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15363 + }, + { + "epoch": 6.138234119057131, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15364 + }, + { + "epoch": 6.138633639632441, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 15365 + }, + { + "epoch": 6.139033160207751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15366 + }, + { + "epoch": 6.13943268078306, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15367 + }, + { + "epoch": 6.13983220135837, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15368 + }, + { + "epoch": 6.14023172193368, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15369 + }, + { + "epoch": 6.140631242508989, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15370 + }, + { + "epoch": 6.141030763084299, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15371 + }, + { + "epoch": 6.141430283659608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15372 + }, + { + "epoch": 6.141829804234918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15373 + }, + { + "epoch": 6.142229324810228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15374 + }, + { + "epoch": 6.142628845385538, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15375 + }, + { + "epoch": 6.143028365960847, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15376 + }, + { + "epoch": 6.1434278865361565, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15377 + }, + { + "epoch": 6.143827407111466, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15378 + }, + { + "epoch": 6.144226927686776, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15379 + }, + { + "epoch": 6.144626448262086, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15380 + }, + { + "epoch": 6.145025968837395, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15381 + }, + { + "epoch": 6.1454254894127045, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 15382 + }, + { + "epoch": 6.145825009988014, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15383 + }, + { + "epoch": 6.146224530563324, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15384 + }, + { + "epoch": 6.146624051138634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15385 + }, + { + "epoch": 6.147023571713944, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15386 + }, + { + "epoch": 6.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15387 + }, + { + "epoch": 6.147822612864562, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15388 + }, + { + "epoch": 6.148222133439872, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15389 + }, + { + "epoch": 6.148621654015182, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15390 + }, + { + "epoch": 6.149021174590492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15391 + }, + { + "epoch": 6.1494206951658015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15392 + }, + { + "epoch": 6.14982021574111, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15393 + }, + { + "epoch": 6.15021973631642, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15394 + }, + { + "epoch": 6.15061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15395 + }, + { + "epoch": 6.15101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15396 + }, + { + "epoch": 6.1514182980423495, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15397 + }, + { + "epoch": 6.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 15398 + }, + { + "epoch": 6.152217339192968, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15399 + }, + { + "epoch": 6.152616859768278, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15400 + }, + { + "epoch": 6.153016380343588, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15401 + }, + { + "epoch": 6.153415900918898, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15402 + }, + { + "epoch": 6.153815421494207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15403 + }, + { + "epoch": 6.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2963, + "step": 15404 + }, + { + "epoch": 6.154614462644826, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15405 + }, + { + "epoch": 6.155013983220136, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15406 + }, + { + "epoch": 6.155413503795446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15407 + }, + { + "epoch": 6.155813024370755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15408 + }, + { + "epoch": 6.156212544946065, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15409 + }, + { + "epoch": 6.156612065521374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15410 + }, + { + "epoch": 6.157011586096684, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15411 + }, + { + "epoch": 6.157411106671994, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15412 + }, + { + "epoch": 6.157810627247303, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15413 + }, + { + "epoch": 6.158210147822613, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15414 + }, + { + "epoch": 6.158609668397922, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15415 + }, + { + "epoch": 6.159009188973232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15416 + }, + { + "epoch": 6.159408709548542, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15417 + }, + { + "epoch": 6.1598082301238515, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15418 + }, + { + "epoch": 6.160207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15419 + }, + { + "epoch": 6.160607271274471, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.2983, + "step": 15420 + }, + { + "epoch": 6.16100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 15421 + }, + { + "epoch": 6.16140631242509, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15422 + }, + { + "epoch": 6.1618058330003995, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15423 + }, + { + "epoch": 6.162205353575709, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15424 + }, + { + "epoch": 6.162604874151019, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15425 + }, + { + "epoch": 6.163004394726328, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15426 + }, + { + "epoch": 6.163403915301638, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15427 + }, + { + "epoch": 6.163803435876948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15428 + }, + { + "epoch": 6.164202956452257, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15429 + }, + { + "epoch": 6.164602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15430 + }, + { + "epoch": 6.165001997602877, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15431 + }, + { + "epoch": 6.165401518178186, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15432 + }, + { + "epoch": 6.165801038753496, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15433 + }, + { + "epoch": 6.166200559328805, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15434 + }, + { + "epoch": 6.166600079904115, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15435 + }, + { + "epoch": 6.166999600479425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15436 + }, + { + "epoch": 6.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 15437 + }, + { + "epoch": 6.167798641630044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15438 + }, + { + "epoch": 6.168198162205353, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15439 + }, + { + "epoch": 6.168597682780663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15440 + }, + { + "epoch": 6.168997203355973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15441 + }, + { + "epoch": 6.169396723931283, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15442 + }, + { + "epoch": 6.169796244506592, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15443 + }, + { + "epoch": 6.1701957650819015, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 15444 + }, + { + "epoch": 6.170595285657211, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15445 + }, + { + "epoch": 6.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15446 + }, + { + "epoch": 6.171394326807831, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15447 + }, + { + "epoch": 6.171793847383141, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15448 + }, + { + "epoch": 6.1721933679584495, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15449 + }, + { + "epoch": 6.172592888533759, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15450 + }, + { + "epoch": 6.172992409109069, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 15451 + }, + { + "epoch": 6.173391929684379, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15452 + }, + { + "epoch": 6.173791450259689, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15453 + }, + { + "epoch": 6.174190970834998, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15454 + }, + { + "epoch": 6.174590491410307, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15455 + }, + { + "epoch": 6.174990011985617, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15456 + }, + { + "epoch": 6.175389532560927, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15457 + }, + { + "epoch": 6.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15458 + }, + { + "epoch": 6.1761885737115465, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15459 + }, + { + "epoch": 6.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15460 + }, + { + "epoch": 6.176987614862165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15461 + }, + { + "epoch": 6.177387135437475, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 15462 + }, + { + "epoch": 6.177786656012785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15463 + }, + { + "epoch": 6.1781861765880945, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15464 + }, + { + "epoch": 6.178585697163404, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15465 + }, + { + "epoch": 6.178985217738713, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15466 + }, + { + "epoch": 6.179384738314023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15467 + }, + { + "epoch": 6.179784258889333, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15468 + }, + { + "epoch": 6.180183779464643, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15469 + }, + { + "epoch": 6.180583300039952, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15470 + }, + { + "epoch": 6.180982820615261, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15471 + }, + { + "epoch": 6.181382341190571, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15472 + }, + { + "epoch": 6.181781861765881, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15473 + }, + { + "epoch": 6.182181382341191, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15474 + }, + { + "epoch": 6.1825809029165, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15475 + }, + { + "epoch": 6.18298042349181, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15476 + }, + { + "epoch": 6.183379944067119, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15477 + }, + { + "epoch": 6.183779464642429, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 15478 + }, + { + "epoch": 6.184178985217739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15479 + }, + { + "epoch": 6.1845785057930485, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15480 + }, + { + "epoch": 6.184978026368358, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15481 + }, + { + "epoch": 6.185377546943668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15482 + }, + { + "epoch": 6.185777067518977, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15483 + }, + { + "epoch": 6.186176588094287, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15484 + }, + { + "epoch": 6.1865761086695965, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15485 + }, + { + "epoch": 6.186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15486 + }, + { + "epoch": 6.187375149820216, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15487 + }, + { + "epoch": 6.187774670395525, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15488 + }, + { + "epoch": 6.188174190970835, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15489 + }, + { + "epoch": 6.1885737115461446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15490 + }, + { + "epoch": 6.188973232121454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15491 + }, + { + "epoch": 6.189372752696764, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15492 + }, + { + "epoch": 6.189772273272074, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15493 + }, + { + "epoch": 6.190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15494 + }, + { + "epoch": 6.190571314422693, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15495 + }, + { + "epoch": 6.190970834998002, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15496 + }, + { + "epoch": 6.191370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 15497 + }, + { + "epoch": 6.191769876148622, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15498 + }, + { + "epoch": 6.192169396723932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15499 + }, + { + "epoch": 6.192568917299241, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15500 + }, + { + "epoch": 6.19296843787455, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15501 + }, + { + "epoch": 6.19336795844986, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15502 + }, + { + "epoch": 6.19376747902517, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15503 + }, + { + "epoch": 6.19416699960048, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15504 + }, + { + "epoch": 6.194566520175789, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15505 + }, + { + "epoch": 6.1949660407510985, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15506 + }, + { + "epoch": 6.195365561326408, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15507 + }, + { + "epoch": 6.195765081901718, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15508 + }, + { + "epoch": 6.196164602477028, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15509 + }, + { + "epoch": 6.196564123052338, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15510 + }, + { + "epoch": 6.1969636436276465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15511 + }, + { + "epoch": 6.197363164202956, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15512 + }, + { + "epoch": 6.197762684778266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15513 + }, + { + "epoch": 6.198162205353576, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15514 + }, + { + "epoch": 6.198561725928886, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15515 + }, + { + "epoch": 6.1989612465041946, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15516 + }, + { + "epoch": 6.199360767079504, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15517 + }, + { + "epoch": 6.199760287654814, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15518 + }, + { + "epoch": 6.200159808230124, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15519 + }, + { + "epoch": 6.200559328805434, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15520 + }, + { + "epoch": 6.2009588493807435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 15521 + }, + { + "epoch": 6.201358369956052, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15522 + }, + { + "epoch": 6.201757890531362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15523 + }, + { + "epoch": 6.202157411106672, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15524 + }, + { + "epoch": 6.202556931681982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 15525 + }, + { + "epoch": 6.2029564522572915, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15526 + }, + { + "epoch": 6.2033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15527 + }, + { + "epoch": 6.20375549340791, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15528 + }, + { + "epoch": 6.20415501398322, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15529 + }, + { + "epoch": 6.20455453455853, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15530 + }, + { + "epoch": 6.20495405513384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15531 + }, + { + "epoch": 6.205353575709149, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15532 + }, + { + "epoch": 6.205753096284458, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15533 + }, + { + "epoch": 6.206152616859768, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15534 + }, + { + "epoch": 6.206552137435078, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15535 + }, + { + "epoch": 6.206951658010388, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15536 + }, + { + "epoch": 6.207351178585697, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15537 + }, + { + "epoch": 6.207750699161007, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15538 + }, + { + "epoch": 6.208150219736316, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15539 + }, + { + "epoch": 6.208549740311626, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15540 + }, + { + "epoch": 6.208949260886936, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15541 + }, + { + "epoch": 6.2093487814622454, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15542 + }, + { + "epoch": 6.209748302037555, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15543 + }, + { + "epoch": 6.210147822612864, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15544 + }, + { + "epoch": 6.210547343188174, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15545 + }, + { + "epoch": 6.210946863763484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15546 + }, + { + "epoch": 6.2113463843387935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15547 + }, + { + "epoch": 6.211745904914103, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15548 + }, + { + "epoch": 6.212145425489413, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15549 + }, + { + "epoch": 6.212544946064722, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15550 + }, + { + "epoch": 6.212944466640032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15551 + }, + { + "epoch": 6.2133439872153415, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15552 + }, + { + "epoch": 6.213743507790651, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15553 + }, + { + "epoch": 6.214143028365961, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15554 + }, + { + "epoch": 6.214542548941271, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15555 + }, + { + "epoch": 6.21494206951658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15556 + }, + { + "epoch": 6.21534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 15557 + }, + { + "epoch": 6.215741110667199, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 15558 + }, + { + "epoch": 6.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 15559 + }, + { + "epoch": 6.216540151817819, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15560 + }, + { + "epoch": 6.216939672393128, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15561 + }, + { + "epoch": 6.217339192968438, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15562 + }, + { + "epoch": 6.217738713543747, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15563 + }, + { + "epoch": 6.218138234119057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15564 + }, + { + "epoch": 6.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15565 + }, + { + "epoch": 6.218937275269677, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15566 + }, + { + "epoch": 6.219336795844986, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15567 + }, + { + "epoch": 6.2197363164202955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15568 + }, + { + "epoch": 6.220135836995605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15569 + }, + { + "epoch": 6.220535357570915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15570 + }, + { + "epoch": 6.220934878146225, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15571 + }, + { + "epoch": 6.221334398721535, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15572 + }, + { + "epoch": 6.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15573 + }, + { + "epoch": 6.222133439872153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15574 + }, + { + "epoch": 6.222532960447463, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 15575 + }, + { + "epoch": 6.222932481022773, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15576 + }, + { + "epoch": 6.223332001598083, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15577 + }, + { + "epoch": 6.2237315221733915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15578 + }, + { + "epoch": 6.224131042748701, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15579 + }, + { + "epoch": 6.224530563324011, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15580 + }, + { + "epoch": 6.224930083899321, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15581 + }, + { + "epoch": 6.225329604474631, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15582 + }, + { + "epoch": 6.2257291250499405, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15583 + }, + { + "epoch": 6.226128645625249, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15584 + }, + { + "epoch": 6.226528166200559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15585 + }, + { + "epoch": 6.226927686775869, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15586 + }, + { + "epoch": 6.227327207351179, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15587 + }, + { + "epoch": 6.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15588 + }, + { + "epoch": 6.228126248501797, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15589 + }, + { + "epoch": 6.228525769077107, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15590 + }, + { + "epoch": 6.228925289652417, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15591 + }, + { + "epoch": 6.229324810227727, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 15592 + }, + { + "epoch": 6.229724330803037, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15593 + }, + { + "epoch": 6.230123851378346, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15594 + }, + { + "epoch": 6.230523371953655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15595 + }, + { + "epoch": 6.230922892528965, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15596 + }, + { + "epoch": 6.231322413104275, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15597 + }, + { + "epoch": 6.231721933679585, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15598 + }, + { + "epoch": 6.232121454254894, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15599 + }, + { + "epoch": 6.232520974830204, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15600 + }, + { + "epoch": 6.232920495405513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15601 + }, + { + "epoch": 6.233320015980823, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15602 + }, + { + "epoch": 6.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15603 + }, + { + "epoch": 6.234119057131442, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15604 + }, + { + "epoch": 6.234518577706752, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 15605 + }, + { + "epoch": 6.234918098282061, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15606 + }, + { + "epoch": 6.235317618857371, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 15607 + }, + { + "epoch": 6.235717139432681, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15608 + }, + { + "epoch": 6.2361166600079905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15609 + }, + { + "epoch": 6.2365161805833, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15610 + }, + { + "epoch": 6.23691570115861, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15611 + }, + { + "epoch": 6.237315221733919, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15612 + }, + { + "epoch": 6.237714742309229, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15613 + }, + { + "epoch": 6.2381142628845385, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 15614 + }, + { + "epoch": 6.238513783459848, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15615 + }, + { + "epoch": 6.238913304035158, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15616 + }, + { + "epoch": 6.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15617 + }, + { + "epoch": 6.239712345185777, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15618 + }, + { + "epoch": 6.240111865761087, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15619 + }, + { + "epoch": 6.240511386336396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15620 + }, + { + "epoch": 6.240910906911706, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15621 + }, + { + "epoch": 6.241310427487016, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15622 + }, + { + "epoch": 6.241709948062325, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15623 + }, + { + "epoch": 6.242109468637635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15624 + }, + { + "epoch": 6.242508989212944, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15625 + }, + { + "epoch": 6.242908509788254, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15626 + }, + { + "epoch": 6.243308030363564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15627 + }, + { + "epoch": 6.243707550938874, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15628 + }, + { + "epoch": 6.244107071514183, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15629 + }, + { + "epoch": 6.244506592089492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15630 + }, + { + "epoch": 6.244906112664802, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15631 + }, + { + "epoch": 6.245305633240112, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15632 + }, + { + "epoch": 6.245705153815422, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15633 + }, + { + "epoch": 6.246104674390731, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15634 + }, + { + "epoch": 6.2465041949660405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 15635 + }, + { + "epoch": 6.24690371554135, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15636 + }, + { + "epoch": 6.24730323611666, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15637 + }, + { + "epoch": 6.24770275669197, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15638 + }, + { + "epoch": 6.24810227726728, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15639 + }, + { + "epoch": 6.2485017978425885, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15640 + }, + { + "epoch": 6.248901318417898, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15641 + }, + { + "epoch": 6.249300838993208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15642 + }, + { + "epoch": 6.249700359568518, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15643 + }, + { + "epoch": 6.250099880143828, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15644 + }, + { + "epoch": 6.2504994007191375, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15645 + }, + { + "epoch": 6.250898921294446, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15646 + }, + { + "epoch": 6.251298441869756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15647 + }, + { + "epoch": 6.251697962445066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15648 + }, + { + "epoch": 6.252097483020376, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15649 + }, + { + "epoch": 6.2524970035956855, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15650 + }, + { + "epoch": 6.252896524170994, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15651 + }, + { + "epoch": 6.253296044746304, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15652 + }, + { + "epoch": 6.253695565321614, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15653 + }, + { + "epoch": 6.254095085896924, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15654 + }, + { + "epoch": 6.2544946064722335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15655 + }, + { + "epoch": 6.254894127047543, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15656 + }, + { + "epoch": 6.255293647622852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15657 + }, + { + "epoch": 6.255693168198162, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15658 + }, + { + "epoch": 6.256092688773472, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15659 + }, + { + "epoch": 6.256492209348782, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15660 + }, + { + "epoch": 6.256891729924091, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15661 + }, + { + "epoch": 6.257291250499401, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15662 + }, + { + "epoch": 6.25769077107471, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15663 + }, + { + "epoch": 6.25809029165002, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15664 + }, + { + "epoch": 6.25848981222533, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15665 + }, + { + "epoch": 6.258889332800639, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15666 + }, + { + "epoch": 6.259288853375949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 15667 + }, + { + "epoch": 6.259688373951258, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15668 + }, + { + "epoch": 6.260087894526568, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15669 + }, + { + "epoch": 6.260487415101878, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15670 + }, + { + "epoch": 6.2608869356771875, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 15671 + }, + { + "epoch": 6.261286456252497, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15672 + }, + { + "epoch": 6.261685976827807, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15673 + }, + { + "epoch": 6.262085497403116, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15674 + }, + { + "epoch": 6.262485017978426, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15675 + }, + { + "epoch": 6.2628845385537355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15676 + }, + { + "epoch": 6.263284059129045, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15677 + }, + { + "epoch": 6.263683579704355, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 15678 + }, + { + "epoch": 6.264083100279665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 15679 + }, + { + "epoch": 6.264482620854974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15680 + }, + { + "epoch": 6.2648821414302835, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15681 + }, + { + "epoch": 6.265281662005593, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15682 + }, + { + "epoch": 6.265681182580903, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15683 + }, + { + "epoch": 6.266080703156213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15684 + }, + { + "epoch": 6.266480223731522, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15685 + }, + { + "epoch": 6.266879744306832, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15686 + }, + { + "epoch": 6.267279264882141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15687 + }, + { + "epoch": 6.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15688 + }, + { + "epoch": 6.268078306032761, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15689 + }, + { + "epoch": 6.26847782660807, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15690 + }, + { + "epoch": 6.26887734718338, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15691 + }, + { + "epoch": 6.269276867758689, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15692 + }, + { + "epoch": 6.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15693 + }, + { + "epoch": 6.270075908909309, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15694 + }, + { + "epoch": 6.270475429484619, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15695 + }, + { + "epoch": 6.270874950059928, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15696 + }, + { + "epoch": 6.2712744706352375, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15697 + }, + { + "epoch": 6.271673991210547, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15698 + }, + { + "epoch": 6.272073511785857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15699 + }, + { + "epoch": 6.272473032361167, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15700 + }, + { + "epoch": 6.272872552936477, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15701 + }, + { + "epoch": 6.2732720735117855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15702 + }, + { + "epoch": 6.273671594087095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 15703 + }, + { + "epoch": 6.274071114662405, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15704 + }, + { + "epoch": 6.274470635237715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15705 + }, + { + "epoch": 6.274870155813025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15706 + }, + { + "epoch": 6.2752696763883336, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 15707 + }, + { + "epoch": 6.275669196963643, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15708 + }, + { + "epoch": 6.276068717538953, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15709 + }, + { + "epoch": 6.276468238114263, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15710 + }, + { + "epoch": 6.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15711 + }, + { + "epoch": 6.2772672792648825, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15712 + }, + { + "epoch": 6.277666799840191, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15713 + }, + { + "epoch": 6.278066320415501, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15714 + }, + { + "epoch": 6.278465840990811, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15715 + }, + { + "epoch": 6.278865361566121, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15716 + }, + { + "epoch": 6.2792648821414305, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15717 + }, + { + "epoch": 6.27966440271674, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15718 + }, + { + "epoch": 6.280063923292049, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15719 + }, + { + "epoch": 6.280463443867359, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15720 + }, + { + "epoch": 6.280862964442669, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15721 + }, + { + "epoch": 6.281262485017979, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15722 + }, + { + "epoch": 6.281662005593288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15723 + }, + { + "epoch": 6.282061526168597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 15724 + }, + { + "epoch": 6.282461046743907, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15725 + }, + { + "epoch": 6.282860567319217, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15726 + }, + { + "epoch": 6.283260087894527, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15727 + }, + { + "epoch": 6.283659608469836, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15728 + }, + { + "epoch": 6.284059129045146, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15729 + }, + { + "epoch": 6.284458649620455, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15730 + }, + { + "epoch": 6.284858170195765, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 15731 + }, + { + "epoch": 6.285257690771075, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 15732 + }, + { + "epoch": 6.285657211346384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15733 + }, + { + "epoch": 6.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15734 + }, + { + "epoch": 6.286456252497004, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15735 + }, + { + "epoch": 6.286855773072313, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15736 + }, + { + "epoch": 6.287255293647623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15737 + }, + { + "epoch": 6.2876548142229325, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 15738 + }, + { + "epoch": 6.288054334798242, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15739 + }, + { + "epoch": 6.288453855373552, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15740 + }, + { + "epoch": 6.288853375948861, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15741 + }, + { + "epoch": 6.289252896524171, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15742 + }, + { + "epoch": 6.2896524170994805, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15743 + }, + { + "epoch": 6.29005193767479, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15744 + }, + { + "epoch": 6.2904514582501, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15745 + }, + { + "epoch": 6.29085097882541, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15746 + }, + { + "epoch": 6.291250499400719, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15747 + }, + { + "epoch": 6.291650019976029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15748 + }, + { + "epoch": 6.292049540551338, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15749 + }, + { + "epoch": 6.292449061126648, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15750 + }, + { + "epoch": 6.292848581701958, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15751 + }, + { + "epoch": 6.293248102277268, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15752 + }, + { + "epoch": 6.293647622852577, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15753 + }, + { + "epoch": 6.294047143427886, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15754 + }, + { + "epoch": 6.294446664003196, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15755 + }, + { + "epoch": 6.294846184578506, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15756 + }, + { + "epoch": 6.295245705153816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15757 + }, + { + "epoch": 6.295645225729125, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 15758 + }, + { + "epoch": 6.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15759 + }, + { + "epoch": 6.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15760 + }, + { + "epoch": 6.296843787455054, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15761 + }, + { + "epoch": 6.297243308030364, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15762 + }, + { + "epoch": 6.297642828605674, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15763 + }, + { + "epoch": 6.2980423491809825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15764 + }, + { + "epoch": 6.298441869756292, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15765 + }, + { + "epoch": 6.298841390331602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15766 + }, + { + "epoch": 6.299240910906912, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15767 + }, + { + "epoch": 6.299640431482222, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 15768 + }, + { + "epoch": 6.300039952057531, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15769 + }, + { + "epoch": 6.30043947263284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15770 + }, + { + "epoch": 6.30083899320815, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15771 + }, + { + "epoch": 6.30123851378346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15772 + }, + { + "epoch": 6.30163803435877, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15773 + }, + { + "epoch": 6.3020375549340795, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15774 + }, + { + "epoch": 6.302437075509388, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15775 + }, + { + "epoch": 6.302836596084698, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15776 + }, + { + "epoch": 6.303236116660008, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15777 + }, + { + "epoch": 6.303635637235318, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15778 + }, + { + "epoch": 6.3040351578106275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15779 + }, + { + "epoch": 6.304434678385936, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15780 + }, + { + "epoch": 6.304834198961246, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15781 + }, + { + "epoch": 6.305233719536556, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15782 + }, + { + "epoch": 6.305633240111866, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15783 + }, + { + "epoch": 6.3060327606871756, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15784 + }, + { + "epoch": 6.306432281262485, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 15785 + }, + { + "epoch": 6.306831801837794, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15786 + }, + { + "epoch": 6.307231322413104, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15787 + }, + { + "epoch": 6.307630842988414, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15788 + }, + { + "epoch": 6.308030363563724, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15789 + }, + { + "epoch": 6.308429884139033, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15790 + }, + { + "epoch": 6.308829404714343, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15791 + }, + { + "epoch": 6.309228925289652, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15792 + }, + { + "epoch": 6.309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15793 + }, + { + "epoch": 6.310027966440272, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15794 + }, + { + "epoch": 6.310427487015581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15795 + }, + { + "epoch": 6.310827007590891, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15796 + }, + { + "epoch": 6.3112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15797 + }, + { + "epoch": 6.31162604874151, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15798 + }, + { + "epoch": 6.31202556931682, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15799 + }, + { + "epoch": 6.3124250898921295, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 15800 + }, + { + "epoch": 6.312824610467439, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15801 + }, + { + "epoch": 6.313224131042749, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 15802 + }, + { + "epoch": 6.313623651618058, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15803 + }, + { + "epoch": 6.314023172193368, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15804 + }, + { + "epoch": 6.3144226927686775, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15805 + }, + { + "epoch": 6.314822213343987, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15806 + }, + { + "epoch": 6.315221733919297, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15807 + }, + { + "epoch": 6.315621254494607, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15808 + }, + { + "epoch": 6.316020775069916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15809 + }, + { + "epoch": 6.316420295645226, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15810 + }, + { + "epoch": 6.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15811 + }, + { + "epoch": 6.317219336795845, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 15812 + }, + { + "epoch": 6.317618857371155, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15813 + }, + { + "epoch": 6.318018377946464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15814 + }, + { + "epoch": 6.318417898521774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 15815 + }, + { + "epoch": 6.318817419097083, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15816 + }, + { + "epoch": 6.319216939672393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15817 + }, + { + "epoch": 6.319616460247703, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15818 + }, + { + "epoch": 6.320015980823013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15819 + }, + { + "epoch": 6.320415501398322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15820 + }, + { + "epoch": 6.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15821 + }, + { + "epoch": 6.321214542548941, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15822 + }, + { + "epoch": 6.321614063124251, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15823 + }, + { + "epoch": 6.322013583699561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15824 + }, + { + "epoch": 6.322413104274871, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15825 + }, + { + "epoch": 6.3228126248501795, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15826 + }, + { + "epoch": 6.323212145425489, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15827 + }, + { + "epoch": 6.323611666000799, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15828 + }, + { + "epoch": 6.324011186576109, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15829 + }, + { + "epoch": 6.324410707151419, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15830 + }, + { + "epoch": 6.3248102277267275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15831 + }, + { + "epoch": 6.325209748302037, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15832 + }, + { + "epoch": 6.325609268877347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15833 + }, + { + "epoch": 6.326008789452657, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15834 + }, + { + "epoch": 6.326408310027967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15835 + }, + { + "epoch": 6.3268078306032765, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15836 + }, + { + "epoch": 6.327207351178585, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15837 + }, + { + "epoch": 6.327606871753895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15838 + }, + { + "epoch": 6.328006392329205, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15839 + }, + { + "epoch": 6.328405912904515, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15840 + }, + { + "epoch": 6.3288054334798245, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15841 + }, + { + "epoch": 6.329204954055134, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15842 + }, + { + "epoch": 6.329604474630443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15843 + }, + { + "epoch": 6.330003995205753, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 15844 + }, + { + "epoch": 6.330403515781063, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15845 + }, + { + "epoch": 6.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 15846 + }, + { + "epoch": 6.331202556931682, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15847 + }, + { + "epoch": 6.331602077506991, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15848 + }, + { + "epoch": 6.332001598082301, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15849 + }, + { + "epoch": 6.332401118657611, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15850 + }, + { + "epoch": 6.332800639232921, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15851 + }, + { + "epoch": 6.33320015980823, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15852 + }, + { + "epoch": 6.33359968038354, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15853 + }, + { + "epoch": 6.333999200958849, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15854 + }, + { + "epoch": 6.334398721534159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15855 + }, + { + "epoch": 6.334798242109469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15856 + }, + { + "epoch": 6.335197762684778, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15857 + }, + { + "epoch": 6.335597283260088, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15858 + }, + { + "epoch": 6.335996803835398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15859 + }, + { + "epoch": 6.336396324410707, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15860 + }, + { + "epoch": 6.336795844986017, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15861 + }, + { + "epoch": 6.3371953655613265, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15862 + }, + { + "epoch": 6.337594886136636, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15863 + }, + { + "epoch": 6.337994406711946, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15864 + }, + { + "epoch": 6.338393927287255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15865 + }, + { + "epoch": 6.338793447862565, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15866 + }, + { + "epoch": 6.3391929684378745, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15867 + }, + { + "epoch": 6.339592489013184, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15868 + }, + { + "epoch": 6.339992009588494, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15869 + }, + { + "epoch": 6.340391530163803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15870 + }, + { + "epoch": 6.340791050739113, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15871 + }, + { + "epoch": 6.3411905713144225, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15872 + }, + { + "epoch": 6.341590091889732, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15873 + }, + { + "epoch": 6.341989612465042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15874 + }, + { + "epoch": 6.342389133040352, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15875 + }, + { + "epoch": 6.342788653615661, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15876 + }, + { + "epoch": 6.343188174190971, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15877 + }, + { + "epoch": 6.34358769476628, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15878 + }, + { + "epoch": 6.34398721534159, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15879 + }, + { + "epoch": 6.3443867359169, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15880 + }, + { + "epoch": 6.34478625649221, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15881 + }, + { + "epoch": 6.345185777067519, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15882 + }, + { + "epoch": 6.345585297642828, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15883 + }, + { + "epoch": 6.345984818218138, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15884 + }, + { + "epoch": 6.346384338793448, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15885 + }, + { + "epoch": 6.346783859368758, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15886 + }, + { + "epoch": 6.347183379944067, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15887 + }, + { + "epoch": 6.3475829005193765, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15888 + }, + { + "epoch": 6.347982421094686, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15889 + }, + { + "epoch": 6.348381941669996, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15890 + }, + { + "epoch": 6.348781462245306, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 15891 + }, + { + "epoch": 6.349180982820616, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15892 + }, + { + "epoch": 6.3495805033959245, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15893 + }, + { + "epoch": 6.349980023971234, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15894 + }, + { + "epoch": 6.350379544546544, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15895 + }, + { + "epoch": 6.350779065121854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15896 + }, + { + "epoch": 6.351178585697164, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15897 + }, + { + "epoch": 6.351578106272473, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15898 + }, + { + "epoch": 6.351977626847782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15899 + }, + { + "epoch": 6.352377147423092, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15900 + }, + { + "epoch": 6.352776667998402, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15901 + }, + { + "epoch": 6.353176188573712, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 15902 + }, + { + "epoch": 6.3535757091490215, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15903 + }, + { + "epoch": 6.35397522972433, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 15904 + }, + { + "epoch": 6.35437475029964, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15905 + }, + { + "epoch": 6.35477427087495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15906 + }, + { + "epoch": 6.35517379145026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15907 + }, + { + "epoch": 6.3555733120255695, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15908 + }, + { + "epoch": 6.355972832600879, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15909 + }, + { + "epoch": 6.356372353176188, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15910 + }, + { + "epoch": 6.356771873751498, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 15911 + }, + { + "epoch": 6.357171394326808, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15912 + }, + { + "epoch": 6.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15913 + }, + { + "epoch": 6.357970435477427, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15914 + }, + { + "epoch": 6.358369956052737, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15915 + }, + { + "epoch": 6.358769476628046, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15916 + }, + { + "epoch": 6.359168997203356, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15917 + }, + { + "epoch": 6.359568517778666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15918 + }, + { + "epoch": 6.359968038353975, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15919 + }, + { + "epoch": 6.360367558929285, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 15920 + }, + { + "epoch": 6.360767079504594, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15921 + }, + { + "epoch": 6.361166600079904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15922 + }, + { + "epoch": 6.361566120655214, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15923 + }, + { + "epoch": 6.361965641230523, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15924 + }, + { + "epoch": 6.362365161805833, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15925 + }, + { + "epoch": 6.362764682381143, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15926 + }, + { + "epoch": 6.363164202956452, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15927 + }, + { + "epoch": 6.363563723531762, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15928 + }, + { + "epoch": 6.3639632441070715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15929 + }, + { + "epoch": 6.364362764682381, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15930 + }, + { + "epoch": 6.364762285257691, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15931 + }, + { + "epoch": 6.365161805833001, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15932 + }, + { + "epoch": 6.36556132640831, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15933 + }, + { + "epoch": 6.3659608469836195, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15934 + }, + { + "epoch": 6.366360367558929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15935 + }, + { + "epoch": 6.366759888134239, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.297, + "step": 15936 + }, + { + "epoch": 6.367159408709549, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15937 + }, + { + "epoch": 6.367558929284858, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15938 + }, + { + "epoch": 6.367958449860168, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15939 + }, + { + "epoch": 6.368357970435477, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15940 + }, + { + "epoch": 6.368757491010787, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15941 + }, + { + "epoch": 6.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 15942 + }, + { + "epoch": 6.369556532161407, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15943 + }, + { + "epoch": 6.369956052736716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15944 + }, + { + "epoch": 6.370355573312025, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15945 + }, + { + "epoch": 6.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15946 + }, + { + "epoch": 6.371154614462645, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15947 + }, + { + "epoch": 6.371554135037955, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15948 + }, + { + "epoch": 6.3719536556132645, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15949 + }, + { + "epoch": 6.372353176188573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15950 + }, + { + "epoch": 6.372752696763883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15951 + }, + { + "epoch": 6.373152217339193, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15952 + }, + { + "epoch": 6.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 15953 + }, + { + "epoch": 6.373951258489813, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15954 + }, + { + "epoch": 6.3743507790651215, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15955 + }, + { + "epoch": 6.374750299640431, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15956 + }, + { + "epoch": 6.375149820215741, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15957 + }, + { + "epoch": 6.375549340791051, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15958 + }, + { + "epoch": 6.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15959 + }, + { + "epoch": 6.3763483819416695, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15960 + }, + { + "epoch": 6.376747902516979, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 15961 + }, + { + "epoch": 6.377147423092289, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15962 + }, + { + "epoch": 6.377546943667599, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15963 + }, + { + "epoch": 6.377946464242909, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15964 + }, + { + "epoch": 6.3783459848182185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15965 + }, + { + "epoch": 6.378745505393527, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15966 + }, + { + "epoch": 6.379145025968837, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 15967 + }, + { + "epoch": 6.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15968 + }, + { + "epoch": 6.379944067119457, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15969 + }, + { + "epoch": 6.3803435876947665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15970 + }, + { + "epoch": 6.380743108270076, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15971 + }, + { + "epoch": 6.381142628845385, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15972 + }, + { + "epoch": 6.381542149420695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15973 + }, + { + "epoch": 6.381941669996005, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15974 + }, + { + "epoch": 6.3823411905713145, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15975 + }, + { + "epoch": 6.382740711146624, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15976 + }, + { + "epoch": 6.383140231721933, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15977 + }, + { + "epoch": 6.383539752297243, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15978 + }, + { + "epoch": 6.383939272872553, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15979 + }, + { + "epoch": 6.384338793447863, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15980 + }, + { + "epoch": 6.384738314023172, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15981 + }, + { + "epoch": 6.385137834598482, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15982 + }, + { + "epoch": 6.385537355173791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15983 + }, + { + "epoch": 6.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15984 + }, + { + "epoch": 6.386336396324411, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15985 + }, + { + "epoch": 6.38673591689972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15986 + }, + { + "epoch": 6.38713543747503, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15987 + }, + { + "epoch": 6.38753495805034, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 15988 + }, + { + "epoch": 6.387934478625649, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15989 + }, + { + "epoch": 6.388333999200959, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15990 + }, + { + "epoch": 6.3887335197762685, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15991 + }, + { + "epoch": 6.389133040351578, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15992 + }, + { + "epoch": 6.389532560926888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15993 + }, + { + "epoch": 6.389932081502197, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15994 + }, + { + "epoch": 6.390331602077507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15995 + }, + { + "epoch": 6.3907311226528165, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15996 + }, + { + "epoch": 6.391130643228126, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15997 + }, + { + "epoch": 6.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15998 + }, + { + "epoch": 6.391929684378746, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15999 + }, + { + "epoch": 6.392329204954055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16000 + }, + { + "epoch": 6.3927287255293646, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16001 + }, + { + "epoch": 6.393128246104674, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 16002 + }, + { + "epoch": 6.393527766679984, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16003 + }, + { + "epoch": 6.393927287255294, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16004 + }, + { + "epoch": 6.394326807830604, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16005 + }, + { + "epoch": 6.394726328405913, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16006 + }, + { + "epoch": 6.395125848981222, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16007 + }, + { + "epoch": 6.395525369556532, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16008 + }, + { + "epoch": 6.395924890131842, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16009 + }, + { + "epoch": 6.396324410707152, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 16010 + }, + { + "epoch": 6.396723931282461, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16011 + }, + { + "epoch": 6.39712345185777, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16012 + }, + { + "epoch": 6.39752297243308, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16013 + }, + { + "epoch": 6.39792249300839, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16014 + }, + { + "epoch": 6.3983220135837, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16015 + }, + { + "epoch": 6.39872153415901, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16016 + }, + { + "epoch": 6.3991210547343185, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16017 + }, + { + "epoch": 6.399520575309628, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16018 + }, + { + "epoch": 6.399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16019 + }, + { + "epoch": 6.400319616460248, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16020 + }, + { + "epoch": 6.400719137035558, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16021 + }, + { + "epoch": 6.401118657610867, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16022 + }, + { + "epoch": 6.401518178186176, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16023 + }, + { + "epoch": 6.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16024 + }, + { + "epoch": 6.402317219336796, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16025 + }, + { + "epoch": 6.402716739912106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16026 + }, + { + "epoch": 6.4031162604874154, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16027 + }, + { + "epoch": 6.403515781062724, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16028 + }, + { + "epoch": 6.403915301638034, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16029 + }, + { + "epoch": 6.404314822213344, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16030 + }, + { + "epoch": 6.404714342788654, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16031 + }, + { + "epoch": 6.4051138633639635, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16032 + }, + { + "epoch": 6.405513383939273, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16033 + }, + { + "epoch": 6.405912904514582, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16034 + }, + { + "epoch": 6.406312425089892, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16035 + }, + { + "epoch": 6.406711945665202, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16036 + }, + { + "epoch": 6.4071114662405115, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16037 + }, + { + "epoch": 6.407510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16038 + }, + { + "epoch": 6.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16039 + }, + { + "epoch": 6.40831002796644, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16040 + }, + { + "epoch": 6.40870954854175, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16041 + }, + { + "epoch": 6.40910906911706, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16042 + }, + { + "epoch": 6.409508589692369, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16043 + }, + { + "epoch": 6.409908110267679, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16044 + }, + { + "epoch": 6.410307630842988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16045 + }, + { + "epoch": 6.410707151418298, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16046 + }, + { + "epoch": 6.411106671993608, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16047 + }, + { + "epoch": 6.411506192568917, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16048 + }, + { + "epoch": 6.411905713144227, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16049 + }, + { + "epoch": 6.412305233719536, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16050 + }, + { + "epoch": 6.412704754294846, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16051 + }, + { + "epoch": 6.413104274870156, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16052 + }, + { + "epoch": 6.4135037954454654, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16053 + }, + { + "epoch": 6.413903316020775, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16054 + }, + { + "epoch": 6.414302836596085, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16055 + }, + { + "epoch": 6.414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16056 + }, + { + "epoch": 6.415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16057 + }, + { + "epoch": 6.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16058 + }, + { + "epoch": 6.415900918897323, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16059 + }, + { + "epoch": 6.416300439472633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16060 + }, + { + "epoch": 6.416699960047943, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16061 + }, + { + "epoch": 6.417099480623252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 16062 + }, + { + "epoch": 6.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16063 + }, + { + "epoch": 6.417898521773871, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 16064 + }, + { + "epoch": 6.418298042349181, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16065 + }, + { + "epoch": 6.418697562924491, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16066 + }, + { + "epoch": 6.4190970834998, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16067 + }, + { + "epoch": 6.41949660407511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16068 + }, + { + "epoch": 6.419896124650419, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16069 + }, + { + "epoch": 6.420295645225729, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16070 + }, + { + "epoch": 6.420695165801039, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16071 + }, + { + "epoch": 6.421094686376349, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16072 + }, + { + "epoch": 6.421494206951658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16073 + }, + { + "epoch": 6.421893727526967, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16074 + }, + { + "epoch": 6.422293248102277, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16075 + }, + { + "epoch": 6.422692768677587, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16076 + }, + { + "epoch": 6.423092289252897, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16077 + }, + { + "epoch": 6.4234918098282066, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16078 + }, + { + "epoch": 6.4238913304035155, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16079 + }, + { + "epoch": 6.424290850978825, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16080 + }, + { + "epoch": 6.424690371554135, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16081 + }, + { + "epoch": 6.425089892129445, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16082 + }, + { + "epoch": 6.425489412704755, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16083 + }, + { + "epoch": 6.4258889332800635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16084 + }, + { + "epoch": 6.426288453855373, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16085 + }, + { + "epoch": 6.426687974430683, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16086 + }, + { + "epoch": 6.427087495005993, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16087 + }, + { + "epoch": 6.427487015581303, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16088 + }, + { + "epoch": 6.427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16089 + }, + { + "epoch": 6.428286056731921, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16090 + }, + { + "epoch": 6.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16091 + }, + { + "epoch": 6.429085097882541, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16092 + }, + { + "epoch": 6.429484618457851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16093 + }, + { + "epoch": 6.4298841390331605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16094 + }, + { + "epoch": 6.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16095 + }, + { + "epoch": 6.430683180183779, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16096 + }, + { + "epoch": 6.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16097 + }, + { + "epoch": 6.431482221334399, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16098 + }, + { + "epoch": 6.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16099 + }, + { + "epoch": 6.432281262485018, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16100 + }, + { + "epoch": 6.432680783060327, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16101 + }, + { + "epoch": 6.433080303635637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16102 + }, + { + "epoch": 6.433479824210947, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 16103 + }, + { + "epoch": 6.433879344786257, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16104 + }, + { + "epoch": 6.434278865361566, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16105 + }, + { + "epoch": 6.434678385936876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16106 + }, + { + "epoch": 6.435077906512185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16107 + }, + { + "epoch": 6.435477427087495, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16108 + }, + { + "epoch": 6.435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16109 + }, + { + "epoch": 6.436276468238114, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16110 + }, + { + "epoch": 6.436675988813424, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16111 + }, + { + "epoch": 6.437075509388734, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16112 + }, + { + "epoch": 6.437475029964043, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16113 + }, + { + "epoch": 6.437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16114 + }, + { + "epoch": 6.438274071114662, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 16115 + }, + { + "epoch": 6.438673591689972, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16116 + }, + { + "epoch": 6.439073112265282, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 16117 + }, + { + "epoch": 6.439472632840591, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16118 + }, + { + "epoch": 6.439872153415901, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16119 + }, + { + "epoch": 6.4402716739912105, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16120 + }, + { + "epoch": 6.44067119456652, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16121 + }, + { + "epoch": 6.44107071514183, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16122 + }, + { + "epoch": 6.44147023571714, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16123 + }, + { + "epoch": 6.441869756292449, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16124 + }, + { + "epoch": 6.4422692768677585, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16125 + }, + { + "epoch": 6.442668797443068, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16126 + }, + { + "epoch": 6.443068318018378, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16127 + }, + { + "epoch": 6.443467838593688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16128 + }, + { + "epoch": 6.443867359168998, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16129 + }, + { + "epoch": 6.444266879744307, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16130 + }, + { + "epoch": 6.444666400319616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16131 + }, + { + "epoch": 6.445065920894926, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16132 + }, + { + "epoch": 6.445465441470236, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16133 + }, + { + "epoch": 6.445864962045546, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16134 + }, + { + "epoch": 6.446264482620855, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 16135 + }, + { + "epoch": 6.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16136 + }, + { + "epoch": 6.447063523771474, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16137 + }, + { + "epoch": 6.447463044346784, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16138 + }, + { + "epoch": 6.447862564922094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16139 + }, + { + "epoch": 6.448262085497403, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16140 + }, + { + "epoch": 6.448661606072712, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16141 + }, + { + "epoch": 6.449061126648022, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16142 + }, + { + "epoch": 6.449460647223332, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16143 + }, + { + "epoch": 6.449860167798642, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16144 + }, + { + "epoch": 6.450259688373952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16145 + }, + { + "epoch": 6.4506592089492605, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16146 + }, + { + "epoch": 6.45105872952457, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16147 + }, + { + "epoch": 6.45145825009988, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16148 + }, + { + "epoch": 6.45185777067519, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16149 + }, + { + "epoch": 6.4522572912505, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16150 + }, + { + "epoch": 6.452656811825809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16151 + }, + { + "epoch": 6.453056332401118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16152 + }, + { + "epoch": 6.453455852976428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16153 + }, + { + "epoch": 6.453855373551738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16154 + }, + { + "epoch": 6.454254894127048, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16155 + }, + { + "epoch": 6.4546544147023575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16156 + }, + { + "epoch": 6.455053935277666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16157 + }, + { + "epoch": 6.455453455852976, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16158 + }, + { + "epoch": 6.455852976428286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16159 + }, + { + "epoch": 6.456252497003596, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16160 + }, + { + "epoch": 6.4566520175789055, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16161 + }, + { + "epoch": 6.457051538154215, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16162 + }, + { + "epoch": 6.457451058729524, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16163 + }, + { + "epoch": 6.457850579304834, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16164 + }, + { + "epoch": 6.458250099880144, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16165 + }, + { + "epoch": 6.4586496204554535, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16166 + }, + { + "epoch": 6.459049141030763, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16167 + }, + { + "epoch": 6.459448661606073, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16168 + }, + { + "epoch": 6.459848182181382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16169 + }, + { + "epoch": 6.460247702756692, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 16170 + }, + { + "epoch": 6.460647223332002, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16171 + }, + { + "epoch": 6.461046743907311, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16172 + }, + { + "epoch": 6.461446264482621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16173 + }, + { + "epoch": 6.46184578505793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16174 + }, + { + "epoch": 6.46224530563324, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16175 + }, + { + "epoch": 6.46264482620855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16176 + }, + { + "epoch": 6.463044346783859, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16177 + }, + { + "epoch": 6.463443867359169, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16178 + }, + { + "epoch": 6.463843387934479, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16179 + }, + { + "epoch": 6.464242908509788, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16180 + }, + { + "epoch": 6.464642429085098, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16181 + }, + { + "epoch": 6.4650419496604075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16182 + }, + { + "epoch": 6.465441470235717, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16183 + }, + { + "epoch": 6.465840990811027, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16184 + }, + { + "epoch": 6.466240511386337, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16185 + }, + { + "epoch": 6.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16186 + }, + { + "epoch": 6.4670395525369555, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16187 + }, + { + "epoch": 6.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 16188 + }, + { + "epoch": 6.467838593687575, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16189 + }, + { + "epoch": 6.468238114262885, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16190 + }, + { + "epoch": 6.468637634838194, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16191 + }, + { + "epoch": 6.4690371554135035, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16192 + }, + { + "epoch": 6.469436675988813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16193 + }, + { + "epoch": 6.469836196564123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16194 + }, + { + "epoch": 6.470235717139433, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16195 + }, + { + "epoch": 6.470635237714743, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16196 + }, + { + "epoch": 6.471034758290052, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16197 + }, + { + "epoch": 6.471434278865361, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16198 + }, + { + "epoch": 6.471833799440671, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16199 + }, + { + "epoch": 6.472233320015981, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16200 + }, + { + "epoch": 6.472632840591291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.2986, + "step": 16201 + }, + { + "epoch": 6.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16202 + }, + { + "epoch": 6.473431881741909, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16203 + }, + { + "epoch": 6.473831402317219, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16204 + }, + { + "epoch": 6.474230922892529, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16205 + }, + { + "epoch": 6.474630443467839, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16206 + }, + { + "epoch": 6.475029964043149, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16207 + }, + { + "epoch": 6.4754294846184575, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16208 + }, + { + "epoch": 6.475829005193767, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16209 + }, + { + "epoch": 6.476228525769077, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16210 + }, + { + "epoch": 6.476628046344387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16211 + }, + { + "epoch": 6.477027566919697, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16212 + }, + { + "epoch": 6.477427087495006, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16213 + }, + { + "epoch": 6.477826608070315, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16214 + }, + { + "epoch": 6.478226128645625, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16215 + }, + { + "epoch": 6.478625649220935, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16216 + }, + { + "epoch": 6.479025169796245, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16217 + }, + { + "epoch": 6.479424690371554, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16218 + }, + { + "epoch": 6.479824210946864, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16219 + }, + { + "epoch": 6.480223731522173, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16220 + }, + { + "epoch": 6.480623252097483, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16221 + }, + { + "epoch": 6.481022772672793, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16222 + }, + { + "epoch": 6.4814222932481025, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16223 + }, + { + "epoch": 6.481821813823412, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16224 + }, + { + "epoch": 6.482221334398721, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16225 + }, + { + "epoch": 6.482620854974031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16226 + }, + { + "epoch": 6.483020375549341, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16227 + }, + { + "epoch": 6.4834198961246505, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16228 + }, + { + "epoch": 6.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16229 + }, + { + "epoch": 6.484218937275269, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16230 + }, + { + "epoch": 6.484618457850579, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16231 + }, + { + "epoch": 6.485017978425889, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16232 + }, + { + "epoch": 6.485417499001199, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16233 + }, + { + "epoch": 6.485817019576508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16234 + }, + { + "epoch": 6.486216540151818, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16235 + }, + { + "epoch": 6.486616060727127, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16236 + }, + { + "epoch": 6.487015581302437, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16237 + }, + { + "epoch": 6.487415101877747, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16238 + }, + { + "epoch": 6.487814622453056, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16239 + }, + { + "epoch": 6.488214143028366, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16240 + }, + { + "epoch": 6.488613663603676, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16241 + }, + { + "epoch": 6.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16242 + }, + { + "epoch": 6.489412704754295, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16243 + }, + { + "epoch": 6.4898122253296044, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16244 + }, + { + "epoch": 6.490211745904914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16245 + }, + { + "epoch": 6.490611266480224, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16246 + }, + { + "epoch": 6.491010787055533, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16247 + }, + { + "epoch": 6.491410307630843, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16248 + }, + { + "epoch": 6.4918098282061525, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16249 + }, + { + "epoch": 6.492209348781462, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16250 + }, + { + "epoch": 6.492608869356772, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 16251 + }, + { + "epoch": 6.493008389932082, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16252 + }, + { + "epoch": 6.493407910507391, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16253 + }, + { + "epoch": 6.4938074310827005, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16254 + }, + { + "epoch": 6.49420695165801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16255 + }, + { + "epoch": 6.49460647223332, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16256 + }, + { + "epoch": 6.49500599280863, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16257 + }, + { + "epoch": 6.49540551338394, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16258 + }, + { + "epoch": 6.495805033959249, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16259 + }, + { + "epoch": 6.496204554534558, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16260 + }, + { + "epoch": 6.496604075109868, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16261 + }, + { + "epoch": 6.497003595685178, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16262 + }, + { + "epoch": 6.497403116260488, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16263 + }, + { + "epoch": 6.497802636835797, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16264 + }, + { + "epoch": 6.498202157411106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16265 + }, + { + "epoch": 6.498601677986416, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16266 + }, + { + "epoch": 6.499001198561726, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16267 + }, + { + "epoch": 6.499400719137036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 16268 + }, + { + "epoch": 6.4998002397123456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16269 + }, + { + "epoch": 6.5001997602876544, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16270 + }, + { + "epoch": 6.500599280862964, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16271 + }, + { + "epoch": 6.500998801438274, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16272 + }, + { + "epoch": 6.501398322013584, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16273 + }, + { + "epoch": 6.501797842588894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16274 + }, + { + "epoch": 6.502197363164203, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16275 + }, + { + "epoch": 6.502596883739512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16276 + }, + { + "epoch": 6.502996404314822, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16277 + }, + { + "epoch": 6.503395924890132, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16278 + }, + { + "epoch": 6.503795445465442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16279 + }, + { + "epoch": 6.504194966040751, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16280 + }, + { + "epoch": 6.50459448661606, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.2957, + "step": 16281 + }, + { + "epoch": 6.50499400719137, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16282 + }, + { + "epoch": 6.50539352776668, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16283 + }, + { + "epoch": 6.50579304834199, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 16284 + }, + { + "epoch": 6.5061925689172995, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16285 + }, + { + "epoch": 6.506592089492609, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16286 + }, + { + "epoch": 6.506991610067918, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16287 + }, + { + "epoch": 6.507391130643228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16288 + }, + { + "epoch": 6.507790651218538, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16289 + }, + { + "epoch": 6.5081901717938475, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 16290 + }, + { + "epoch": 6.508589692369157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16291 + }, + { + "epoch": 6.508989212944467, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16292 + }, + { + "epoch": 6.509388733519776, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16293 + }, + { + "epoch": 6.509788254095086, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16294 + }, + { + "epoch": 6.5101877746703956, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16295 + }, + { + "epoch": 6.510587295245705, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 16296 + }, + { + "epoch": 6.510986815821015, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16297 + }, + { + "epoch": 6.511386336396324, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16298 + }, + { + "epoch": 6.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16299 + }, + { + "epoch": 6.512185377546944, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16300 + }, + { + "epoch": 6.512584898122253, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16301 + }, + { + "epoch": 6.512984418697563, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16302 + }, + { + "epoch": 6.513383939272872, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16303 + }, + { + "epoch": 6.513783459848182, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16304 + }, + { + "epoch": 6.514182980423492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16305 + }, + { + "epoch": 6.514582500998801, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16306 + }, + { + "epoch": 6.514982021574111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16307 + }, + { + "epoch": 6.515381542149421, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16308 + }, + { + "epoch": 6.515781062724731, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16309 + }, + { + "epoch": 6.51618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16310 + }, + { + "epoch": 6.5165801038753495, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16311 + }, + { + "epoch": 6.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16312 + }, + { + "epoch": 6.517379145025969, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16313 + }, + { + "epoch": 6.517778665601279, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16314 + }, + { + "epoch": 6.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16315 + }, + { + "epoch": 6.5185777067518975, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16316 + }, + { + "epoch": 6.518977227327207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16317 + }, + { + "epoch": 6.519376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 16318 + }, + { + "epoch": 6.519776268477827, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16319 + }, + { + "epoch": 6.520175789053136, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16320 + }, + { + "epoch": 6.520575309628446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16321 + }, + { + "epoch": 6.520974830203755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16322 + }, + { + "epoch": 6.521374350779065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16323 + }, + { + "epoch": 6.521773871354375, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16324 + }, + { + "epoch": 6.522173391929685, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16325 + }, + { + "epoch": 6.5225729125049945, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16326 + }, + { + "epoch": 6.522972433080303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16327 + }, + { + "epoch": 6.523371953655613, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16328 + }, + { + "epoch": 6.523771474230923, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 16329 + }, + { + "epoch": 6.524170994806233, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16330 + }, + { + "epoch": 6.5245705153815425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16331 + }, + { + "epoch": 6.524970035956851, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16332 + }, + { + "epoch": 6.525369556532161, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 16333 + }, + { + "epoch": 6.525769077107471, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16334 + }, + { + "epoch": 6.526168597682781, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16335 + }, + { + "epoch": 6.526568118258091, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16336 + }, + { + "epoch": 6.5269676388333995, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16337 + }, + { + "epoch": 6.527367159408709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16338 + }, + { + "epoch": 6.527766679984019, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16339 + }, + { + "epoch": 6.528166200559329, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16340 + }, + { + "epoch": 6.528565721134639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16341 + }, + { + "epoch": 6.528965241709948, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16342 + }, + { + "epoch": 6.529364762285257, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16343 + }, + { + "epoch": 6.529764282860567, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16344 + }, + { + "epoch": 6.530163803435877, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16345 + }, + { + "epoch": 6.530563324011187, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 16346 + }, + { + "epoch": 6.5309628445864965, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 16347 + }, + { + "epoch": 6.531362365161806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16348 + }, + { + "epoch": 6.531761885737115, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16349 + }, + { + "epoch": 6.532161406312425, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16350 + }, + { + "epoch": 6.532560926887735, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16351 + }, + { + "epoch": 6.5329604474630445, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16352 + }, + { + "epoch": 6.533359968038354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16353 + }, + { + "epoch": 6.533759488613663, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 16354 + }, + { + "epoch": 6.534159009188973, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16355 + }, + { + "epoch": 6.534558529764283, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16356 + }, + { + "epoch": 6.5349580503395925, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16357 + }, + { + "epoch": 6.535357570914902, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 16358 + }, + { + "epoch": 6.535757091490212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16359 + }, + { + "epoch": 6.536156612065521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16360 + }, + { + "epoch": 6.536556132640831, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16361 + }, + { + "epoch": 6.536955653216141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 16362 + }, + { + "epoch": 6.53735517379145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16363 + }, + { + "epoch": 6.53775469436676, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16364 + }, + { + "epoch": 6.53815421494207, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 16365 + }, + { + "epoch": 6.538553735517379, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16366 + }, + { + "epoch": 6.538953256092689, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16367 + }, + { + "epoch": 6.539352776667998, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16368 + }, + { + "epoch": 6.539752297243308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16369 + }, + { + "epoch": 6.540151817818618, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16370 + }, + { + "epoch": 6.540551338393927, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16371 + }, + { + "epoch": 6.540950858969237, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16372 + }, + { + "epoch": 6.5413503795445465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16373 + }, + { + "epoch": 6.541749900119856, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16374 + }, + { + "epoch": 6.542149420695166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16375 + }, + { + "epoch": 6.542548941270476, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16376 + }, + { + "epoch": 6.542948461845785, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16377 + }, + { + "epoch": 6.5433479824210945, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16378 + }, + { + "epoch": 6.543747502996404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16379 + }, + { + "epoch": 6.544147023571714, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16380 + }, + { + "epoch": 6.544546544147024, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16381 + }, + { + "epoch": 6.544946064722334, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16382 + }, + { + "epoch": 6.5453455852976425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16383 + }, + { + "epoch": 6.545745105872952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16384 + }, + { + "epoch": 6.546144626448262, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16385 + }, + { + "epoch": 6.546544147023572, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16386 + }, + { + "epoch": 6.546943667598882, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16387 + }, + { + "epoch": 6.547343188174191, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 16388 + }, + { + "epoch": 6.5477427087495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16389 + }, + { + "epoch": 6.54814222932481, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16390 + }, + { + "epoch": 6.54854174990012, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16391 + }, + { + "epoch": 6.54894127047543, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16392 + }, + { + "epoch": 6.549340791050739, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16393 + }, + { + "epoch": 6.549740311626048, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16394 + }, + { + "epoch": 6.550139832201358, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16395 + }, + { + "epoch": 6.550539352776668, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16396 + }, + { + "epoch": 6.550938873351978, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 16397 + }, + { + "epoch": 6.551338393927288, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16398 + }, + { + "epoch": 6.551737914502597, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16399 + }, + { + "epoch": 6.552137435077906, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16400 + }, + { + "epoch": 6.552536955653216, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16401 + }, + { + "epoch": 6.552936476228526, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16402 + }, + { + "epoch": 6.553335996803836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16403 + }, + { + "epoch": 6.553735517379145, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 16404 + }, + { + "epoch": 6.554135037954454, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16405 + }, + { + "epoch": 6.554534558529764, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16406 + }, + { + "epoch": 6.554934079105074, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16407 + }, + { + "epoch": 6.555333599680384, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16408 + }, + { + "epoch": 6.555733120255693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16409 + }, + { + "epoch": 6.556132640831002, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16410 + }, + { + "epoch": 6.556532161406312, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16411 + }, + { + "epoch": 6.556931681981622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16412 + }, + { + "epoch": 6.557331202556932, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16413 + }, + { + "epoch": 6.5577307231322415, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16414 + }, + { + "epoch": 6.558130243707551, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16415 + }, + { + "epoch": 6.558529764282861, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16416 + }, + { + "epoch": 6.55892928485817, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16417 + }, + { + "epoch": 6.55932880543348, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16418 + }, + { + "epoch": 6.5597283260087895, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16419 + }, + { + "epoch": 6.560127846584099, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16420 + }, + { + "epoch": 6.560527367159409, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16421 + }, + { + "epoch": 6.560926887734718, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16422 + }, + { + "epoch": 6.561326408310028, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16423 + }, + { + "epoch": 6.561725928885338, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16424 + }, + { + "epoch": 6.562125449460647, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16425 + }, + { + "epoch": 6.562524970035957, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16426 + }, + { + "epoch": 6.562924490611266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16427 + }, + { + "epoch": 6.563324011186576, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16428 + }, + { + "epoch": 6.563723531761886, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16429 + }, + { + "epoch": 6.564123052337195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16430 + }, + { + "epoch": 6.564522572912505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16431 + }, + { + "epoch": 6.564922093487815, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16432 + }, + { + "epoch": 6.565321614063124, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16433 + }, + { + "epoch": 6.565721134638434, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16434 + }, + { + "epoch": 6.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16435 + }, + { + "epoch": 6.566520175789053, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16436 + }, + { + "epoch": 6.566919696364363, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16437 + }, + { + "epoch": 6.567319216939673, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16438 + }, + { + "epoch": 6.567718737514982, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 16439 + }, + { + "epoch": 6.5681182580902915, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16440 + }, + { + "epoch": 6.568517778665601, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16441 + }, + { + "epoch": 6.568917299240911, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16442 + }, + { + "epoch": 6.569316819816221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16443 + }, + { + "epoch": 6.56971634039153, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16444 + }, + { + "epoch": 6.5701158609668395, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16445 + }, + { + "epoch": 6.570515381542149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16446 + }, + { + "epoch": 6.570914902117459, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 16447 + }, + { + "epoch": 6.571314422692769, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16448 + }, + { + "epoch": 6.571713943268079, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16449 + }, + { + "epoch": 6.572113463843388, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16450 + }, + { + "epoch": 6.572512984418697, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16451 + }, + { + "epoch": 6.572912504994007, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16452 + }, + { + "epoch": 6.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16453 + }, + { + "epoch": 6.573711546144627, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16454 + }, + { + "epoch": 6.5741110667199365, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16455 + }, + { + "epoch": 6.574510587295245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16456 + }, + { + "epoch": 6.574910107870555, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16457 + }, + { + "epoch": 6.575309628445865, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16458 + }, + { + "epoch": 6.575709149021175, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16459 + }, + { + "epoch": 6.5761086695964845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16460 + }, + { + "epoch": 6.5765081901717934, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16461 + }, + { + "epoch": 6.576907710747103, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16462 + }, + { + "epoch": 6.577307231322413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16463 + }, + { + "epoch": 6.577706751897723, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16464 + }, + { + "epoch": 6.578106272473033, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16465 + }, + { + "epoch": 6.578505793048342, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16466 + }, + { + "epoch": 6.578905313623651, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16467 + }, + { + "epoch": 6.579304834198961, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16468 + }, + { + "epoch": 6.579704354774271, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16469 + }, + { + "epoch": 6.580103875349581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16470 + }, + { + "epoch": 6.58050339592489, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16471 + }, + { + "epoch": 6.5809029165002, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16472 + }, + { + "epoch": 6.581302437075509, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16473 + }, + { + "epoch": 6.581701957650819, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16474 + }, + { + "epoch": 6.582101478226129, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16475 + }, + { + "epoch": 6.5825009988014385, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16476 + }, + { + "epoch": 6.582900519376748, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16477 + }, + { + "epoch": 6.583300039952057, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16478 + }, + { + "epoch": 6.583699560527367, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16479 + }, + { + "epoch": 6.584099081102677, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16480 + }, + { + "epoch": 6.5844986016779865, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16481 + }, + { + "epoch": 6.584898122253296, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16482 + }, + { + "epoch": 6.585297642828605, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16483 + }, + { + "epoch": 6.585697163403915, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16484 + }, + { + "epoch": 6.586096683979225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16485 + }, + { + "epoch": 6.5864962045545346, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16486 + }, + { + "epoch": 6.586895725129844, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16487 + }, + { + "epoch": 6.587295245705154, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16488 + }, + { + "epoch": 6.587694766280464, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16489 + }, + { + "epoch": 6.588094286855773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16490 + }, + { + "epoch": 6.588493807431083, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16491 + }, + { + "epoch": 6.588893328006392, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16492 + }, + { + "epoch": 6.589292848581702, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16493 + }, + { + "epoch": 6.589692369157012, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16494 + }, + { + "epoch": 6.590091889732321, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16495 + }, + { + "epoch": 6.590491410307631, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16496 + }, + { + "epoch": 6.59089093088294, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16497 + }, + { + "epoch": 6.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16498 + }, + { + "epoch": 6.59168997203356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16499 + }, + { + "epoch": 6.592089492608869, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16500 + }, + { + "epoch": 6.592489013184179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16501 + }, + { + "epoch": 6.5928885337594885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16502 + }, + { + "epoch": 6.593288054334798, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16503 + }, + { + "epoch": 6.593687574910108, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16504 + }, + { + "epoch": 6.594087095485418, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16505 + }, + { + "epoch": 6.594486616060728, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16506 + }, + { + "epoch": 6.5948861366360365, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16507 + }, + { + "epoch": 6.595285657211346, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16508 + }, + { + "epoch": 6.595685177786656, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16509 + }, + { + "epoch": 6.596084698361966, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16510 + }, + { + "epoch": 6.596484218937276, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16511 + }, + { + "epoch": 6.5968837395125846, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16512 + }, + { + "epoch": 6.597283260087894, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16513 + }, + { + "epoch": 6.597682780663204, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16514 + }, + { + "epoch": 6.598082301238514, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16515 + }, + { + "epoch": 6.598481821813824, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 16516 + }, + { + "epoch": 6.598881342389133, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 16517 + }, + { + "epoch": 6.599280862964442, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16518 + }, + { + "epoch": 6.599680383539752, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16519 + }, + { + "epoch": 6.600079904115062, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16520 + }, + { + "epoch": 6.600479424690372, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16521 + }, + { + "epoch": 6.6008789452656815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16522 + }, + { + "epoch": 6.60127846584099, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16523 + }, + { + "epoch": 6.6016779864163, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16524 + }, + { + "epoch": 6.60207750699161, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16525 + }, + { + "epoch": 6.60247702756692, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16526 + }, + { + "epoch": 6.60287654814223, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16527 + }, + { + "epoch": 6.603276068717539, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16528 + }, + { + "epoch": 6.603675589292848, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16529 + }, + { + "epoch": 6.604075109868158, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16530 + }, + { + "epoch": 6.604474630443468, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16531 + }, + { + "epoch": 6.604874151018778, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16532 + }, + { + "epoch": 6.605273671594087, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16533 + }, + { + "epoch": 6.605673192169396, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16534 + }, + { + "epoch": 6.606072712744706, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16535 + }, + { + "epoch": 6.606472233320016, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16536 + }, + { + "epoch": 6.606871753895326, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16537 + }, + { + "epoch": 6.6072712744706354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16538 + }, + { + "epoch": 6.607670795045945, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16539 + }, + { + "epoch": 6.608070315621254, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16540 + }, + { + "epoch": 6.608469836196564, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16541 + }, + { + "epoch": 6.608869356771874, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 16542 + }, + { + "epoch": 6.6092688773471835, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16543 + }, + { + "epoch": 6.609668397922493, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16544 + }, + { + "epoch": 6.610067918497803, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16545 + }, + { + "epoch": 6.610467439073112, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16546 + }, + { + "epoch": 6.610866959648422, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16547 + }, + { + "epoch": 6.6112664802237315, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16548 + }, + { + "epoch": 6.611666000799041, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16549 + }, + { + "epoch": 6.612065521374351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16550 + }, + { + "epoch": 6.61246504194966, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16551 + }, + { + "epoch": 6.61286456252497, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16552 + }, + { + "epoch": 6.61326408310028, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16553 + }, + { + "epoch": 6.613663603675589, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16554 + }, + { + "epoch": 6.614063124250899, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16555 + }, + { + "epoch": 6.614462644826209, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16556 + }, + { + "epoch": 6.614862165401518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16557 + }, + { + "epoch": 6.615261685976828, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16558 + }, + { + "epoch": 6.615661206552137, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16559 + }, + { + "epoch": 6.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16560 + }, + { + "epoch": 6.616460247702757, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16561 + }, + { + "epoch": 6.616859768278067, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16562 + }, + { + "epoch": 6.617259288853376, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16563 + }, + { + "epoch": 6.6176588094286855, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16564 + }, + { + "epoch": 6.618058330003995, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16565 + }, + { + "epoch": 6.618457850579305, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16566 + }, + { + "epoch": 6.618857371154615, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16567 + }, + { + "epoch": 6.619256891729924, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16568 + }, + { + "epoch": 6.6196564123052335, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16569 + }, + { + "epoch": 6.620055932880543, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16570 + }, + { + "epoch": 6.620455453455853, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16571 + }, + { + "epoch": 6.620854974031163, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16572 + }, + { + "epoch": 6.621254494606472, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 16573 + }, + { + "epoch": 6.6216540151817815, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16574 + }, + { + "epoch": 6.622053535757091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 16575 + }, + { + "epoch": 6.622453056332401, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16576 + }, + { + "epoch": 6.622852576907711, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16577 + }, + { + "epoch": 6.623252097483021, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16578 + }, + { + "epoch": 6.6236516180583305, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16579 + }, + { + "epoch": 6.624051138633639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16580 + }, + { + "epoch": 6.624450659208949, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16581 + }, + { + "epoch": 6.624850179784259, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16582 + }, + { + "epoch": 6.625249700359569, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16583 + }, + { + "epoch": 6.6256492209348785, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16584 + }, + { + "epoch": 6.626048741510187, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16585 + }, + { + "epoch": 6.626448262085497, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16586 + }, + { + "epoch": 6.626847782660807, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16587 + }, + { + "epoch": 6.627247303236117, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16588 + }, + { + "epoch": 6.627646823811427, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16589 + }, + { + "epoch": 6.6280463443867355, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16590 + }, + { + "epoch": 6.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16591 + }, + { + "epoch": 6.628845385537355, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16592 + }, + { + "epoch": 6.629244906112665, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16593 + }, + { + "epoch": 6.629644426687975, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16594 + }, + { + "epoch": 6.630043947263284, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16595 + }, + { + "epoch": 6.630443467838594, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16596 + }, + { + "epoch": 6.630842988413903, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16597 + }, + { + "epoch": 6.631242508989213, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16598 + }, + { + "epoch": 6.631642029564523, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 16599 + }, + { + "epoch": 6.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16600 + }, + { + "epoch": 6.632441070715142, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16601 + }, + { + "epoch": 6.632840591290451, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16602 + }, + { + "epoch": 6.633240111865761, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16603 + }, + { + "epoch": 6.633639632441071, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16604 + }, + { + "epoch": 6.6340391530163805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16605 + }, + { + "epoch": 6.63443867359169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16606 + }, + { + "epoch": 6.634838194166999, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16607 + }, + { + "epoch": 6.635237714742309, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16608 + }, + { + "epoch": 6.635637235317619, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 16609 + }, + { + "epoch": 6.6360367558929285, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16610 + }, + { + "epoch": 6.636436276468238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16611 + }, + { + "epoch": 6.636835797043548, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16612 + }, + { + "epoch": 6.637235317618857, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16613 + }, + { + "epoch": 6.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16614 + }, + { + "epoch": 6.638034358769477, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16615 + }, + { + "epoch": 6.638433879344786, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16616 + }, + { + "epoch": 6.638833399920096, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 16617 + }, + { + "epoch": 6.639232920495406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16618 + }, + { + "epoch": 6.639632441070715, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16619 + }, + { + "epoch": 6.640031961646025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16620 + }, + { + "epoch": 6.640431482221334, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16621 + }, + { + "epoch": 6.640831002796644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16622 + }, + { + "epoch": 6.641230523371954, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16623 + }, + { + "epoch": 6.641630043947263, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16624 + }, + { + "epoch": 6.642029564522573, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16625 + }, + { + "epoch": 6.642429085097882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16626 + }, + { + "epoch": 6.642828605673192, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16627 + }, + { + "epoch": 6.643228126248502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16628 + }, + { + "epoch": 6.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16629 + }, + { + "epoch": 6.644027167399121, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16630 + }, + { + "epoch": 6.6444266879744305, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16631 + }, + { + "epoch": 6.64482620854974, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16632 + }, + { + "epoch": 6.64522572912505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16633 + }, + { + "epoch": 6.64562524970036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16634 + }, + { + "epoch": 6.64602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16635 + }, + { + "epoch": 6.6464242908509785, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16636 + }, + { + "epoch": 6.646823811426288, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16637 + }, + { + "epoch": 6.647223332001598, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16638 + }, + { + "epoch": 6.647622852576908, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16639 + }, + { + "epoch": 6.648022373152218, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16640 + }, + { + "epoch": 6.648421893727527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16641 + }, + { + "epoch": 6.648821414302836, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16642 + }, + { + "epoch": 6.649220934878146, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16643 + }, + { + "epoch": 6.649620455453456, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16644 + }, + { + "epoch": 6.650019976028766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16645 + }, + { + "epoch": 6.6504194966040755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16646 + }, + { + "epoch": 6.650819017179384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16647 + }, + { + "epoch": 6.651218537754694, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16648 + }, + { + "epoch": 6.651618058330004, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16649 + }, + { + "epoch": 6.652017578905314, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16650 + }, + { + "epoch": 6.6524170994806235, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16651 + }, + { + "epoch": 6.652816620055933, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16652 + }, + { + "epoch": 6.653216140631242, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16653 + }, + { + "epoch": 6.653615661206552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16654 + }, + { + "epoch": 6.654015181781862, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16655 + }, + { + "epoch": 6.654414702357172, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16656 + }, + { + "epoch": 6.654814222932481, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16657 + }, + { + "epoch": 6.65521374350779, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16658 + }, + { + "epoch": 6.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16659 + }, + { + "epoch": 6.65601278465841, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16660 + }, + { + "epoch": 6.65641230523372, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16661 + }, + { + "epoch": 6.656811825809029, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16662 + }, + { + "epoch": 6.657211346384338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16663 + }, + { + "epoch": 6.657610866959648, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16664 + }, + { + "epoch": 6.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16665 + }, + { + "epoch": 6.658409908110268, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16666 + }, + { + "epoch": 6.6588094286855775, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16667 + }, + { + "epoch": 6.659208949260887, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16668 + }, + { + "epoch": 6.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16669 + }, + { + "epoch": 6.660007990411506, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16670 + }, + { + "epoch": 6.660407510986816, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16671 + }, + { + "epoch": 6.6608070315621255, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16672 + }, + { + "epoch": 6.661206552137435, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16673 + }, + { + "epoch": 6.661606072712745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16674 + }, + { + "epoch": 6.662005593288054, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16675 + }, + { + "epoch": 6.662405113863364, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16676 + }, + { + "epoch": 6.6628046344386735, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16677 + }, + { + "epoch": 6.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16678 + }, + { + "epoch": 6.663603675589293, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16679 + }, + { + "epoch": 6.664003196164602, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16680 + }, + { + "epoch": 6.664402716739912, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16681 + }, + { + "epoch": 6.664802237315222, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16682 + }, + { + "epoch": 6.665201757890531, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16683 + }, + { + "epoch": 6.665601278465841, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16684 + }, + { + "epoch": 6.666000799041151, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16685 + }, + { + "epoch": 6.666400319616461, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16686 + }, + { + "epoch": 6.66679984019177, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16687 + }, + { + "epoch": 6.667199360767079, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16688 + }, + { + "epoch": 6.667598881342389, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16689 + }, + { + "epoch": 6.667998401917699, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16690 + }, + { + "epoch": 6.668397922493009, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16691 + }, + { + "epoch": 6.668797443068318, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16692 + }, + { + "epoch": 6.6691969636436275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16693 + }, + { + "epoch": 6.669596484218937, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16694 + }, + { + "epoch": 6.669996004794247, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16695 + }, + { + "epoch": 6.670395525369557, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16696 + }, + { + "epoch": 6.670795045944866, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16697 + }, + { + "epoch": 6.6711945665201755, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16698 + }, + { + "epoch": 6.671594087095485, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16699 + }, + { + "epoch": 6.671993607670795, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16700 + }, + { + "epoch": 6.672393128246105, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16701 + }, + { + "epoch": 6.672792648821415, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16702 + }, + { + "epoch": 6.6731921693967235, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16703 + }, + { + "epoch": 6.673591689972033, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16704 + }, + { + "epoch": 6.673991210547343, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16705 + }, + { + "epoch": 6.674390731122653, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 16706 + }, + { + "epoch": 6.674790251697963, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16707 + }, + { + "epoch": 6.6751897722732725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16708 + }, + { + "epoch": 6.675589292848581, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16709 + }, + { + "epoch": 6.675988813423891, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16710 + }, + { + "epoch": 6.676388333999201, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16711 + }, + { + "epoch": 6.676787854574511, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16712 + }, + { + "epoch": 6.6771873751498205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16713 + }, + { + "epoch": 6.677586895725129, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16714 + }, + { + "epoch": 6.677986416300439, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16715 + }, + { + "epoch": 6.678385936875749, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16716 + }, + { + "epoch": 6.678785457451059, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16717 + }, + { + "epoch": 6.679184978026369, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16718 + }, + { + "epoch": 6.679584498601678, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16719 + }, + { + "epoch": 6.679984019176987, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 16720 + }, + { + "epoch": 6.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16721 + }, + { + "epoch": 6.680783060327607, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16722 + }, + { + "epoch": 6.681182580902917, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16723 + }, + { + "epoch": 6.681582101478226, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16724 + }, + { + "epoch": 6.681981622053536, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16725 + }, + { + "epoch": 6.682381142628845, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16726 + }, + { + "epoch": 6.682780663204155, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16727 + }, + { + "epoch": 6.683180183779465, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16728 + }, + { + "epoch": 6.683579704354774, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16729 + }, + { + "epoch": 6.683979224930084, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16730 + }, + { + "epoch": 6.684378745505393, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16731 + }, + { + "epoch": 6.684778266080703, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16732 + }, + { + "epoch": 6.685177786656013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16733 + }, + { + "epoch": 6.6855773072313225, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16734 + }, + { + "epoch": 6.685976827806632, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16735 + }, + { + "epoch": 6.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16736 + }, + { + "epoch": 6.686775868957251, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16737 + }, + { + "epoch": 6.687175389532561, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 16738 + }, + { + "epoch": 6.6875749101078705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16739 + }, + { + "epoch": 6.68797443068318, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16740 + }, + { + "epoch": 6.68837395125849, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16741 + }, + { + "epoch": 6.6887734718338, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16742 + }, + { + "epoch": 6.689172992409109, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16743 + }, + { + "epoch": 6.689572512984419, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16744 + }, + { + "epoch": 6.689972033559728, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16745 + }, + { + "epoch": 6.690371554135038, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16746 + }, + { + "epoch": 6.690771074710348, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16747 + }, + { + "epoch": 6.691170595285657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16748 + }, + { + "epoch": 6.691570115860967, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16749 + }, + { + "epoch": 6.691969636436276, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16750 + }, + { + "epoch": 6.692369157011586, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 16751 + }, + { + "epoch": 6.692768677586896, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 16752 + }, + { + "epoch": 6.693168198162205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16753 + }, + { + "epoch": 6.693567718737515, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16754 + }, + { + "epoch": 6.6939672393128244, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16755 + }, + { + "epoch": 6.694366759888134, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16756 + }, + { + "epoch": 6.694766280463444, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16757 + }, + { + "epoch": 6.695165801038754, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16758 + }, + { + "epoch": 6.695565321614064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16759 + }, + { + "epoch": 6.6959648421893725, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16760 + }, + { + "epoch": 6.696364362764682, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16761 + }, + { + "epoch": 6.696763883339992, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16762 + }, + { + "epoch": 6.697163403915302, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16763 + }, + { + "epoch": 6.697562924490612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16764 + }, + { + "epoch": 6.6979624450659205, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16765 + }, + { + "epoch": 6.69836196564123, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16766 + }, + { + "epoch": 6.69876148621654, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16767 + }, + { + "epoch": 6.69916100679185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16768 + }, + { + "epoch": 6.69956052736716, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16769 + }, + { + "epoch": 6.699960047942469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16770 + }, + { + "epoch": 6.700359568517778, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16771 + }, + { + "epoch": 6.700759089093088, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16772 + }, + { + "epoch": 6.701158609668398, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16773 + }, + { + "epoch": 6.701558130243708, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16774 + }, + { + "epoch": 6.7019576508190175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16775 + }, + { + "epoch": 6.702357171394327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16776 + }, + { + "epoch": 6.702756691969636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16777 + }, + { + "epoch": 6.703156212544946, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16778 + }, + { + "epoch": 6.703555733120256, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16779 + }, + { + "epoch": 6.7039552536955656, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16780 + }, + { + "epoch": 6.704354774270875, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16781 + }, + { + "epoch": 6.704754294846184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16782 + }, + { + "epoch": 6.705153815421494, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16783 + }, + { + "epoch": 6.705553335996804, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16784 + }, + { + "epoch": 6.705952856572114, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16785 + }, + { + "epoch": 6.706352377147423, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16786 + }, + { + "epoch": 6.706751897722732, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16787 + }, + { + "epoch": 6.707151418298042, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16788 + }, + { + "epoch": 6.707550938873352, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16789 + }, + { + "epoch": 6.707950459448662, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16790 + }, + { + "epoch": 6.708349980023971, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16791 + }, + { + "epoch": 6.708749500599281, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16792 + }, + { + "epoch": 6.70914902117459, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16793 + }, + { + "epoch": 6.7095485417499, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16794 + }, + { + "epoch": 6.70994806232521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16795 + }, + { + "epoch": 6.7103475829005195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16796 + }, + { + "epoch": 6.710747103475829, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16797 + }, + { + "epoch": 6.711146624051139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16798 + }, + { + "epoch": 6.711546144626448, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16799 + }, + { + "epoch": 6.711945665201758, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16800 + }, + { + "epoch": 6.7123451857770675, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16801 + }, + { + "epoch": 6.712744706352377, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.2968, + "step": 16802 + }, + { + "epoch": 6.713144226927687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16803 + }, + { + "epoch": 6.713543747502996, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16804 + }, + { + "epoch": 6.713943268078306, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16805 + }, + { + "epoch": 6.714342788653616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16806 + }, + { + "epoch": 6.714742309228925, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16807 + }, + { + "epoch": 6.715141829804235, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 16808 + }, + { + "epoch": 6.715541350379545, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16809 + }, + { + "epoch": 6.715940870954854, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 16810 + }, + { + "epoch": 6.716340391530164, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16811 + }, + { + "epoch": 6.716739912105473, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16812 + }, + { + "epoch": 6.717139432680783, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16813 + }, + { + "epoch": 6.717538953256093, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16814 + }, + { + "epoch": 6.717938473831403, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16815 + }, + { + "epoch": 6.718337994406712, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16816 + }, + { + "epoch": 6.718737514982021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16817 + }, + { + "epoch": 6.719137035557331, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16818 + }, + { + "epoch": 6.719536556132641, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16819 + }, + { + "epoch": 6.719936076707951, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16820 + }, + { + "epoch": 6.72033559728326, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16821 + }, + { + "epoch": 6.7207351178585695, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 16822 + }, + { + "epoch": 6.721134638433879, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16823 + }, + { + "epoch": 6.721534159009189, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 16824 + }, + { + "epoch": 6.721933679584499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16825 + }, + { + "epoch": 6.722333200159808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16826 + }, + { + "epoch": 6.7227327207351175, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16827 + }, + { + "epoch": 6.723132241310427, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16828 + }, + { + "epoch": 6.723531761885737, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16829 + }, + { + "epoch": 6.723931282461047, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16830 + }, + { + "epoch": 6.724330803036357, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16831 + }, + { + "epoch": 6.7247303236116664, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16832 + }, + { + "epoch": 6.725129844186975, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16833 + }, + { + "epoch": 6.725529364762285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16834 + }, + { + "epoch": 6.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16835 + }, + { + "epoch": 6.726328405912905, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16836 + }, + { + "epoch": 6.7267279264882145, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16837 + }, + { + "epoch": 6.727127447063523, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16838 + }, + { + "epoch": 6.727526967638833, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16839 + }, + { + "epoch": 6.727926488214143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16840 + }, + { + "epoch": 6.728326008789453, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16841 + }, + { + "epoch": 6.7287255293647625, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16842 + }, + { + "epoch": 6.729125049940071, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16843 + }, + { + "epoch": 6.729524570515381, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16844 + }, + { + "epoch": 6.729924091090691, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16845 + }, + { + "epoch": 6.730323611666001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16846 + }, + { + "epoch": 6.730723132241311, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16847 + }, + { + "epoch": 6.73112265281662, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16848 + }, + { + "epoch": 6.73152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16849 + }, + { + "epoch": 6.731921693967239, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16850 + }, + { + "epoch": 6.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16851 + }, + { + "epoch": 6.732720735117859, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16852 + }, + { + "epoch": 6.733120255693168, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16853 + }, + { + "epoch": 6.733519776268478, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16854 + }, + { + "epoch": 6.733919296843787, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16855 + }, + { + "epoch": 6.734318817419097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16856 + }, + { + "epoch": 6.734718337994407, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16857 + }, + { + "epoch": 6.7351178585697165, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16858 + }, + { + "epoch": 6.735517379145026, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16859 + }, + { + "epoch": 6.735916899720335, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16860 + }, + { + "epoch": 6.736316420295645, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16861 + }, + { + "epoch": 6.736715940870955, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16862 + }, + { + "epoch": 6.7371154614462645, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 16863 + }, + { + "epoch": 6.737514982021574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16864 + }, + { + "epoch": 6.737914502596884, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16865 + }, + { + "epoch": 6.738314023172194, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16866 + }, + { + "epoch": 6.738713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16867 + }, + { + "epoch": 6.7391130643228125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16868 + }, + { + "epoch": 6.739512584898122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16869 + }, + { + "epoch": 6.739912105473432, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16870 + }, + { + "epoch": 6.740311626048742, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16871 + }, + { + "epoch": 6.740711146624051, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16872 + }, + { + "epoch": 6.741110667199361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16873 + }, + { + "epoch": 6.74151018777467, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16874 + }, + { + "epoch": 6.74190970834998, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16875 + }, + { + "epoch": 6.74230922892529, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16876 + }, + { + "epoch": 6.742708749500599, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16877 + }, + { + "epoch": 6.743108270075909, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16878 + }, + { + "epoch": 6.743507790651218, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16879 + }, + { + "epoch": 6.743907311226528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16880 + }, + { + "epoch": 6.744306831801838, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16881 + }, + { + "epoch": 6.744706352377148, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16882 + }, + { + "epoch": 6.745105872952457, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16883 + }, + { + "epoch": 6.7455053935277665, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16884 + }, + { + "epoch": 6.745904914103076, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16885 + }, + { + "epoch": 6.746304434678386, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 16886 + }, + { + "epoch": 6.746703955253696, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16887 + }, + { + "epoch": 6.747103475829006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16888 + }, + { + "epoch": 6.7475029964043145, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16889 + }, + { + "epoch": 6.747902516979624, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16890 + }, + { + "epoch": 6.748302037554934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16891 + }, + { + "epoch": 6.748701558130244, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16892 + }, + { + "epoch": 6.749101078705554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16893 + }, + { + "epoch": 6.7495005992808625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16894 + }, + { + "epoch": 6.749900119856172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16895 + }, + { + "epoch": 6.750299640431482, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16896 + }, + { + "epoch": 6.750699161006792, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16897 + }, + { + "epoch": 6.751098681582102, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16898 + }, + { + "epoch": 6.7514982021574115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.296, + "step": 16899 + }, + { + "epoch": 6.75189772273272, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16900 + }, + { + "epoch": 6.75229724330803, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16901 + }, + { + "epoch": 6.75269676388334, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16902 + }, + { + "epoch": 6.75309628445865, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16903 + }, + { + "epoch": 6.7534958050339595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16904 + }, + { + "epoch": 6.753895325609269, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16905 + }, + { + "epoch": 6.754294846184578, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16906 + }, + { + "epoch": 6.754694366759888, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 16907 + }, + { + "epoch": 6.755093887335198, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16908 + }, + { + "epoch": 6.755493407910508, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16909 + }, + { + "epoch": 6.755892928485817, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 16910 + }, + { + "epoch": 6.756292449061126, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16911 + }, + { + "epoch": 6.756691969636436, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16912 + }, + { + "epoch": 6.757091490211746, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16913 + }, + { + "epoch": 6.757491010787056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16914 + }, + { + "epoch": 6.757890531362365, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16915 + }, + { + "epoch": 6.758290051937674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16916 + }, + { + "epoch": 6.758689572512984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16917 + }, + { + "epoch": 6.759089093088294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16918 + }, + { + "epoch": 6.759488613663604, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16919 + }, + { + "epoch": 6.759888134238913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16920 + }, + { + "epoch": 6.760287654814223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16921 + }, + { + "epoch": 6.760687175389533, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16922 + }, + { + "epoch": 6.761086695964842, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16923 + }, + { + "epoch": 6.761486216540152, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16924 + }, + { + "epoch": 6.7618857371154615, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16925 + }, + { + "epoch": 6.762285257690771, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16926 + }, + { + "epoch": 6.762684778266081, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16927 + }, + { + "epoch": 6.76308429884139, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16928 + }, + { + "epoch": 6.7634838194167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16929 + }, + { + "epoch": 6.7638833399920095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16930 + }, + { + "epoch": 6.764282860567319, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16931 + }, + { + "epoch": 6.764682381142629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16932 + }, + { + "epoch": 6.765081901717938, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16933 + }, + { + "epoch": 6.765481422293248, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16934 + }, + { + "epoch": 6.765880942868558, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16935 + }, + { + "epoch": 6.766280463443867, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16936 + }, + { + "epoch": 6.766679984019177, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16937 + }, + { + "epoch": 6.767079504594487, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16938 + }, + { + "epoch": 6.767479025169797, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16939 + }, + { + "epoch": 6.767878545745106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16940 + }, + { + "epoch": 6.768278066320415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16941 + }, + { + "epoch": 6.768677586895725, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16942 + }, + { + "epoch": 6.769077107471035, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16943 + }, + { + "epoch": 6.769476628046345, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16944 + }, + { + "epoch": 6.769876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16945 + }, + { + "epoch": 6.770275669196963, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16946 + }, + { + "epoch": 6.770675189772273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16947 + }, + { + "epoch": 6.771074710347583, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16948 + }, + { + "epoch": 6.771474230922893, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16949 + }, + { + "epoch": 6.771873751498202, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16950 + }, + { + "epoch": 6.7722732720735115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16951 + }, + { + "epoch": 6.772672792648821, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16952 + }, + { + "epoch": 6.773072313224131, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16953 + }, + { + "epoch": 6.773471833799441, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16954 + }, + { + "epoch": 6.773871354374751, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16955 + }, + { + "epoch": 6.77427087495006, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16956 + }, + { + "epoch": 6.774670395525369, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16957 + }, + { + "epoch": 6.775069916100679, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16958 + }, + { + "epoch": 6.775469436675989, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16959 + }, + { + "epoch": 6.775868957251299, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16960 + }, + { + "epoch": 6.7762684778266085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16961 + }, + { + "epoch": 6.776667998401917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16962 + }, + { + "epoch": 6.777067518977227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16963 + }, + { + "epoch": 6.777467039552537, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 16964 + }, + { + "epoch": 6.777866560127847, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16965 + }, + { + "epoch": 6.7782660807031565, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16966 + }, + { + "epoch": 6.778665601278465, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 16967 + }, + { + "epoch": 6.779065121853775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16968 + }, + { + "epoch": 6.779464642429085, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16969 + }, + { + "epoch": 6.779864163004395, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 16970 + }, + { + "epoch": 6.7802636835797045, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16971 + }, + { + "epoch": 6.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16972 + }, + { + "epoch": 6.781062724730323, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16973 + }, + { + "epoch": 6.781462245305633, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16974 + }, + { + "epoch": 6.781861765880943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16975 + }, + { + "epoch": 6.782261286456253, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16976 + }, + { + "epoch": 6.782660807031562, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16977 + }, + { + "epoch": 6.783060327606872, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16978 + }, + { + "epoch": 6.783459848182181, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16979 + }, + { + "epoch": 6.783859368757491, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16980 + }, + { + "epoch": 6.784258889332801, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16981 + }, + { + "epoch": 6.78465840990811, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16982 + }, + { + "epoch": 6.78505793048342, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16983 + }, + { + "epoch": 6.785457451058729, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16984 + }, + { + "epoch": 6.785856971634039, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16985 + }, + { + "epoch": 6.786256492209349, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16986 + }, + { + "epoch": 6.7866560127846585, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16987 + }, + { + "epoch": 6.787055533359968, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16988 + }, + { + "epoch": 6.787455053935278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16989 + }, + { + "epoch": 6.787854574510587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16990 + }, + { + "epoch": 6.788254095085897, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16991 + }, + { + "epoch": 6.7886536156612065, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16992 + }, + { + "epoch": 6.789053136236516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16993 + }, + { + "epoch": 6.789452656811826, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16994 + }, + { + "epoch": 6.789852177387136, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16995 + }, + { + "epoch": 6.790251697962445, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16996 + }, + { + "epoch": 6.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16997 + }, + { + "epoch": 6.791050739113064, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16998 + }, + { + "epoch": 6.791450259688374, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16999 + }, + { + "epoch": 6.791849780263684, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17000 + }, + { + "epoch": 6.792249300838993, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17001 + }, + { + "epoch": 6.792648821414303, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17002 + }, + { + "epoch": 6.793048341989612, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17003 + }, + { + "epoch": 6.793447862564922, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17004 + }, + { + "epoch": 6.793847383140232, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17005 + }, + { + "epoch": 6.794246903715541, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17006 + }, + { + "epoch": 6.794646424290851, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17007 + }, + { + "epoch": 6.79504594486616, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17008 + }, + { + "epoch": 6.79544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17009 + }, + { + "epoch": 6.79584498601678, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17010 + }, + { + "epoch": 6.79624450659209, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17011 + }, + { + "epoch": 6.7966440271674, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17012 + }, + { + "epoch": 6.7970435477427085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17013 + }, + { + "epoch": 6.797443068318018, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17014 + }, + { + "epoch": 6.797842588893328, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17015 + }, + { + "epoch": 6.798242109468638, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17016 + }, + { + "epoch": 6.798641630043948, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17017 + }, + { + "epoch": 6.7990411506192565, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17018 + }, + { + "epoch": 6.799440671194566, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17019 + }, + { + "epoch": 6.799840191769876, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17020 + }, + { + "epoch": 6.800239712345186, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17021 + }, + { + "epoch": 6.800639232920496, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17022 + }, + { + "epoch": 6.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17023 + }, + { + "epoch": 6.801438274071114, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17024 + }, + { + "epoch": 6.801837794646424, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17025 + }, + { + "epoch": 6.802237315221734, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17026 + }, + { + "epoch": 6.802636835797044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17027 + }, + { + "epoch": 6.8030363563723535, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17028 + }, + { + "epoch": 6.803435876947663, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17029 + }, + { + "epoch": 6.803835397522972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17030 + }, + { + "epoch": 6.804234918098282, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17031 + }, + { + "epoch": 6.804634438673592, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17032 + }, + { + "epoch": 6.8050339592489015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17033 + }, + { + "epoch": 6.805433479824211, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17034 + }, + { + "epoch": 6.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17035 + }, + { + "epoch": 6.80623252097483, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17036 + }, + { + "epoch": 6.80663204155014, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17037 + }, + { + "epoch": 6.80703156212545, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17038 + }, + { + "epoch": 6.807431082700759, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17039 + }, + { + "epoch": 6.807830603276068, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17040 + }, + { + "epoch": 6.808230123851378, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17041 + }, + { + "epoch": 6.808629644426688, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17042 + }, + { + "epoch": 6.809029165001998, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17043 + }, + { + "epoch": 6.809428685577307, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17044 + }, + { + "epoch": 6.809828206152617, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17045 + }, + { + "epoch": 6.810227726727927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17046 + }, + { + "epoch": 6.810627247303236, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17047 + }, + { + "epoch": 6.811026767878546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17048 + }, + { + "epoch": 6.8114262884538554, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17049 + }, + { + "epoch": 6.811825809029165, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17050 + }, + { + "epoch": 6.812225329604475, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17051 + }, + { + "epoch": 6.812624850179784, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17052 + }, + { + "epoch": 6.813024370755094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17053 + }, + { + "epoch": 6.8134238913304035, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17054 + }, + { + "epoch": 6.813823411905713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 17055 + }, + { + "epoch": 6.814222932481023, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17056 + }, + { + "epoch": 6.814622453056332, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 17057 + }, + { + "epoch": 6.815021973631642, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17058 + }, + { + "epoch": 6.8154214942069515, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17059 + }, + { + "epoch": 6.815821014782261, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17060 + }, + { + "epoch": 6.816220535357571, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 17061 + }, + { + "epoch": 6.816620055932881, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 17062 + }, + { + "epoch": 6.81701957650819, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17063 + }, + { + "epoch": 6.8174190970835, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17064 + }, + { + "epoch": 6.817818617658809, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17065 + }, + { + "epoch": 6.818218138234119, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17066 + }, + { + "epoch": 6.818617658809429, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17067 + }, + { + "epoch": 6.819017179384739, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 17068 + }, + { + "epoch": 6.819416699960048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17069 + }, + { + "epoch": 6.819816220535357, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17070 + }, + { + "epoch": 6.820215741110667, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17071 + }, + { + "epoch": 6.820615261685977, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17072 + }, + { + "epoch": 6.821014782261287, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17073 + }, + { + "epoch": 6.821414302836596, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17074 + }, + { + "epoch": 6.8218138234119055, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 17075 + }, + { + "epoch": 6.822213343987215, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17076 + }, + { + "epoch": 6.822612864562525, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 17077 + }, + { + "epoch": 6.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17078 + }, + { + "epoch": 6.823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17079 + }, + { + "epoch": 6.8238114262884535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17080 + }, + { + "epoch": 6.824210946863763, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17081 + }, + { + "epoch": 6.824610467439073, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17082 + }, + { + "epoch": 6.825009988014383, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17083 + }, + { + "epoch": 6.825409508589693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17084 + }, + { + "epoch": 6.825809029165002, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17085 + }, + { + "epoch": 6.826208549740311, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17086 + }, + { + "epoch": 6.826608070315621, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17087 + }, + { + "epoch": 6.827007590890931, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17088 + }, + { + "epoch": 6.827407111466241, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17089 + }, + { + "epoch": 6.8278066320415505, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17090 + }, + { + "epoch": 6.828206152616859, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17091 + }, + { + "epoch": 6.828605673192169, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 17092 + }, + { + "epoch": 6.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17093 + }, + { + "epoch": 6.829404714342789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17094 + }, + { + "epoch": 6.8298042349180985, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17095 + }, + { + "epoch": 6.830203755493407, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17096 + }, + { + "epoch": 6.830603276068717, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17097 + }, + { + "epoch": 6.831002796644027, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17098 + }, + { + "epoch": 6.831402317219337, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17099 + }, + { + "epoch": 6.831801837794647, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17100 + }, + { + "epoch": 6.832201358369956, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17101 + }, + { + "epoch": 6.832600878945266, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17102 + }, + { + "epoch": 6.833000399520575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17103 + }, + { + "epoch": 6.833399920095885, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17104 + }, + { + "epoch": 6.833799440671195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17105 + }, + { + "epoch": 6.834198961246504, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17106 + }, + { + "epoch": 6.834598481821814, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17107 + }, + { + "epoch": 6.834998002397123, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17108 + }, + { + "epoch": 6.835397522972433, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17109 + }, + { + "epoch": 6.835797043547743, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17110 + }, + { + "epoch": 6.836196564123052, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17111 + }, + { + "epoch": 6.836596084698362, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17112 + }, + { + "epoch": 6.836995605273671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17113 + }, + { + "epoch": 6.837395125848981, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17114 + }, + { + "epoch": 6.837794646424291, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17115 + }, + { + "epoch": 6.8381941669996005, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17116 + }, + { + "epoch": 6.83859368757491, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17117 + }, + { + "epoch": 6.83899320815022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17118 + }, + { + "epoch": 6.83939272872553, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17119 + }, + { + "epoch": 6.839792249300839, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17120 + }, + { + "epoch": 6.8401917698761485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17121 + }, + { + "epoch": 6.840591290451458, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17122 + }, + { + "epoch": 6.840990811026768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17123 + }, + { + "epoch": 6.841390331602078, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17124 + }, + { + "epoch": 6.841789852177387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17125 + }, + { + "epoch": 6.842189372752697, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17126 + }, + { + "epoch": 6.842588893328006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17127 + }, + { + "epoch": 6.842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17128 + }, + { + "epoch": 6.843387934478626, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17129 + }, + { + "epoch": 6.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 17130 + }, + { + "epoch": 6.844186975629245, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17131 + }, + { + "epoch": 6.844586496204554, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17132 + }, + { + "epoch": 6.844986016779864, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17133 + }, + { + "epoch": 6.845385537355174, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17134 + }, + { + "epoch": 6.845785057930484, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17135 + }, + { + "epoch": 6.8461845785057935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17136 + }, + { + "epoch": 6.846584099081102, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17137 + }, + { + "epoch": 6.846983619656412, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17138 + }, + { + "epoch": 6.847383140231722, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17139 + }, + { + "epoch": 6.847782660807032, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17140 + }, + { + "epoch": 6.848182181382342, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17141 + }, + { + "epoch": 6.8485817019576505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17142 + }, + { + "epoch": 6.84898122253296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17143 + }, + { + "epoch": 6.84938074310827, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17144 + }, + { + "epoch": 6.84978026368358, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17145 + }, + { + "epoch": 6.85017978425889, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17146 + }, + { + "epoch": 6.8505793048341985, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17147 + }, + { + "epoch": 6.850978825409508, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17148 + }, + { + "epoch": 6.851378345984818, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17149 + }, + { + "epoch": 6.851777866560128, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17150 + }, + { + "epoch": 6.852177387135438, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 17151 + }, + { + "epoch": 6.8525769077107475, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.2944, + "step": 17152 + }, + { + "epoch": 6.852976428286056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17153 + }, + { + "epoch": 6.853375948861366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17154 + }, + { + "epoch": 6.853775469436676, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17155 + }, + { + "epoch": 6.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17156 + }, + { + "epoch": 6.8545745105872955, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17157 + }, + { + "epoch": 6.854974031162605, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17158 + }, + { + "epoch": 6.855373551737914, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17159 + }, + { + "epoch": 6.855773072313224, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17160 + }, + { + "epoch": 6.856172592888534, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17161 + }, + { + "epoch": 6.8565721134638435, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 17162 + }, + { + "epoch": 6.856971634039153, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17163 + }, + { + "epoch": 6.857371154614462, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17164 + }, + { + "epoch": 6.857770675189772, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17165 + }, + { + "epoch": 6.858170195765082, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17166 + }, + { + "epoch": 6.858569716340392, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17167 + }, + { + "epoch": 6.858969236915701, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17168 + }, + { + "epoch": 6.859368757491011, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17169 + }, + { + "epoch": 6.85976827806632, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17170 + }, + { + "epoch": 6.86016779864163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17171 + }, + { + "epoch": 6.86056731921694, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17172 + }, + { + "epoch": 6.860966839792249, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17173 + }, + { + "epoch": 6.861366360367559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17174 + }, + { + "epoch": 6.861765880942869, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17175 + }, + { + "epoch": 6.862165401518178, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17176 + }, + { + "epoch": 6.862564922093488, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17177 + }, + { + "epoch": 6.8629644426687975, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17178 + }, + { + "epoch": 6.863363963244107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17179 + }, + { + "epoch": 6.863763483819417, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17180 + }, + { + "epoch": 6.864163004394726, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17181 + }, + { + "epoch": 6.864562524970036, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17182 + }, + { + "epoch": 6.8649620455453455, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17183 + }, + { + "epoch": 6.865361566120655, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17184 + }, + { + "epoch": 6.865761086695965, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17185 + }, + { + "epoch": 6.866160607271274, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17186 + }, + { + "epoch": 6.866560127846584, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17187 + }, + { + "epoch": 6.8669596484218935, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17188 + }, + { + "epoch": 6.867359168997203, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17189 + }, + { + "epoch": 6.867758689572513, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17190 + }, + { + "epoch": 6.868158210147823, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17191 + }, + { + "epoch": 6.868557730723133, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17192 + }, + { + "epoch": 6.868957251298442, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17193 + }, + { + "epoch": 6.869356771873751, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17194 + }, + { + "epoch": 6.869756292449061, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17195 + }, + { + "epoch": 6.870155813024371, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17196 + }, + { + "epoch": 6.870555333599681, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17197 + }, + { + "epoch": 6.87095485417499, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17198 + }, + { + "epoch": 6.871354374750299, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17199 + }, + { + "epoch": 6.871753895325609, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17200 + }, + { + "epoch": 6.872153415900919, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17201 + }, + { + "epoch": 6.872552936476229, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 17202 + }, + { + "epoch": 6.872952457051538, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17203 + }, + { + "epoch": 6.8733519776268475, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17204 + }, + { + "epoch": 6.873751498202157, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17205 + }, + { + "epoch": 6.874151018777467, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17206 + }, + { + "epoch": 6.874550539352777, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 17207 + }, + { + "epoch": 6.874950059928087, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17208 + }, + { + "epoch": 6.875349580503396, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17209 + }, + { + "epoch": 6.875749101078705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17210 + }, + { + "epoch": 6.876148621654015, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17211 + }, + { + "epoch": 6.876548142229325, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17212 + }, + { + "epoch": 6.876947662804635, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17213 + }, + { + "epoch": 6.877347183379944, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17214 + }, + { + "epoch": 6.877746703955253, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17215 + }, + { + "epoch": 6.878146224530563, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17216 + }, + { + "epoch": 6.878545745105873, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17217 + }, + { + "epoch": 6.878945265681183, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17218 + }, + { + "epoch": 6.8793447862564925, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17219 + }, + { + "epoch": 6.879744306831801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17220 + }, + { + "epoch": 6.880143827407111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17221 + }, + { + "epoch": 6.880543347982421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17222 + }, + { + "epoch": 6.880942868557731, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17223 + }, + { + "epoch": 6.8813423891330405, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17224 + }, + { + "epoch": 6.88174190970835, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17225 + }, + { + "epoch": 6.88214143028366, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17226 + }, + { + "epoch": 6.882540950858969, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17227 + }, + { + "epoch": 6.882940471434279, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17228 + }, + { + "epoch": 6.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17229 + }, + { + "epoch": 6.883739512584898, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17230 + }, + { + "epoch": 6.884139033160208, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17231 + }, + { + "epoch": 6.884538553735517, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17232 + }, + { + "epoch": 6.884938074310827, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17233 + }, + { + "epoch": 6.885337594886137, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17234 + }, + { + "epoch": 6.885737115461446, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17235 + }, + { + "epoch": 6.886136636036756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17236 + }, + { + "epoch": 6.886536156612065, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17237 + }, + { + "epoch": 6.886935677187375, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17238 + }, + { + "epoch": 6.887335197762685, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17239 + }, + { + "epoch": 6.887734718337994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17240 + }, + { + "epoch": 6.888134238913304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17241 + }, + { + "epoch": 6.888533759488614, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17242 + }, + { + "epoch": 6.888933280063923, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17243 + }, + { + "epoch": 6.889332800639233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17244 + }, + { + "epoch": 6.8897323212145425, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17245 + }, + { + "epoch": 6.890131841789852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17246 + }, + { + "epoch": 6.890531362365162, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17247 + }, + { + "epoch": 6.890930882940472, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17248 + }, + { + "epoch": 6.891330403515781, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17249 + }, + { + "epoch": 6.8917299240910905, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17250 + }, + { + "epoch": 6.8921294446664, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17251 + }, + { + "epoch": 6.89252896524171, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 17252 + }, + { + "epoch": 6.89292848581702, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17253 + }, + { + "epoch": 6.893328006392329, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17254 + }, + { + "epoch": 6.893727526967639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17255 + }, + { + "epoch": 6.894127047542948, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17256 + }, + { + "epoch": 6.894526568118258, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17257 + }, + { + "epoch": 6.894926088693568, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17258 + }, + { + "epoch": 6.895325609268878, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17259 + }, + { + "epoch": 6.895725129844187, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17260 + }, + { + "epoch": 6.896124650419496, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17261 + }, + { + "epoch": 6.896524170994806, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17262 + }, + { + "epoch": 6.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17263 + }, + { + "epoch": 6.897323212145426, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17264 + }, + { + "epoch": 6.8977227327207356, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17265 + }, + { + "epoch": 6.8981222532960444, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17266 + }, + { + "epoch": 6.898521773871354, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17267 + }, + { + "epoch": 6.898921294446664, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17268 + }, + { + "epoch": 6.899320815021974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17269 + }, + { + "epoch": 6.899720335597284, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17270 + }, + { + "epoch": 6.9001198561725925, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17271 + }, + { + "epoch": 6.900519376747902, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17272 + }, + { + "epoch": 6.900918897323212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17273 + }, + { + "epoch": 6.901318417898522, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17274 + }, + { + "epoch": 6.901717938473832, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17275 + }, + { + "epoch": 6.9021174590491405, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.2994, + "step": 17276 + }, + { + "epoch": 6.90251697962445, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17277 + }, + { + "epoch": 6.90291650019976, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17278 + }, + { + "epoch": 6.90331602077507, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17279 + }, + { + "epoch": 6.90371554135038, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17280 + }, + { + "epoch": 6.9041150619256895, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17281 + }, + { + "epoch": 6.904514582500999, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17282 + }, + { + "epoch": 6.904914103076308, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17283 + }, + { + "epoch": 6.905313623651618, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17284 + }, + { + "epoch": 6.905713144226928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17285 + }, + { + "epoch": 6.9061126648022375, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17286 + }, + { + "epoch": 6.906512185377547, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17287 + }, + { + "epoch": 6.906911705952856, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17288 + }, + { + "epoch": 6.907311226528166, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17289 + }, + { + "epoch": 6.907710747103476, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17290 + }, + { + "epoch": 6.9081102676787856, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17291 + }, + { + "epoch": 6.908509788254095, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17292 + }, + { + "epoch": 6.908909308829404, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17293 + }, + { + "epoch": 6.909308829404714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17294 + }, + { + "epoch": 6.909708349980024, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17295 + }, + { + "epoch": 6.910107870555334, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17296 + }, + { + "epoch": 6.910507391130643, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17297 + }, + { + "epoch": 6.910906911705953, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17298 + }, + { + "epoch": 6.911306432281263, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17299 + }, + { + "epoch": 6.911705952856572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17300 + }, + { + "epoch": 6.912105473431882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17301 + }, + { + "epoch": 6.912504994007191, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17302 + }, + { + "epoch": 6.912904514582501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17303 + }, + { + "epoch": 6.913304035157811, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17304 + }, + { + "epoch": 6.91370355573312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17305 + }, + { + "epoch": 6.91410307630843, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17306 + }, + { + "epoch": 6.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17307 + }, + { + "epoch": 6.914902117459049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 17308 + }, + { + "epoch": 6.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17309 + }, + { + "epoch": 6.915701158609668, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17310 + }, + { + "epoch": 6.916100679184978, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17311 + }, + { + "epoch": 6.9165001997602875, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17312 + }, + { + "epoch": 6.916899720335597, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17313 + }, + { + "epoch": 6.917299240910907, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17314 + }, + { + "epoch": 6.917698761486217, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 17315 + }, + { + "epoch": 6.918098282061527, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17316 + }, + { + "epoch": 6.918497802636836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17317 + }, + { + "epoch": 6.918897323212145, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17318 + }, + { + "epoch": 6.919296843787455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17319 + }, + { + "epoch": 6.919696364362765, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17320 + }, + { + "epoch": 6.920095884938075, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17321 + }, + { + "epoch": 6.920495405513384, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17322 + }, + { + "epoch": 6.920894926088693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17323 + }, + { + "epoch": 6.921294446664003, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17324 + }, + { + "epoch": 6.921693967239313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 17325 + }, + { + "epoch": 6.922093487814623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17326 + }, + { + "epoch": 6.922493008389932, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17327 + }, + { + "epoch": 6.922892528965241, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17328 + }, + { + "epoch": 6.923292049540551, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17329 + }, + { + "epoch": 6.923691570115861, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 17330 + }, + { + "epoch": 6.924091090691171, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17331 + }, + { + "epoch": 6.924490611266481, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17332 + }, + { + "epoch": 6.9248901318417895, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17333 + }, + { + "epoch": 6.925289652417099, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17334 + }, + { + "epoch": 6.925689172992409, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17335 + }, + { + "epoch": 6.926088693567719, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17336 + }, + { + "epoch": 6.926488214143029, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17337 + }, + { + "epoch": 6.926887734718338, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17338 + }, + { + "epoch": 6.927287255293647, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17339 + }, + { + "epoch": 6.927686775868957, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 17340 + }, + { + "epoch": 6.928086296444267, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17341 + }, + { + "epoch": 6.928485817019577, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17342 + }, + { + "epoch": 6.9288853375948865, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17343 + }, + { + "epoch": 6.929284858170195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17344 + }, + { + "epoch": 6.929684378745505, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17345 + }, + { + "epoch": 6.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 17346 + }, + { + "epoch": 6.930483419896125, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17347 + }, + { + "epoch": 6.9308829404714345, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 17348 + }, + { + "epoch": 6.931282461046744, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17349 + }, + { + "epoch": 6.931681981622053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17350 + }, + { + "epoch": 6.932081502197363, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 17351 + }, + { + "epoch": 6.932481022772673, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17352 + }, + { + "epoch": 6.9328805433479825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17353 + }, + { + "epoch": 6.933280063923292, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17354 + }, + { + "epoch": 6.933679584498602, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17355 + }, + { + "epoch": 6.934079105073911, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17356 + }, + { + "epoch": 6.934478625649221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17357 + }, + { + "epoch": 6.934878146224531, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17358 + }, + { + "epoch": 6.93527766679984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17359 + }, + { + "epoch": 6.93567718737515, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17360 + }, + { + "epoch": 6.936076707950459, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17361 + }, + { + "epoch": 6.936476228525769, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 17362 + }, + { + "epoch": 6.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17363 + }, + { + "epoch": 6.937275269676388, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17364 + }, + { + "epoch": 6.937674790251698, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17365 + }, + { + "epoch": 6.938074310827007, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17366 + }, + { + "epoch": 6.938473831402317, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17367 + }, + { + "epoch": 6.938873351977627, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 17368 + }, + { + "epoch": 6.9392728725529365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17369 + }, + { + "epoch": 6.939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17370 + }, + { + "epoch": 6.940071913703556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17371 + }, + { + "epoch": 6.940471434278866, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17372 + }, + { + "epoch": 6.940870954854175, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17373 + }, + { + "epoch": 6.9412704754294845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17374 + }, + { + "epoch": 6.941669996004794, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17375 + }, + { + "epoch": 6.942069516580104, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17376 + }, + { + "epoch": 6.942469037155414, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17377 + }, + { + "epoch": 6.942868557730723, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17378 + }, + { + "epoch": 6.9432680783060325, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17379 + }, + { + "epoch": 6.943667598881342, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 17380 + }, + { + "epoch": 6.944067119456652, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17381 + }, + { + "epoch": 6.944466640031962, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17382 + }, + { + "epoch": 6.944866160607271, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17383 + }, + { + "epoch": 6.945265681182581, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 17384 + }, + { + "epoch": 6.94566520175789, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17385 + }, + { + "epoch": 6.9460647223332, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17386 + }, + { + "epoch": 6.94646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17387 + }, + { + "epoch": 6.94686376348382, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 17388 + }, + { + "epoch": 6.9472632840591295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17389 + }, + { + "epoch": 6.947662804634438, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17390 + }, + { + "epoch": 6.948062325209748, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17391 + }, + { + "epoch": 6.948461845785058, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17392 + }, + { + "epoch": 6.948861366360368, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17393 + }, + { + "epoch": 6.949260886935678, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17394 + }, + { + "epoch": 6.9496604075109865, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17395 + }, + { + "epoch": 6.950059928086296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 17396 + }, + { + "epoch": 6.950459448661606, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 17397 + }, + { + "epoch": 6.950858969236916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17398 + }, + { + "epoch": 6.951258489812226, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17399 + }, + { + "epoch": 6.9516580103875345, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17400 + }, + { + "epoch": 6.952057530962844, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17401 + }, + { + "epoch": 6.952457051538154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17402 + }, + { + "epoch": 6.952856572113464, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17403 + }, + { + "epoch": 6.953256092688774, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17404 + }, + { + "epoch": 6.953655613264083, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17405 + }, + { + "epoch": 6.954055133839393, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17406 + }, + { + "epoch": 6.954454654414702, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17407 + }, + { + "epoch": 6.954854174990012, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17408 + }, + { + "epoch": 6.955253695565322, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17409 + }, + { + "epoch": 6.9556532161406315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17410 + }, + { + "epoch": 6.956052736715941, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 17411 + }, + { + "epoch": 6.95645225729125, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17412 + }, + { + "epoch": 6.95685177786656, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17413 + }, + { + "epoch": 6.95725129844187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17414 + }, + { + "epoch": 6.9576508190171795, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17415 + }, + { + "epoch": 6.958050339592489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17416 + }, + { + "epoch": 6.958449860167798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17417 + }, + { + "epoch": 6.958849380743108, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17418 + }, + { + "epoch": 6.959248901318418, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17419 + }, + { + "epoch": 6.959648421893728, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17420 + }, + { + "epoch": 6.960047942469037, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17421 + }, + { + "epoch": 6.960447463044347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17422 + }, + { + "epoch": 6.960846983619656, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17423 + }, + { + "epoch": 6.961246504194966, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17424 + }, + { + "epoch": 6.961646024770276, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17425 + }, + { + "epoch": 6.962045545345585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17426 + }, + { + "epoch": 6.962445065920895, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17427 + }, + { + "epoch": 6.962844586496205, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17428 + }, + { + "epoch": 6.963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17429 + }, + { + "epoch": 6.963643627646824, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17430 + }, + { + "epoch": 6.964043148222133, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17431 + }, + { + "epoch": 6.964442668797443, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17432 + }, + { + "epoch": 6.964842189372753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17433 + }, + { + "epoch": 6.965241709948062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17434 + }, + { + "epoch": 6.965641230523372, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17435 + }, + { + "epoch": 6.9660407510986815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17436 + }, + { + "epoch": 6.966440271673991, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17437 + }, + { + "epoch": 6.966839792249301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17438 + }, + { + "epoch": 6.967239312824611, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 17439 + }, + { + "epoch": 6.96763883339992, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17440 + }, + { + "epoch": 6.9680383539752295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17441 + }, + { + "epoch": 6.968437874550539, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17442 + }, + { + "epoch": 6.968837395125849, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17443 + }, + { + "epoch": 6.969236915701159, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17444 + }, + { + "epoch": 6.969636436276469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17445 + }, + { + "epoch": 6.970035956851778, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17446 + }, + { + "epoch": 6.970435477427087, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17447 + }, + { + "epoch": 6.970834998002397, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17448 + }, + { + "epoch": 6.971234518577707, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17449 + }, + { + "epoch": 6.971634039153017, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17450 + }, + { + "epoch": 6.972033559728326, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17451 + }, + { + "epoch": 6.972433080303635, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17452 + }, + { + "epoch": 6.972832600878945, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17453 + }, + { + "epoch": 6.973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17454 + }, + { + "epoch": 6.973631642029565, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17455 + }, + { + "epoch": 6.974031162604874, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17456 + }, + { + "epoch": 6.974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17457 + }, + { + "epoch": 6.974830203755493, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17458 + }, + { + "epoch": 6.975229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17459 + }, + { + "epoch": 6.975629244906113, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17460 + }, + { + "epoch": 6.976028765481423, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17461 + }, + { + "epoch": 6.976428286056732, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 17462 + }, + { + "epoch": 6.976827806632041, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17463 + }, + { + "epoch": 6.977227327207351, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17464 + }, + { + "epoch": 6.977626847782661, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17465 + }, + { + "epoch": 6.978026368357971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17466 + }, + { + "epoch": 6.97842588893328, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17467 + }, + { + "epoch": 6.978825409508589, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17468 + }, + { + "epoch": 6.979224930083899, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17469 + }, + { + "epoch": 6.979624450659209, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17470 + }, + { + "epoch": 6.980023971234519, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17471 + }, + { + "epoch": 6.9804234918098285, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17472 + }, + { + "epoch": 6.980823012385137, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17473 + }, + { + "epoch": 6.981222532960447, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17474 + }, + { + "epoch": 6.981622053535757, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17475 + }, + { + "epoch": 6.982021574111067, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17476 + }, + { + "epoch": 6.9824210946863765, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17477 + }, + { + "epoch": 6.982820615261686, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17478 + }, + { + "epoch": 6.983220135836996, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17479 + }, + { + "epoch": 6.983619656412305, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17480 + }, + { + "epoch": 6.984019176987615, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17481 + }, + { + "epoch": 6.9844186975629245, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17482 + }, + { + "epoch": 6.984818218138234, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17483 + }, + { + "epoch": 6.985217738713544, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17484 + }, + { + "epoch": 6.985617259288853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17485 + }, + { + "epoch": 6.986016779864163, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17486 + }, + { + "epoch": 6.986416300439473, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17487 + }, + { + "epoch": 6.986815821014782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17488 + }, + { + "epoch": 6.987215341590092, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17489 + }, + { + "epoch": 6.987614862165401, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17490 + }, + { + "epoch": 6.988014382740711, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17491 + }, + { + "epoch": 6.988413903316021, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17492 + }, + { + "epoch": 6.98881342389133, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17493 + }, + { + "epoch": 6.98921294446664, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17494 + }, + { + "epoch": 6.98961246504195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17495 + }, + { + "epoch": 6.99001198561726, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17496 + }, + { + "epoch": 6.990411506192569, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17497 + }, + { + "epoch": 6.9908110267678785, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17498 + }, + { + "epoch": 6.991210547343188, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17499 + }, + { + "epoch": 6.991610067918498, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17500 + }, + { + "epoch": 6.992009588493808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17501 + }, + { + "epoch": 6.992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 17502 + }, + { + "epoch": 6.9928086296444265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17503 + }, + { + "epoch": 6.993208150219736, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17504 + }, + { + "epoch": 6.993607670795046, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 17505 + }, + { + "epoch": 6.994007191370356, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17506 + }, + { + "epoch": 6.994406711945665, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17507 + }, + { + "epoch": 6.9948062325209746, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17508 + }, + { + "epoch": 6.995205753096284, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17509 + }, + { + "epoch": 6.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17510 + }, + { + "epoch": 6.996004794246904, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17511 + }, + { + "epoch": 6.996404314822214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17512 + }, + { + "epoch": 6.996803835397523, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 17513 + }, + { + "epoch": 6.997203355972832, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17514 + }, + { + "epoch": 6.997602876548142, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17515 + }, + { + "epoch": 6.998002397123452, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17516 + }, + { + "epoch": 6.998401917698762, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 17517 + }, + { + "epoch": 6.9988014382740715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17518 + }, + { + "epoch": 6.99920095884938, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 17519 + }, + { + "epoch": 6.99960047942469, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17520 + }, + { + "epoch": 7.0, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17521 + }, + { + "epoch": 7.00039952057531, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17522 + }, + { + "epoch": 7.00079904115062, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17523 + }, + { + "epoch": 7.0011985617259285, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17524 + }, + { + "epoch": 7.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17525 + }, + { + "epoch": 7.001997602876548, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17526 + }, + { + "epoch": 7.002397123451858, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17527 + }, + { + "epoch": 7.002796644027168, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17528 + }, + { + "epoch": 7.003196164602477, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17529 + }, + { + "epoch": 7.003595685177786, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17530 + }, + { + "epoch": 7.003995205753096, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17531 + }, + { + "epoch": 7.004394726328406, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17532 + }, + { + "epoch": 7.004794246903716, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17533 + }, + { + "epoch": 7.0051937674790254, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17534 + }, + { + "epoch": 7.005593288054335, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17535 + }, + { + "epoch": 7.005992808629644, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17536 + }, + { + "epoch": 7.006392329204954, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17537 + }, + { + "epoch": 7.006791849780264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17538 + }, + { + "epoch": 7.0071913703555735, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17539 + }, + { + "epoch": 7.007590890930883, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17540 + }, + { + "epoch": 7.007990411506192, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17541 + }, + { + "epoch": 7.008389932081502, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17542 + }, + { + "epoch": 7.008789452656812, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17543 + }, + { + "epoch": 7.0091889732321215, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17544 + }, + { + "epoch": 7.009588493807431, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17545 + }, + { + "epoch": 7.009988014382741, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17546 + }, + { + "epoch": 7.01038753495805, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17547 + }, + { + "epoch": 7.01078705553336, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17548 + }, + { + "epoch": 7.01118657610867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 17549 + }, + { + "epoch": 7.011586096683979, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17550 + }, + { + "epoch": 7.011985617259289, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17551 + }, + { + "epoch": 7.012385137834598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17552 + }, + { + "epoch": 7.012784658409908, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17553 + }, + { + "epoch": 7.013184178985218, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17554 + }, + { + "epoch": 7.013583699560527, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17555 + }, + { + "epoch": 7.013983220135837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17556 + }, + { + "epoch": 7.014382740711147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17557 + }, + { + "epoch": 7.014782261286456, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17558 + }, + { + "epoch": 7.015181781861766, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17559 + }, + { + "epoch": 7.0155813024370755, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17560 + }, + { + "epoch": 7.015980823012385, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17561 + }, + { + "epoch": 7.016380343587695, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17562 + }, + { + "epoch": 7.016779864163005, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17563 + }, + { + "epoch": 7.017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17564 + }, + { + "epoch": 7.0175789053136235, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 17565 + }, + { + "epoch": 7.017978425888933, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17566 + }, + { + "epoch": 7.018377946464243, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 17567 + }, + { + "epoch": 7.018777467039553, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17568 + }, + { + "epoch": 7.019176987614862, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17569 + }, + { + "epoch": 7.0195765081901715, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17570 + }, + { + "epoch": 7.019976028765481, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17571 + }, + { + "epoch": 7.020375549340791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17572 + }, + { + "epoch": 7.020775069916101, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17573 + }, + { + "epoch": 7.021174590491411, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17574 + }, + { + "epoch": 7.02157411106672, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 17575 + }, + { + "epoch": 7.021973631642029, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17576 + }, + { + "epoch": 7.022373152217339, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17577 + }, + { + "epoch": 7.022772672792649, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17578 + }, + { + "epoch": 7.023172193367959, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17579 + }, + { + "epoch": 7.0235717139432685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 17580 + }, + { + "epoch": 7.023971234518577, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17581 + }, + { + "epoch": 7.024370755093887, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17582 + }, + { + "epoch": 7.024770275669197, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17583 + }, + { + "epoch": 7.025169796244507, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17584 + }, + { + "epoch": 7.025569316819817, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17585 + }, + { + "epoch": 7.0259688373951255, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17586 + }, + { + "epoch": 7.026368357970435, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17587 + }, + { + "epoch": 7.026767878545745, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17588 + }, + { + "epoch": 7.027167399121055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17589 + }, + { + "epoch": 7.027566919696365, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 17590 + }, + { + "epoch": 7.027966440271674, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17591 + }, + { + "epoch": 7.028365960846983, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17592 + }, + { + "epoch": 7.028765481422293, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17593 + }, + { + "epoch": 7.029165001997603, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17594 + }, + { + "epoch": 7.029564522572913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17595 + }, + { + "epoch": 7.029964043148222, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17596 + }, + { + "epoch": 7.030363563723531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17597 + }, + { + "epoch": 7.030763084298841, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 17598 + }, + { + "epoch": 7.031162604874151, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17599 + }, + { + "epoch": 7.031562125449461, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17600 + }, + { + "epoch": 7.0319616460247705, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 17601 + }, + { + "epoch": 7.03236116660008, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17602 + }, + { + "epoch": 7.032760687175389, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17603 + }, + { + "epoch": 7.033160207750699, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17604 + }, + { + "epoch": 7.033559728326009, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17605 + }, + { + "epoch": 7.0339592489013185, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17606 + }, + { + "epoch": 7.034358769476628, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17607 + }, + { + "epoch": 7.034758290051938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17608 + }, + { + "epoch": 7.035157810627247, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17609 + }, + { + "epoch": 7.035557331202557, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17610 + }, + { + "epoch": 7.035956851777867, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17611 + }, + { + "epoch": 7.036356372353176, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17612 + }, + { + "epoch": 7.036755892928486, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17613 + }, + { + "epoch": 7.037155413503795, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17614 + }, + { + "epoch": 7.037554934079105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17615 + }, + { + "epoch": 7.037954454654415, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 17616 + }, + { + "epoch": 7.038353975229724, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17617 + }, + { + "epoch": 7.038753495805034, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17618 + }, + { + "epoch": 7.039153016380344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17619 + }, + { + "epoch": 7.039552536955653, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17620 + }, + { + "epoch": 7.039952057530963, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17621 + }, + { + "epoch": 7.040351578106272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 17622 + }, + { + "epoch": 7.040751098681582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17623 + }, + { + "epoch": 7.041150619256892, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17624 + }, + { + "epoch": 7.041550139832202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 17625 + }, + { + "epoch": 7.041949660407511, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17626 + }, + { + "epoch": 7.0423491809828205, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17627 + }, + { + "epoch": 7.04274870155813, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17628 + }, + { + "epoch": 7.04314822213344, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17629 + }, + { + "epoch": 7.04354774270875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17630 + }, + { + "epoch": 7.043947263284059, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17631 + }, + { + "epoch": 7.0443467838593685, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17632 + }, + { + "epoch": 7.044746304434678, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17633 + }, + { + "epoch": 7.045145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17634 + }, + { + "epoch": 7.045545345585298, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17635 + }, + { + "epoch": 7.045944866160608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17636 + }, + { + "epoch": 7.046344386735917, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.2991, + "step": 17637 + }, + { + "epoch": 7.046743907311226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17638 + }, + { + "epoch": 7.047143427886536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17639 + }, + { + "epoch": 7.047542948461846, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17640 + }, + { + "epoch": 7.047942469037156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17641 + }, + { + "epoch": 7.048341989612465, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17642 + }, + { + "epoch": 7.048741510187774, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17643 + }, + { + "epoch": 7.049141030763084, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17644 + }, + { + "epoch": 7.049540551338394, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17645 + }, + { + "epoch": 7.049940071913704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17646 + }, + { + "epoch": 7.0503395924890135, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17647 + }, + { + "epoch": 7.050739113064322, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17648 + }, + { + "epoch": 7.051138633639632, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17649 + }, + { + "epoch": 7.051538154214942, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17650 + }, + { + "epoch": 7.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17651 + }, + { + "epoch": 7.052337195365562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17652 + }, + { + "epoch": 7.052736715940871, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17653 + }, + { + "epoch": 7.05313623651618, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17654 + }, + { + "epoch": 7.05353575709149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17655 + }, + { + "epoch": 7.0539352776668, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17656 + }, + { + "epoch": 7.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17657 + }, + { + "epoch": 7.054734318817419, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17658 + }, + { + "epoch": 7.055133839392728, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17659 + }, + { + "epoch": 7.055533359968038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17660 + }, + { + "epoch": 7.055932880543348, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17661 + }, + { + "epoch": 7.056332401118658, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17662 + }, + { + "epoch": 7.0567319216939675, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17663 + }, + { + "epoch": 7.057131442269277, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17664 + }, + { + "epoch": 7.057530962844586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17665 + }, + { + "epoch": 7.057930483419896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17666 + }, + { + "epoch": 7.058330003995206, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17667 + }, + { + "epoch": 7.0587295245705155, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17668 + }, + { + "epoch": 7.059129045145825, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17669 + }, + { + "epoch": 7.059528565721135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17670 + }, + { + "epoch": 7.059928086296444, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17671 + }, + { + "epoch": 7.060327606871754, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17672 + }, + { + "epoch": 7.0607271274470635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17673 + }, + { + "epoch": 7.061126648022373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17674 + }, + { + "epoch": 7.061526168597683, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17675 + }, + { + "epoch": 7.061925689172992, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17676 + }, + { + "epoch": 7.062325209748302, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17677 + }, + { + "epoch": 7.062724730323612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17678 + }, + { + "epoch": 7.063124250898921, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17679 + }, + { + "epoch": 7.063523771474231, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17680 + }, + { + "epoch": 7.063923292049541, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17681 + }, + { + "epoch": 7.06432281262485, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 17682 + }, + { + "epoch": 7.06472233320016, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17683 + }, + { + "epoch": 7.065121853775469, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17684 + }, + { + "epoch": 7.065521374350779, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17685 + }, + { + "epoch": 7.065920894926089, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17686 + }, + { + "epoch": 7.066320415501398, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17687 + }, + { + "epoch": 7.066719936076708, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17688 + }, + { + "epoch": 7.0671194566520175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 17689 + }, + { + "epoch": 7.067518977227327, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17690 + }, + { + "epoch": 7.067918497802637, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17691 + }, + { + "epoch": 7.068318018377947, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17692 + }, + { + "epoch": 7.068717538953256, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17693 + }, + { + "epoch": 7.0691170595285655, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17694 + }, + { + "epoch": 7.069516580103875, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17695 + }, + { + "epoch": 7.069916100679185, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17696 + }, + { + "epoch": 7.070315621254495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17697 + }, + { + "epoch": 7.070715141829805, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17698 + }, + { + "epoch": 7.0711146624051135, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 17699 + }, + { + "epoch": 7.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17700 + }, + { + "epoch": 7.071913703555733, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17701 + }, + { + "epoch": 7.072313224131043, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 17702 + }, + { + "epoch": 7.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17703 + }, + { + "epoch": 7.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17704 + }, + { + "epoch": 7.073511785856971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17705 + }, + { + "epoch": 7.073911306432281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17706 + }, + { + "epoch": 7.074310827007591, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17707 + }, + { + "epoch": 7.074710347582901, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17708 + }, + { + "epoch": 7.0751098681582105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17709 + }, + { + "epoch": 7.075509388733519, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17710 + }, + { + "epoch": 7.075908909308829, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17711 + }, + { + "epoch": 7.076308429884139, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17712 + }, + { + "epoch": 7.076707950459449, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17713 + }, + { + "epoch": 7.077107471034759, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17714 + }, + { + "epoch": 7.077506991610068, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17715 + }, + { + "epoch": 7.077906512185377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17716 + }, + { + "epoch": 7.078306032760687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17717 + }, + { + "epoch": 7.078705553335997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17718 + }, + { + "epoch": 7.079105073911307, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17719 + }, + { + "epoch": 7.079504594486616, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17720 + }, + { + "epoch": 7.079904115061925, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17721 + }, + { + "epoch": 7.080303635637235, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17722 + }, + { + "epoch": 7.080703156212545, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17723 + }, + { + "epoch": 7.081102676787855, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17724 + }, + { + "epoch": 7.081502197363164, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17725 + }, + { + "epoch": 7.081901717938474, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17726 + }, + { + "epoch": 7.082301238513783, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17727 + }, + { + "epoch": 7.082700759089093, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17728 + }, + { + "epoch": 7.083100279664403, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17729 + }, + { + "epoch": 7.0834998002397125, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17730 + }, + { + "epoch": 7.083899320815022, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 17731 + }, + { + "epoch": 7.084298841390331, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17732 + }, + { + "epoch": 7.084698361965641, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17733 + }, + { + "epoch": 7.085097882540951, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17734 + }, + { + "epoch": 7.0854974031162605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17735 + }, + { + "epoch": 7.08589692369157, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17736 + }, + { + "epoch": 7.08629644426688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17737 + }, + { + "epoch": 7.086695964842189, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17738 + }, + { + "epoch": 7.087095485417499, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17739 + }, + { + "epoch": 7.087495005992809, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17740 + }, + { + "epoch": 7.087894526568118, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17741 + }, + { + "epoch": 7.088294047143428, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17742 + }, + { + "epoch": 7.088693567718738, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17743 + }, + { + "epoch": 7.089093088294047, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17744 + }, + { + "epoch": 7.089492608869357, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17745 + }, + { + "epoch": 7.089892129444666, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17746 + }, + { + "epoch": 7.090291650019976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17747 + }, + { + "epoch": 7.090691170595286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17748 + }, + { + "epoch": 7.091090691170595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 17749 + }, + { + "epoch": 7.091490211745905, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17750 + }, + { + "epoch": 7.0918897323212144, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17751 + }, + { + "epoch": 7.092289252896524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17752 + }, + { + "epoch": 7.092688773471834, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17753 + }, + { + "epoch": 7.093088294047144, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17754 + }, + { + "epoch": 7.093487814622453, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17755 + }, + { + "epoch": 7.0938873351977625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17756 + }, + { + "epoch": 7.094286855773072, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17757 + }, + { + "epoch": 7.094686376348382, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17758 + }, + { + "epoch": 7.095085896923692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17759 + }, + { + "epoch": 7.095485417499002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17760 + }, + { + "epoch": 7.0958849380743105, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17761 + }, + { + "epoch": 7.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17762 + }, + { + "epoch": 7.09668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17763 + }, + { + "epoch": 7.09708349980024, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17764 + }, + { + "epoch": 7.09748302037555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17765 + }, + { + "epoch": 7.097882540950859, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17766 + }, + { + "epoch": 7.098282061526168, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17767 + }, + { + "epoch": 7.098681582101478, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17768 + }, + { + "epoch": 7.099081102676788, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17769 + }, + { + "epoch": 7.099480623252098, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17770 + }, + { + "epoch": 7.0998801438274075, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17771 + }, + { + "epoch": 7.100279664402716, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17772 + }, + { + "epoch": 7.100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17773 + }, + { + "epoch": 7.101078705553336, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17774 + }, + { + "epoch": 7.101478226128646, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17775 + }, + { + "epoch": 7.1018777467039556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17776 + }, + { + "epoch": 7.1022772672792644, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17777 + }, + { + "epoch": 7.102676787854574, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17778 + }, + { + "epoch": 7.103076308429884, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17779 + }, + { + "epoch": 7.103475829005194, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17780 + }, + { + "epoch": 7.103875349580504, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17781 + }, + { + "epoch": 7.104274870155813, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17782 + }, + { + "epoch": 7.104674390731122, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17783 + }, + { + "epoch": 7.105073911306432, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17784 + }, + { + "epoch": 7.105473431881742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 17785 + }, + { + "epoch": 7.105872952457052, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17786 + }, + { + "epoch": 7.106272473032361, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17787 + }, + { + "epoch": 7.106671993607671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17788 + }, + { + "epoch": 7.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17789 + }, + { + "epoch": 7.10747103475829, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17790 + }, + { + "epoch": 7.1078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17791 + }, + { + "epoch": 7.1082700759089095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17792 + }, + { + "epoch": 7.108669596484219, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17793 + }, + { + "epoch": 7.109069117059528, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17794 + }, + { + "epoch": 7.109468637634838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17795 + }, + { + "epoch": 7.109868158210148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17796 + }, + { + "epoch": 7.1102676787854575, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17797 + }, + { + "epoch": 7.110667199360767, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17798 + }, + { + "epoch": 7.111066719936077, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17799 + }, + { + "epoch": 7.111466240511386, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17800 + }, + { + "epoch": 7.111865761086696, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17801 + }, + { + "epoch": 7.112265281662006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 17802 + }, + { + "epoch": 7.112664802237315, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17803 + }, + { + "epoch": 7.113064322812625, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17804 + }, + { + "epoch": 7.113463843387935, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17805 + }, + { + "epoch": 7.113863363963244, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17806 + }, + { + "epoch": 7.114262884538554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17807 + }, + { + "epoch": 7.114662405113863, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17808 + }, + { + "epoch": 7.115061925689173, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17809 + }, + { + "epoch": 7.115461446264483, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17810 + }, + { + "epoch": 7.115860966839792, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17811 + }, + { + "epoch": 7.116260487415102, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 17812 + }, + { + "epoch": 7.116660007990411, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17813 + }, + { + "epoch": 7.117059528565721, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17814 + }, + { + "epoch": 7.117459049141031, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17815 + }, + { + "epoch": 7.117858569716341, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17816 + }, + { + "epoch": 7.11825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17817 + }, + { + "epoch": 7.1186576108669595, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 17818 + }, + { + "epoch": 7.119057131442269, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17819 + }, + { + "epoch": 7.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17820 + }, + { + "epoch": 7.119856172592889, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17821 + }, + { + "epoch": 7.120255693168198, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17822 + }, + { + "epoch": 7.1206552137435075, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17823 + }, + { + "epoch": 7.121054734318817, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17824 + }, + { + "epoch": 7.121454254894127, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17825 + }, + { + "epoch": 7.121853775469437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17826 + }, + { + "epoch": 7.122253296044747, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17827 + }, + { + "epoch": 7.122652816620056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17828 + }, + { + "epoch": 7.123052337195365, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 17829 + }, + { + "epoch": 7.123451857770675, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 17830 + }, + { + "epoch": 7.123851378345985, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17831 + }, + { + "epoch": 7.124250898921295, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17832 + }, + { + "epoch": 7.1246504194966045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17833 + }, + { + "epoch": 7.125049940071913, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17834 + }, + { + "epoch": 7.125449460647223, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17835 + }, + { + "epoch": 7.125848981222533, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17836 + }, + { + "epoch": 7.126248501797843, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17837 + }, + { + "epoch": 7.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17838 + }, + { + "epoch": 7.127047542948461, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17839 + }, + { + "epoch": 7.127447063523771, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17840 + }, + { + "epoch": 7.127846584099081, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17841 + }, + { + "epoch": 7.128246104674391, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17842 + }, + { + "epoch": 7.128645625249701, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17843 + }, + { + "epoch": 7.12904514582501, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17844 + }, + { + "epoch": 7.129444666400319, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 17845 + }, + { + "epoch": 7.129844186975629, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17846 + }, + { + "epoch": 7.130243707550939, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17847 + }, + { + "epoch": 7.130643228126249, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17848 + }, + { + "epoch": 7.131042748701558, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17849 + }, + { + "epoch": 7.131442269276867, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17850 + }, + { + "epoch": 7.131841789852177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17851 + }, + { + "epoch": 7.132241310427487, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17852 + }, + { + "epoch": 7.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17853 + }, + { + "epoch": 7.1330403515781065, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17854 + }, + { + "epoch": 7.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17855 + }, + { + "epoch": 7.133839392728725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17856 + }, + { + "epoch": 7.134238913304035, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17857 + }, + { + "epoch": 7.134638433879345, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17858 + }, + { + "epoch": 7.1350379544546545, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17859 + }, + { + "epoch": 7.135437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17860 + }, + { + "epoch": 7.135836995605274, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17861 + }, + { + "epoch": 7.136236516180583, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17862 + }, + { + "epoch": 7.136636036755893, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17863 + }, + { + "epoch": 7.1370355573312025, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17864 + }, + { + "epoch": 7.137435077906512, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17865 + }, + { + "epoch": 7.137834598481822, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17866 + }, + { + "epoch": 7.138234119057131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 17867 + }, + { + "epoch": 7.138633639632441, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17868 + }, + { + "epoch": 7.139033160207751, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17869 + }, + { + "epoch": 7.13943268078306, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17870 + }, + { + "epoch": 7.13983220135837, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17871 + }, + { + "epoch": 7.14023172193368, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17872 + }, + { + "epoch": 7.140631242508989, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17873 + }, + { + "epoch": 7.141030763084299, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17874 + }, + { + "epoch": 7.141430283659608, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17875 + }, + { + "epoch": 7.141829804234918, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17876 + }, + { + "epoch": 7.142229324810228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17877 + }, + { + "epoch": 7.142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17878 + }, + { + "epoch": 7.143028365960847, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17879 + }, + { + "epoch": 7.1434278865361565, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17880 + }, + { + "epoch": 7.143827407111466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17881 + }, + { + "epoch": 7.144226927686776, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17882 + }, + { + "epoch": 7.144626448262086, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17883 + }, + { + "epoch": 7.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17884 + }, + { + "epoch": 7.1454254894127045, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17885 + }, + { + "epoch": 7.145825009988014, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 17886 + }, + { + "epoch": 7.146224530563324, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17887 + }, + { + "epoch": 7.146624051138634, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17888 + }, + { + "epoch": 7.147023571713944, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17889 + }, + { + "epoch": 7.1474230922892525, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17890 + }, + { + "epoch": 7.147822612864562, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17891 + }, + { + "epoch": 7.148222133439872, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17892 + }, + { + "epoch": 7.148621654015182, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 17893 + }, + { + "epoch": 7.149021174590492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17894 + }, + { + "epoch": 7.1494206951658015, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17895 + }, + { + "epoch": 7.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17896 + }, + { + "epoch": 7.15021973631642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17897 + }, + { + "epoch": 7.15061925689173, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17898 + }, + { + "epoch": 7.15101877746704, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17899 + }, + { + "epoch": 7.1514182980423495, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17900 + }, + { + "epoch": 7.151817818617658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17901 + }, + { + "epoch": 7.152217339192968, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17902 + }, + { + "epoch": 7.152616859768278, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17903 + }, + { + "epoch": 7.153016380343588, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17904 + }, + { + "epoch": 7.153415900918898, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17905 + }, + { + "epoch": 7.153815421494207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17906 + }, + { + "epoch": 7.154214942069516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 17907 + }, + { + "epoch": 7.154614462644826, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17908 + }, + { + "epoch": 7.155013983220136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17909 + }, + { + "epoch": 7.155413503795446, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17910 + }, + { + "epoch": 7.155813024370755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17911 + }, + { + "epoch": 7.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17912 + }, + { + "epoch": 7.156612065521374, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17913 + }, + { + "epoch": 7.157011586096684, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17914 + }, + { + "epoch": 7.157411106671994, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17915 + }, + { + "epoch": 7.157810627247303, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17916 + }, + { + "epoch": 7.158210147822613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17917 + }, + { + "epoch": 7.158609668397922, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17918 + }, + { + "epoch": 7.159009188973232, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17919 + }, + { + "epoch": 7.159408709548542, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17920 + }, + { + "epoch": 7.1598082301238515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17921 + }, + { + "epoch": 7.160207750699161, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17922 + }, + { + "epoch": 7.160607271274471, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17923 + }, + { + "epoch": 7.16100679184978, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17924 + }, + { + "epoch": 7.16140631242509, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17925 + }, + { + "epoch": 7.1618058330003995, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17926 + }, + { + "epoch": 7.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17927 + }, + { + "epoch": 7.162604874151019, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17928 + }, + { + "epoch": 7.163004394726328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17929 + }, + { + "epoch": 7.163403915301638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17930 + }, + { + "epoch": 7.163803435876948, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17931 + }, + { + "epoch": 7.164202956452257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17932 + }, + { + "epoch": 7.164602477027567, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17933 + }, + { + "epoch": 7.165001997602877, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17934 + }, + { + "epoch": 7.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17935 + }, + { + "epoch": 7.165801038753496, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17936 + }, + { + "epoch": 7.166200559328805, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17937 + }, + { + "epoch": 7.166600079904115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17938 + }, + { + "epoch": 7.166999600479425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17939 + }, + { + "epoch": 7.167399121054734, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17940 + }, + { + "epoch": 7.167798641630044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17941 + }, + { + "epoch": 7.168198162205353, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17942 + }, + { + "epoch": 7.168597682780663, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17943 + }, + { + "epoch": 7.168997203355973, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17944 + }, + { + "epoch": 7.169396723931283, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 17945 + }, + { + "epoch": 7.169796244506592, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17946 + }, + { + "epoch": 7.1701957650819015, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17947 + }, + { + "epoch": 7.170595285657211, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17948 + }, + { + "epoch": 7.170994806232521, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17949 + }, + { + "epoch": 7.171394326807831, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17950 + }, + { + "epoch": 7.171793847383141, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17951 + }, + { + "epoch": 7.1721933679584495, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17952 + }, + { + "epoch": 7.172592888533759, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17953 + }, + { + "epoch": 7.172992409109069, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17954 + }, + { + "epoch": 7.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17955 + }, + { + "epoch": 7.173791450259689, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17956 + }, + { + "epoch": 7.174190970834998, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17957 + }, + { + "epoch": 7.174590491410307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17958 + }, + { + "epoch": 7.174990011985617, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17959 + }, + { + "epoch": 7.175389532560927, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17960 + }, + { + "epoch": 7.175789053136237, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17961 + }, + { + "epoch": 7.1761885737115465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17962 + }, + { + "epoch": 7.176588094286855, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17963 + }, + { + "epoch": 7.176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17964 + }, + { + "epoch": 7.177387135437475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17965 + }, + { + "epoch": 7.177786656012785, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17966 + }, + { + "epoch": 7.1781861765880945, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17967 + }, + { + "epoch": 7.178585697163404, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17968 + }, + { + "epoch": 7.178985217738713, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17969 + }, + { + "epoch": 7.179384738314023, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17970 + }, + { + "epoch": 7.179784258889333, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17971 + }, + { + "epoch": 7.180183779464643, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17972 + }, + { + "epoch": 7.180583300039952, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17973 + }, + { + "epoch": 7.180982820615261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17974 + }, + { + "epoch": 7.181382341190571, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17975 + }, + { + "epoch": 7.181781861765881, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17976 + }, + { + "epoch": 7.182181382341191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17977 + }, + { + "epoch": 7.1825809029165, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17978 + }, + { + "epoch": 7.18298042349181, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17979 + }, + { + "epoch": 7.183379944067119, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17980 + }, + { + "epoch": 7.183779464642429, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17981 + }, + { + "epoch": 7.184178985217739, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17982 + }, + { + "epoch": 7.1845785057930485, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17983 + }, + { + "epoch": 7.184978026368358, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17984 + }, + { + "epoch": 7.185377546943668, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17985 + }, + { + "epoch": 7.185777067518977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17986 + }, + { + "epoch": 7.186176588094287, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17987 + }, + { + "epoch": 7.1865761086695965, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17988 + }, + { + "epoch": 7.186975629244906, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17989 + }, + { + "epoch": 7.187375149820216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17990 + }, + { + "epoch": 7.187774670395525, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17991 + }, + { + "epoch": 7.188174190970835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17992 + }, + { + "epoch": 7.1885737115461446, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17993 + }, + { + "epoch": 7.188973232121454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17994 + }, + { + "epoch": 7.189372752696764, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17995 + }, + { + "epoch": 7.189772273272074, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17996 + }, + { + "epoch": 7.190171793847383, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17997 + }, + { + "epoch": 7.190571314422693, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17998 + }, + { + "epoch": 7.190970834998002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17999 + }, + { + "epoch": 7.191370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 18000 + }, + { + "epoch": 7.191769876148622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18001 + }, + { + "epoch": 7.192169396723932, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18002 + }, + { + "epoch": 7.192568917299241, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18003 + }, + { + "epoch": 7.19296843787455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18004 + }, + { + "epoch": 7.19336795844986, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18005 + }, + { + "epoch": 7.19376747902517, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18006 + }, + { + "epoch": 7.19416699960048, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18007 + }, + { + "epoch": 7.194566520175789, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18008 + }, + { + "epoch": 7.1949660407510985, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18009 + }, + { + "epoch": 7.195365561326408, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18010 + }, + { + "epoch": 7.195765081901718, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18011 + }, + { + "epoch": 7.196164602477028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18012 + }, + { + "epoch": 7.196564123052338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18013 + }, + { + "epoch": 7.1969636436276465, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18014 + }, + { + "epoch": 7.197363164202956, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18015 + }, + { + "epoch": 7.197762684778266, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18016 + }, + { + "epoch": 7.198162205353576, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18017 + }, + { + "epoch": 7.198561725928886, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18018 + }, + { + "epoch": 7.1989612465041946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18019 + }, + { + "epoch": 7.199360767079504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18020 + }, + { + "epoch": 7.199760287654814, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18021 + }, + { + "epoch": 7.200159808230124, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18022 + }, + { + "epoch": 7.200559328805434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18023 + }, + { + "epoch": 7.2009588493807435, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18024 + }, + { + "epoch": 7.201358369956052, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18025 + }, + { + "epoch": 7.201757890531362, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18026 + }, + { + "epoch": 7.202157411106672, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18027 + }, + { + "epoch": 7.202556931681982, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18028 + }, + { + "epoch": 7.2029564522572915, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18029 + }, + { + "epoch": 7.2033559728326, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18030 + }, + { + "epoch": 7.20375549340791, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18031 + }, + { + "epoch": 7.20415501398322, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18032 + }, + { + "epoch": 7.20455453455853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18033 + }, + { + "epoch": 7.20495405513384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 18034 + }, + { + "epoch": 7.205353575709149, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18035 + }, + { + "epoch": 7.205753096284458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18036 + }, + { + "epoch": 7.206152616859768, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18037 + }, + { + "epoch": 7.206552137435078, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18038 + }, + { + "epoch": 7.206951658010388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18039 + }, + { + "epoch": 7.207351178585697, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18040 + }, + { + "epoch": 7.207750699161007, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18041 + }, + { + "epoch": 7.208150219736316, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18042 + }, + { + "epoch": 7.208549740311626, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18043 + }, + { + "epoch": 7.208949260886936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18044 + }, + { + "epoch": 7.2093487814622454, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18045 + }, + { + "epoch": 7.209748302037555, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18046 + }, + { + "epoch": 7.210147822612864, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18047 + }, + { + "epoch": 7.210547343188174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18048 + }, + { + "epoch": 7.210946863763484, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18049 + }, + { + "epoch": 7.2113463843387935, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18050 + }, + { + "epoch": 7.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18051 + }, + { + "epoch": 7.212145425489413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18052 + }, + { + "epoch": 7.212544946064722, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18053 + }, + { + "epoch": 7.212944466640032, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18054 + }, + { + "epoch": 7.2133439872153415, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18055 + }, + { + "epoch": 7.213743507790651, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18056 + }, + { + "epoch": 7.214143028365961, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 18057 + }, + { + "epoch": 7.214542548941271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18058 + }, + { + "epoch": 7.21494206951658, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18059 + }, + { + "epoch": 7.21534159009189, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18060 + }, + { + "epoch": 7.215741110667199, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18061 + }, + { + "epoch": 7.216140631242509, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18062 + }, + { + "epoch": 7.216540151817819, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18063 + }, + { + "epoch": 7.216939672393128, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18064 + }, + { + "epoch": 7.217339192968438, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18065 + }, + { + "epoch": 7.217738713543747, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18066 + }, + { + "epoch": 7.218138234119057, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18067 + }, + { + "epoch": 7.218537754694367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18068 + }, + { + "epoch": 7.218937275269677, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18069 + }, + { + "epoch": 7.219336795844986, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18070 + }, + { + "epoch": 7.2197363164202955, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18071 + }, + { + "epoch": 7.220135836995605, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18072 + }, + { + "epoch": 7.220535357570915, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18073 + }, + { + "epoch": 7.220934878146225, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18074 + }, + { + "epoch": 7.221334398721535, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18075 + }, + { + "epoch": 7.2217339192968435, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18076 + }, + { + "epoch": 7.222133439872153, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18077 + }, + { + "epoch": 7.222532960447463, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 18078 + }, + { + "epoch": 7.222932481022773, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18079 + }, + { + "epoch": 7.223332001598083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18080 + }, + { + "epoch": 7.2237315221733915, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18081 + }, + { + "epoch": 7.224131042748701, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 18082 + }, + { + "epoch": 7.224530563324011, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18083 + }, + { + "epoch": 7.224930083899321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18084 + }, + { + "epoch": 7.225329604474631, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18085 + }, + { + "epoch": 7.2257291250499405, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18086 + }, + { + "epoch": 7.226128645625249, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18087 + }, + { + "epoch": 7.226528166200559, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18088 + }, + { + "epoch": 7.226927686775869, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18089 + }, + { + "epoch": 7.227327207351179, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18090 + }, + { + "epoch": 7.2277267279264885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18091 + }, + { + "epoch": 7.228126248501797, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18092 + }, + { + "epoch": 7.228525769077107, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18093 + }, + { + "epoch": 7.228925289652417, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18094 + }, + { + "epoch": 7.229324810227727, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18095 + }, + { + "epoch": 7.229724330803037, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18096 + }, + { + "epoch": 7.230123851378346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18097 + }, + { + "epoch": 7.230523371953655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18098 + }, + { + "epoch": 7.230922892528965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18099 + }, + { + "epoch": 7.231322413104275, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18100 + }, + { + "epoch": 7.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18101 + }, + { + "epoch": 7.232121454254894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18102 + }, + { + "epoch": 7.232520974830204, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18103 + }, + { + "epoch": 7.232920495405513, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18104 + }, + { + "epoch": 7.233320015980823, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18105 + }, + { + "epoch": 7.233719536556133, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18106 + }, + { + "epoch": 7.234119057131442, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18107 + }, + { + "epoch": 7.234518577706752, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18108 + }, + { + "epoch": 7.234918098282061, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18109 + }, + { + "epoch": 7.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18110 + }, + { + "epoch": 7.235717139432681, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18111 + }, + { + "epoch": 7.2361166600079905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18112 + }, + { + "epoch": 7.2365161805833, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18113 + }, + { + "epoch": 7.23691570115861, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18114 + }, + { + "epoch": 7.237315221733919, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18115 + }, + { + "epoch": 7.237714742309229, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 18116 + }, + { + "epoch": 7.2381142628845385, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18117 + }, + { + "epoch": 7.238513783459848, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18118 + }, + { + "epoch": 7.238913304035158, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18119 + }, + { + "epoch": 7.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18120 + }, + { + "epoch": 7.239712345185777, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18121 + }, + { + "epoch": 7.240111865761087, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18122 + }, + { + "epoch": 7.240511386336396, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18123 + }, + { + "epoch": 7.240910906911706, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18124 + }, + { + "epoch": 7.241310427487016, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18125 + }, + { + "epoch": 7.241709948062325, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18126 + }, + { + "epoch": 7.242109468637635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18127 + }, + { + "epoch": 7.242508989212944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18128 + }, + { + "epoch": 7.242908509788254, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18129 + }, + { + "epoch": 7.243308030363564, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 18130 + }, + { + "epoch": 7.243707550938874, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18131 + }, + { + "epoch": 7.244107071514183, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18132 + }, + { + "epoch": 7.244506592089492, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18133 + }, + { + "epoch": 7.244906112664802, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18134 + }, + { + "epoch": 7.245305633240112, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 18135 + }, + { + "epoch": 7.245705153815422, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 18136 + }, + { + "epoch": 7.246104674390731, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18137 + }, + { + "epoch": 7.2465041949660405, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18138 + }, + { + "epoch": 7.24690371554135, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18139 + }, + { + "epoch": 7.24730323611666, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18140 + }, + { + "epoch": 7.24770275669197, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18141 + }, + { + "epoch": 7.24810227726728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18142 + }, + { + "epoch": 7.2485017978425885, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18143 + }, + { + "epoch": 7.248901318417898, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18144 + }, + { + "epoch": 7.249300838993208, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18145 + }, + { + "epoch": 7.249700359568518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18146 + }, + { + "epoch": 7.250099880143828, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18147 + }, + { + "epoch": 7.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18148 + }, + { + "epoch": 7.250898921294446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18149 + }, + { + "epoch": 7.251298441869756, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18150 + }, + { + "epoch": 7.251697962445066, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18151 + }, + { + "epoch": 7.252097483020376, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18152 + }, + { + "epoch": 7.2524970035956855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18153 + }, + { + "epoch": 7.252896524170994, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18154 + }, + { + "epoch": 7.253296044746304, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18155 + }, + { + "epoch": 7.253695565321614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18156 + }, + { + "epoch": 7.254095085896924, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18157 + }, + { + "epoch": 7.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18158 + }, + { + "epoch": 7.254894127047543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18159 + }, + { + "epoch": 7.255293647622852, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18160 + }, + { + "epoch": 7.255693168198162, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18161 + }, + { + "epoch": 7.256092688773472, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18162 + }, + { + "epoch": 7.256492209348782, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18163 + }, + { + "epoch": 7.256891729924091, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18164 + }, + { + "epoch": 7.257291250499401, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 18165 + }, + { + "epoch": 7.25769077107471, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18166 + }, + { + "epoch": 7.25809029165002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18167 + }, + { + "epoch": 7.25848981222533, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18168 + }, + { + "epoch": 7.258889332800639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18169 + }, + { + "epoch": 7.259288853375949, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 18170 + }, + { + "epoch": 7.259688373951258, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18171 + }, + { + "epoch": 7.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18172 + }, + { + "epoch": 7.260487415101878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18173 + }, + { + "epoch": 7.2608869356771875, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 18174 + }, + { + "epoch": 7.261286456252497, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18175 + }, + { + "epoch": 7.261685976827807, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18176 + }, + { + "epoch": 7.262085497403116, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18177 + }, + { + "epoch": 7.262485017978426, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18178 + }, + { + "epoch": 7.2628845385537355, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18179 + }, + { + "epoch": 7.263284059129045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18180 + }, + { + "epoch": 7.263683579704355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18181 + }, + { + "epoch": 7.264083100279665, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18182 + }, + { + "epoch": 7.264482620854974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18183 + }, + { + "epoch": 7.2648821414302835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18184 + }, + { + "epoch": 7.265281662005593, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18185 + }, + { + "epoch": 7.265681182580903, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18186 + }, + { + "epoch": 7.266080703156213, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18187 + }, + { + "epoch": 7.266480223731522, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 18188 + }, + { + "epoch": 7.266879744306832, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18189 + }, + { + "epoch": 7.267279264882141, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18190 + }, + { + "epoch": 7.267678785457451, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18191 + }, + { + "epoch": 7.268078306032761, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18192 + }, + { + "epoch": 7.26847782660807, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 18193 + }, + { + "epoch": 7.26887734718338, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18194 + }, + { + "epoch": 7.269276867758689, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18195 + }, + { + "epoch": 7.269676388333999, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18196 + }, + { + "epoch": 7.270075908909309, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18197 + }, + { + "epoch": 7.270475429484619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18198 + }, + { + "epoch": 7.270874950059928, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18199 + }, + { + "epoch": 7.2712744706352375, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18200 + }, + { + "epoch": 7.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18201 + }, + { + "epoch": 7.272073511785857, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 18202 + }, + { + "epoch": 7.272473032361167, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18203 + }, + { + "epoch": 7.272872552936477, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18204 + }, + { + "epoch": 7.2732720735117855, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18205 + }, + { + "epoch": 7.273671594087095, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18206 + }, + { + "epoch": 7.274071114662405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18207 + }, + { + "epoch": 7.274470635237715, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18208 + }, + { + "epoch": 7.274870155813025, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18209 + }, + { + "epoch": 7.2752696763883336, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18210 + }, + { + "epoch": 7.275669196963643, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18211 + }, + { + "epoch": 7.276068717538953, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2981, + "step": 18212 + }, + { + "epoch": 7.276468238114263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18213 + }, + { + "epoch": 7.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18214 + }, + { + "epoch": 7.2772672792648825, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18215 + }, + { + "epoch": 7.277666799840191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18216 + }, + { + "epoch": 7.278066320415501, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18217 + }, + { + "epoch": 7.278465840990811, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18218 + }, + { + "epoch": 7.278865361566121, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18219 + }, + { + "epoch": 7.2792648821414305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18220 + }, + { + "epoch": 7.27966440271674, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18221 + }, + { + "epoch": 7.280063923292049, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 18222 + }, + { + "epoch": 7.280463443867359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18223 + }, + { + "epoch": 7.280862964442669, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18224 + }, + { + "epoch": 7.281262485017979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18225 + }, + { + "epoch": 7.281662005593288, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18226 + }, + { + "epoch": 7.282061526168597, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18227 + }, + { + "epoch": 7.282461046743907, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18228 + }, + { + "epoch": 7.282860567319217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 18229 + }, + { + "epoch": 7.283260087894527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18230 + }, + { + "epoch": 7.283659608469836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 18231 + }, + { + "epoch": 7.284059129045146, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18232 + }, + { + "epoch": 7.284458649620455, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18233 + }, + { + "epoch": 7.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18234 + }, + { + "epoch": 7.285257690771075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18235 + }, + { + "epoch": 7.285657211346384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18236 + }, + { + "epoch": 7.286056731921694, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18237 + }, + { + "epoch": 7.286456252497004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18238 + }, + { + "epoch": 7.286855773072313, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18239 + }, + { + "epoch": 7.287255293647623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18240 + }, + { + "epoch": 7.2876548142229325, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18241 + }, + { + "epoch": 7.288054334798242, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18242 + }, + { + "epoch": 7.288453855373552, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18243 + }, + { + "epoch": 7.288853375948861, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18244 + }, + { + "epoch": 7.289252896524171, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18245 + }, + { + "epoch": 7.2896524170994805, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18246 + }, + { + "epoch": 7.29005193767479, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18247 + }, + { + "epoch": 7.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18248 + }, + { + "epoch": 7.29085097882541, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18249 + }, + { + "epoch": 7.291250499400719, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18250 + }, + { + "epoch": 7.291650019976029, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18251 + }, + { + "epoch": 7.292049540551338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18252 + }, + { + "epoch": 7.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18253 + }, + { + "epoch": 7.292848581701958, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18254 + }, + { + "epoch": 7.293248102277268, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18255 + }, + { + "epoch": 7.293647622852577, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18256 + }, + { + "epoch": 7.294047143427886, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18257 + }, + { + "epoch": 7.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18258 + }, + { + "epoch": 7.294846184578506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18259 + }, + { + "epoch": 7.295245705153816, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18260 + }, + { + "epoch": 7.295645225729125, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18261 + }, + { + "epoch": 7.2960447463044344, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18262 + }, + { + "epoch": 7.296444266879744, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18263 + }, + { + "epoch": 7.296843787455054, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18264 + }, + { + "epoch": 7.297243308030364, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18265 + }, + { + "epoch": 7.297642828605674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18266 + }, + { + "epoch": 7.2980423491809825, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18267 + }, + { + "epoch": 7.298441869756292, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 18268 + }, + { + "epoch": 7.298841390331602, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18269 + }, + { + "epoch": 7.299240910906912, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18270 + }, + { + "epoch": 7.299640431482222, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18271 + }, + { + "epoch": 7.300039952057531, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18272 + }, + { + "epoch": 7.30043947263284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18273 + }, + { + "epoch": 7.30083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18274 + }, + { + "epoch": 7.30123851378346, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18275 + }, + { + "epoch": 7.30163803435877, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18276 + }, + { + "epoch": 7.3020375549340795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18277 + }, + { + "epoch": 7.302437075509388, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18278 + }, + { + "epoch": 7.302836596084698, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18279 + }, + { + "epoch": 7.303236116660008, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18280 + }, + { + "epoch": 7.303635637235318, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18281 + }, + { + "epoch": 7.3040351578106275, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18282 + }, + { + "epoch": 7.304434678385936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18283 + }, + { + "epoch": 7.304834198961246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 18284 + }, + { + "epoch": 7.305233719536556, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18285 + }, + { + "epoch": 7.305633240111866, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18286 + }, + { + "epoch": 7.3060327606871756, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18287 + }, + { + "epoch": 7.306432281262485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18288 + }, + { + "epoch": 7.306831801837794, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18289 + }, + { + "epoch": 7.307231322413104, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18290 + }, + { + "epoch": 7.307630842988414, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18291 + }, + { + "epoch": 7.308030363563724, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18292 + }, + { + "epoch": 7.308429884139033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18293 + }, + { + "epoch": 7.308829404714343, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18294 + }, + { + "epoch": 7.309228925289652, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18295 + }, + { + "epoch": 7.309628445864962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18296 + }, + { + "epoch": 7.310027966440272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18297 + }, + { + "epoch": 7.310427487015581, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18298 + }, + { + "epoch": 7.310827007590891, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18299 + }, + { + "epoch": 7.3112265281662, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18300 + }, + { + "epoch": 7.31162604874151, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18301 + }, + { + "epoch": 7.31202556931682, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18302 + }, + { + "epoch": 7.3124250898921295, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18303 + }, + { + "epoch": 7.312824610467439, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18304 + }, + { + "epoch": 7.313224131042749, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18305 + }, + { + "epoch": 7.313623651618058, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18306 + }, + { + "epoch": 7.314023172193368, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 18307 + }, + { + "epoch": 7.3144226927686775, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18308 + }, + { + "epoch": 7.314822213343987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18309 + }, + { + "epoch": 7.315221733919297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18310 + }, + { + "epoch": 7.315621254494607, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18311 + }, + { + "epoch": 7.316020775069916, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18312 + }, + { + "epoch": 7.316420295645226, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 18313 + }, + { + "epoch": 7.316819816220535, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18314 + }, + { + "epoch": 7.317219336795845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18315 + }, + { + "epoch": 7.317618857371155, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18316 + }, + { + "epoch": 7.318018377946464, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 18317 + }, + { + "epoch": 7.318417898521774, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18318 + }, + { + "epoch": 7.318817419097083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18319 + }, + { + "epoch": 7.319216939672393, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18320 + }, + { + "epoch": 7.319616460247703, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18321 + }, + { + "epoch": 7.320015980823013, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18322 + }, + { + "epoch": 7.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18323 + }, + { + "epoch": 7.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18324 + }, + { + "epoch": 7.321214542548941, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18325 + }, + { + "epoch": 7.321614063124251, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18326 + }, + { + "epoch": 7.322013583699561, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18327 + }, + { + "epoch": 7.322413104274871, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18328 + }, + { + "epoch": 7.3228126248501795, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18329 + }, + { + "epoch": 7.323212145425489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18330 + }, + { + "epoch": 7.323611666000799, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18331 + }, + { + "epoch": 7.324011186576109, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18332 + }, + { + "epoch": 7.324410707151419, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18333 + }, + { + "epoch": 7.3248102277267275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18334 + }, + { + "epoch": 7.325209748302037, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18335 + }, + { + "epoch": 7.325609268877347, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18336 + }, + { + "epoch": 7.326008789452657, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18337 + }, + { + "epoch": 7.326408310027967, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18338 + }, + { + "epoch": 7.3268078306032765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18339 + }, + { + "epoch": 7.327207351178585, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18340 + }, + { + "epoch": 7.327606871753895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18341 + }, + { + "epoch": 7.328006392329205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18342 + }, + { + "epoch": 7.328405912904515, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18343 + }, + { + "epoch": 7.3288054334798245, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18344 + }, + { + "epoch": 7.329204954055134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18345 + }, + { + "epoch": 7.329604474630443, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18346 + }, + { + "epoch": 7.330003995205753, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18347 + }, + { + "epoch": 7.330403515781063, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 18348 + }, + { + "epoch": 7.3308030363563725, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18349 + }, + { + "epoch": 7.331202556931682, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18350 + }, + { + "epoch": 7.331602077506991, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18351 + }, + { + "epoch": 7.332001598082301, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18352 + }, + { + "epoch": 7.332401118657611, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18353 + }, + { + "epoch": 7.332800639232921, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18354 + }, + { + "epoch": 7.33320015980823, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18355 + }, + { + "epoch": 7.33359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18356 + }, + { + "epoch": 7.333999200958849, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18357 + }, + { + "epoch": 7.334398721534159, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18358 + }, + { + "epoch": 7.334798242109469, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18359 + }, + { + "epoch": 7.335197762684778, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18360 + }, + { + "epoch": 7.335597283260088, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18361 + }, + { + "epoch": 7.335996803835398, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18362 + }, + { + "epoch": 7.336396324410707, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18363 + }, + { + "epoch": 7.336795844986017, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18364 + }, + { + "epoch": 7.3371953655613265, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18365 + }, + { + "epoch": 7.337594886136636, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18366 + }, + { + "epoch": 7.337994406711946, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18367 + }, + { + "epoch": 7.338393927287255, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18368 + }, + { + "epoch": 7.338793447862565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18369 + }, + { + "epoch": 7.3391929684378745, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18370 + }, + { + "epoch": 7.339592489013184, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18371 + }, + { + "epoch": 7.339992009588494, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18372 + }, + { + "epoch": 7.340391530163803, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18373 + }, + { + "epoch": 7.340791050739113, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18374 + }, + { + "epoch": 7.3411905713144225, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18375 + }, + { + "epoch": 7.341590091889732, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18376 + }, + { + "epoch": 7.341989612465042, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18377 + }, + { + "epoch": 7.342389133040352, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18378 + }, + { + "epoch": 7.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18379 + }, + { + "epoch": 7.343188174190971, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18380 + }, + { + "epoch": 7.34358769476628, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18381 + }, + { + "epoch": 7.34398721534159, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18382 + }, + { + "epoch": 7.3443867359169, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 18383 + }, + { + "epoch": 7.34478625649221, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18384 + }, + { + "epoch": 7.345185777067519, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18385 + }, + { + "epoch": 7.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18386 + }, + { + "epoch": 7.345984818218138, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18387 + }, + { + "epoch": 7.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18388 + }, + { + "epoch": 7.346783859368758, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18389 + }, + { + "epoch": 7.347183379944067, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18390 + }, + { + "epoch": 7.3475829005193765, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18391 + }, + { + "epoch": 7.347982421094686, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18392 + }, + { + "epoch": 7.348381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18393 + }, + { + "epoch": 7.348781462245306, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18394 + }, + { + "epoch": 7.349180982820616, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18395 + }, + { + "epoch": 7.3495805033959245, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18396 + }, + { + "epoch": 7.349980023971234, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18397 + }, + { + "epoch": 7.350379544546544, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18398 + }, + { + "epoch": 7.350779065121854, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18399 + }, + { + "epoch": 7.351178585697164, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18400 + }, + { + "epoch": 7.351578106272473, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18401 + }, + { + "epoch": 7.351977626847782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18402 + }, + { + "epoch": 7.352377147423092, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 18403 + }, + { + "epoch": 7.352776667998402, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18404 + }, + { + "epoch": 7.353176188573712, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18405 + }, + { + "epoch": 7.3535757091490215, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18406 + }, + { + "epoch": 7.35397522972433, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18407 + }, + { + "epoch": 7.35437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18408 + }, + { + "epoch": 7.35477427087495, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18409 + }, + { + "epoch": 7.35517379145026, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18410 + }, + { + "epoch": 7.3555733120255695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18411 + }, + { + "epoch": 7.355972832600879, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18412 + }, + { + "epoch": 7.356372353176188, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18413 + }, + { + "epoch": 7.356771873751498, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18414 + }, + { + "epoch": 7.357171394326808, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 18415 + }, + { + "epoch": 7.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18416 + }, + { + "epoch": 7.357970435477427, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18417 + }, + { + "epoch": 7.358369956052737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18418 + }, + { + "epoch": 7.358769476628046, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18419 + }, + { + "epoch": 7.359168997203356, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18420 + }, + { + "epoch": 7.359568517778666, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 18421 + }, + { + "epoch": 7.359968038353975, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18422 + }, + { + "epoch": 7.360367558929285, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 18423 + }, + { + "epoch": 7.360767079504594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18424 + }, + { + "epoch": 7.361166600079904, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18425 + }, + { + "epoch": 7.361566120655214, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18426 + }, + { + "epoch": 7.361965641230523, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18427 + }, + { + "epoch": 7.362365161805833, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18428 + }, + { + "epoch": 7.362764682381143, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18429 + }, + { + "epoch": 7.363164202956452, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18430 + }, + { + "epoch": 7.363563723531762, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18431 + }, + { + "epoch": 7.3639632441070715, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18432 + }, + { + "epoch": 7.364362764682381, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18433 + }, + { + "epoch": 7.364762285257691, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18434 + }, + { + "epoch": 7.365161805833001, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18435 + }, + { + "epoch": 7.36556132640831, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18436 + }, + { + "epoch": 7.3659608469836195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18437 + }, + { + "epoch": 7.366360367558929, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18438 + }, + { + "epoch": 7.366759888134239, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18439 + }, + { + "epoch": 7.367159408709549, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18440 + }, + { + "epoch": 7.367558929284858, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18441 + }, + { + "epoch": 7.367958449860168, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18442 + }, + { + "epoch": 7.368357970435477, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18443 + }, + { + "epoch": 7.368757491010787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18444 + }, + { + "epoch": 7.369157011586097, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18445 + }, + { + "epoch": 7.369556532161407, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18446 + }, + { + "epoch": 7.369956052736716, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18447 + }, + { + "epoch": 7.370355573312025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18448 + }, + { + "epoch": 7.370755093887335, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18449 + }, + { + "epoch": 7.371154614462645, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18450 + }, + { + "epoch": 7.371554135037955, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18451 + }, + { + "epoch": 7.3719536556132645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18452 + }, + { + "epoch": 7.372353176188573, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18453 + }, + { + "epoch": 7.372752696763883, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18454 + }, + { + "epoch": 7.373152217339193, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18455 + }, + { + "epoch": 7.373551737914503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18456 + }, + { + "epoch": 7.373951258489813, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 18457 + }, + { + "epoch": 7.3743507790651215, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18458 + }, + { + "epoch": 7.374750299640431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18459 + }, + { + "epoch": 7.375149820215741, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18460 + }, + { + "epoch": 7.375549340791051, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18461 + }, + { + "epoch": 7.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18462 + }, + { + "epoch": 7.3763483819416695, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18463 + }, + { + "epoch": 7.376747902516979, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18464 + }, + { + "epoch": 7.377147423092289, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18465 + }, + { + "epoch": 7.377546943667599, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 18466 + }, + { + "epoch": 7.377946464242909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18467 + }, + { + "epoch": 7.3783459848182185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18468 + }, + { + "epoch": 7.378745505393527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18469 + }, + { + "epoch": 7.379145025968837, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18470 + }, + { + "epoch": 7.379544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18471 + }, + { + "epoch": 7.379944067119457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18472 + }, + { + "epoch": 7.3803435876947665, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18473 + }, + { + "epoch": 7.380743108270076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18474 + }, + { + "epoch": 7.381142628845385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18475 + }, + { + "epoch": 7.381542149420695, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18476 + }, + { + "epoch": 7.381941669996005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18477 + }, + { + "epoch": 7.3823411905713145, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18478 + }, + { + "epoch": 7.382740711146624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18479 + }, + { + "epoch": 7.383140231721933, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18480 + }, + { + "epoch": 7.383539752297243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18481 + }, + { + "epoch": 7.383939272872553, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18482 + }, + { + "epoch": 7.384338793447863, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18483 + }, + { + "epoch": 7.384738314023172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18484 + }, + { + "epoch": 7.385137834598482, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18485 + }, + { + "epoch": 7.385537355173791, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18486 + }, + { + "epoch": 7.385936875749101, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 18487 + }, + { + "epoch": 7.386336396324411, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18488 + }, + { + "epoch": 7.38673591689972, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18489 + }, + { + "epoch": 7.38713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18490 + }, + { + "epoch": 7.38753495805034, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18491 + }, + { + "epoch": 7.387934478625649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18492 + }, + { + "epoch": 7.388333999200959, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18493 + }, + { + "epoch": 7.3887335197762685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18494 + }, + { + "epoch": 7.389133040351578, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18495 + }, + { + "epoch": 7.389532560926888, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18496 + }, + { + "epoch": 7.389932081502197, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18497 + }, + { + "epoch": 7.390331602077507, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18498 + }, + { + "epoch": 7.3907311226528165, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18499 + }, + { + "epoch": 7.391130643228126, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18500 + }, + { + "epoch": 7.391530163803436, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18501 + }, + { + "epoch": 7.391929684378746, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 18502 + }, + { + "epoch": 7.392329204954055, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18503 + }, + { + "epoch": 7.3927287255293646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18504 + }, + { + "epoch": 7.393128246104674, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18505 + }, + { + "epoch": 7.393527766679984, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18506 + }, + { + "epoch": 7.393927287255294, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18507 + }, + { + "epoch": 7.394326807830604, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18508 + }, + { + "epoch": 7.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18509 + }, + { + "epoch": 7.395125848981222, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18510 + }, + { + "epoch": 7.395525369556532, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18511 + }, + { + "epoch": 7.395924890131842, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18512 + }, + { + "epoch": 7.396324410707152, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18513 + }, + { + "epoch": 7.396723931282461, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 18514 + }, + { + "epoch": 7.39712345185777, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18515 + }, + { + "epoch": 7.39752297243308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18516 + }, + { + "epoch": 7.39792249300839, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18517 + }, + { + "epoch": 7.3983220135837, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18518 + }, + { + "epoch": 7.39872153415901, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18519 + }, + { + "epoch": 7.3991210547343185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 18520 + }, + { + "epoch": 7.399520575309628, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18521 + }, + { + "epoch": 7.399920095884938, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18522 + }, + { + "epoch": 7.400319616460248, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18523 + }, + { + "epoch": 7.400719137035558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18524 + }, + { + "epoch": 7.401118657610867, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18525 + }, + { + "epoch": 7.401518178186176, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18526 + }, + { + "epoch": 7.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18527 + }, + { + "epoch": 7.402317219336796, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18528 + }, + { + "epoch": 7.402716739912106, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18529 + }, + { + "epoch": 7.4031162604874154, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18530 + }, + { + "epoch": 7.403515781062724, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 18531 + }, + { + "epoch": 7.403915301638034, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18532 + }, + { + "epoch": 7.404314822213344, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18533 + }, + { + "epoch": 7.404714342788654, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18534 + }, + { + "epoch": 7.4051138633639635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 18535 + }, + { + "epoch": 7.405513383939273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18536 + }, + { + "epoch": 7.405912904514582, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18537 + }, + { + "epoch": 7.406312425089892, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18538 + }, + { + "epoch": 7.406711945665202, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18539 + }, + { + "epoch": 7.4071114662405115, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18540 + }, + { + "epoch": 7.407510986815821, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18541 + }, + { + "epoch": 7.407910507391131, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18542 + }, + { + "epoch": 7.40831002796644, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18543 + }, + { + "epoch": 7.40870954854175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18544 + }, + { + "epoch": 7.40910906911706, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18545 + }, + { + "epoch": 7.409508589692369, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18546 + }, + { + "epoch": 7.409908110267679, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18547 + }, + { + "epoch": 7.410307630842988, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18548 + }, + { + "epoch": 7.410707151418298, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18549 + }, + { + "epoch": 7.411106671993608, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18550 + }, + { + "epoch": 7.411506192568917, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18551 + }, + { + "epoch": 7.411905713144227, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18552 + }, + { + "epoch": 7.412305233719536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18553 + }, + { + "epoch": 7.412704754294846, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18554 + }, + { + "epoch": 7.413104274870156, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18555 + }, + { + "epoch": 7.4135037954454654, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18556 + }, + { + "epoch": 7.413903316020775, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18557 + }, + { + "epoch": 7.414302836596085, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18558 + }, + { + "epoch": 7.414702357171394, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18559 + }, + { + "epoch": 7.415101877746704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18560 + }, + { + "epoch": 7.4155013983220135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18561 + }, + { + "epoch": 7.415900918897323, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18562 + }, + { + "epoch": 7.416300439472633, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18563 + }, + { + "epoch": 7.416699960047943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18564 + }, + { + "epoch": 7.417099480623252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18565 + }, + { + "epoch": 7.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.2993, + "step": 18566 + }, + { + "epoch": 7.417898521773871, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18567 + }, + { + "epoch": 7.418298042349181, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18568 + }, + { + "epoch": 7.418697562924491, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18569 + }, + { + "epoch": 7.4190970834998, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18570 + }, + { + "epoch": 7.41949660407511, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18571 + }, + { + "epoch": 7.419896124650419, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18572 + }, + { + "epoch": 7.420295645225729, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18573 + }, + { + "epoch": 7.420695165801039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18574 + }, + { + "epoch": 7.421094686376349, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18575 + }, + { + "epoch": 7.421494206951658, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18576 + }, + { + "epoch": 7.421893727526967, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18577 + }, + { + "epoch": 7.422293248102277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18578 + }, + { + "epoch": 7.422692768677587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18579 + }, + { + "epoch": 7.423092289252897, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18580 + }, + { + "epoch": 7.4234918098282066, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 18581 + }, + { + "epoch": 7.4238913304035155, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18582 + }, + { + "epoch": 7.424290850978825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18583 + }, + { + "epoch": 7.424690371554135, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18584 + }, + { + "epoch": 7.425089892129445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18585 + }, + { + "epoch": 7.425489412704755, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18586 + }, + { + "epoch": 7.4258889332800635, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18587 + }, + { + "epoch": 7.426288453855373, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18588 + }, + { + "epoch": 7.426687974430683, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18589 + }, + { + "epoch": 7.427087495005993, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18590 + }, + { + "epoch": 7.427487015581303, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18591 + }, + { + "epoch": 7.427886536156612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18592 + }, + { + "epoch": 7.428286056731921, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18593 + }, + { + "epoch": 7.428685577307231, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18594 + }, + { + "epoch": 7.429085097882541, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18595 + }, + { + "epoch": 7.429484618457851, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18596 + }, + { + "epoch": 7.4298841390331605, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18597 + }, + { + "epoch": 7.43028365960847, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18598 + }, + { + "epoch": 7.430683180183779, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18599 + }, + { + "epoch": 7.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18600 + }, + { + "epoch": 7.431482221334399, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18601 + }, + { + "epoch": 7.4318817419097085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18602 + }, + { + "epoch": 7.432281262485018, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18603 + }, + { + "epoch": 7.432680783060327, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18604 + }, + { + "epoch": 7.433080303635637, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18605 + }, + { + "epoch": 7.433479824210947, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18606 + }, + { + "epoch": 7.433879344786257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18607 + }, + { + "epoch": 7.434278865361566, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18608 + }, + { + "epoch": 7.434678385936876, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18609 + }, + { + "epoch": 7.435077906512185, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 18610 + }, + { + "epoch": 7.435477427087495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18611 + }, + { + "epoch": 7.435876947662805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18612 + }, + { + "epoch": 7.436276468238114, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 18613 + }, + { + "epoch": 7.436675988813424, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18614 + }, + { + "epoch": 7.437075509388734, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18615 + }, + { + "epoch": 7.437475029964043, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18616 + }, + { + "epoch": 7.437874550539353, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18617 + }, + { + "epoch": 7.438274071114662, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18618 + }, + { + "epoch": 7.438673591689972, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18619 + }, + { + "epoch": 7.439073112265282, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18620 + }, + { + "epoch": 7.439472632840591, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18621 + }, + { + "epoch": 7.439872153415901, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18622 + }, + { + "epoch": 7.4402716739912105, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18623 + }, + { + "epoch": 7.44067119456652, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18624 + }, + { + "epoch": 7.44107071514183, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3015, + "step": 18625 + }, + { + "epoch": 7.44147023571714, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18626 + }, + { + "epoch": 7.441869756292449, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18627 + }, + { + "epoch": 7.4422692768677585, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18628 + }, + { + "epoch": 7.442668797443068, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18629 + }, + { + "epoch": 7.443068318018378, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18630 + }, + { + "epoch": 7.443467838593688, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18631 + }, + { + "epoch": 7.443867359168998, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 18632 + }, + { + "epoch": 7.444266879744307, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18633 + }, + { + "epoch": 7.444666400319616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 18634 + }, + { + "epoch": 7.445065920894926, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 18635 + }, + { + "epoch": 7.445465441470236, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18636 + }, + { + "epoch": 7.445864962045546, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18637 + }, + { + "epoch": 7.446264482620855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18638 + }, + { + "epoch": 7.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18639 + }, + { + "epoch": 7.447063523771474, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 18640 + }, + { + "epoch": 7.447463044346784, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18641 + }, + { + "epoch": 7.447862564922094, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18642 + }, + { + "epoch": 7.448262085497403, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18643 + }, + { + "epoch": 7.448661606072712, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18644 + }, + { + "epoch": 7.449061126648022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18645 + }, + { + "epoch": 7.449460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18646 + }, + { + "epoch": 7.449860167798642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18647 + }, + { + "epoch": 7.450259688373952, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18648 + }, + { + "epoch": 7.4506592089492605, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18649 + }, + { + "epoch": 7.45105872952457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18650 + }, + { + "epoch": 7.45145825009988, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18651 + }, + { + "epoch": 7.45185777067519, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18652 + }, + { + "epoch": 7.4522572912505, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18653 + }, + { + "epoch": 7.452656811825809, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18654 + }, + { + "epoch": 7.453056332401118, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18655 + }, + { + "epoch": 7.453455852976428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18656 + }, + { + "epoch": 7.453855373551738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 18657 + }, + { + "epoch": 7.454254894127048, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18658 + }, + { + "epoch": 7.4546544147023575, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18659 + }, + { + "epoch": 7.455053935277666, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18660 + }, + { + "epoch": 7.455453455852976, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18661 + }, + { + "epoch": 7.455852976428286, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18662 + }, + { + "epoch": 7.456252497003596, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18663 + }, + { + "epoch": 7.4566520175789055, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18664 + }, + { + "epoch": 7.457051538154215, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18665 + }, + { + "epoch": 7.457451058729524, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18666 + }, + { + "epoch": 7.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18667 + }, + { + "epoch": 7.458250099880144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18668 + }, + { + "epoch": 7.4586496204554535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18669 + }, + { + "epoch": 7.459049141030763, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18670 + }, + { + "epoch": 7.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18671 + }, + { + "epoch": 7.459848182181382, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18672 + }, + { + "epoch": 7.460247702756692, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18673 + }, + { + "epoch": 7.460647223332002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18674 + }, + { + "epoch": 7.461046743907311, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 18675 + }, + { + "epoch": 7.461446264482621, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18676 + }, + { + "epoch": 7.46184578505793, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18677 + }, + { + "epoch": 7.46224530563324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18678 + }, + { + "epoch": 7.46264482620855, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18679 + }, + { + "epoch": 7.463044346783859, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18680 + }, + { + "epoch": 7.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18681 + }, + { + "epoch": 7.463843387934479, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18682 + }, + { + "epoch": 7.464242908509788, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18683 + }, + { + "epoch": 7.464642429085098, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18684 + }, + { + "epoch": 7.4650419496604075, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18685 + }, + { + "epoch": 7.465441470235717, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18686 + }, + { + "epoch": 7.465840990811027, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18687 + }, + { + "epoch": 7.466240511386337, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18688 + }, + { + "epoch": 7.466640031961646, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18689 + }, + { + "epoch": 7.4670395525369555, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18690 + }, + { + "epoch": 7.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18691 + }, + { + "epoch": 7.467838593687575, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18692 + }, + { + "epoch": 7.468238114262885, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18693 + }, + { + "epoch": 7.468637634838194, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18694 + }, + { + "epoch": 7.4690371554135035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18695 + }, + { + "epoch": 7.469436675988813, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18696 + }, + { + "epoch": 7.469836196564123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18697 + }, + { + "epoch": 7.470235717139433, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18698 + }, + { + "epoch": 7.470635237714743, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18699 + }, + { + "epoch": 7.471034758290052, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18700 + }, + { + "epoch": 7.471434278865361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 18701 + }, + { + "epoch": 7.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18702 + }, + { + "epoch": 7.472233320015981, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.2967, + "step": 18703 + }, + { + "epoch": 7.472632840591291, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18704 + }, + { + "epoch": 7.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18705 + }, + { + "epoch": 7.473431881741909, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18706 + }, + { + "epoch": 7.473831402317219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18707 + }, + { + "epoch": 7.474230922892529, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18708 + }, + { + "epoch": 7.474630443467839, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 18709 + }, + { + "epoch": 7.475029964043149, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 18710 + }, + { + "epoch": 7.4754294846184575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18711 + }, + { + "epoch": 7.475829005193767, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18712 + }, + { + "epoch": 7.476228525769077, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18713 + }, + { + "epoch": 7.476628046344387, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18714 + }, + { + "epoch": 7.477027566919697, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18715 + }, + { + "epoch": 7.477427087495006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 18716 + }, + { + "epoch": 7.477826608070315, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18717 + }, + { + "epoch": 7.478226128645625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18718 + }, + { + "epoch": 7.478625649220935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 18719 + }, + { + "epoch": 7.479025169796245, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18720 + }, + { + "epoch": 7.479424690371554, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18721 + }, + { + "epoch": 7.479824210946864, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18722 + }, + { + "epoch": 7.480223731522173, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18723 + }, + { + "epoch": 7.480623252097483, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18724 + }, + { + "epoch": 7.481022772672793, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18725 + }, + { + "epoch": 7.4814222932481025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18726 + }, + { + "epoch": 7.481821813823412, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18727 + }, + { + "epoch": 7.482221334398721, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18728 + }, + { + "epoch": 7.482620854974031, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18729 + }, + { + "epoch": 7.483020375549341, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18730 + }, + { + "epoch": 7.4834198961246505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18731 + }, + { + "epoch": 7.48381941669996, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18732 + }, + { + "epoch": 7.484218937275269, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18733 + }, + { + "epoch": 7.484618457850579, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18734 + }, + { + "epoch": 7.485017978425889, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18735 + }, + { + "epoch": 7.485417499001199, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18736 + }, + { + "epoch": 7.485817019576508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18737 + }, + { + "epoch": 7.486216540151818, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18738 + }, + { + "epoch": 7.486616060727127, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18739 + }, + { + "epoch": 7.487015581302437, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18740 + }, + { + "epoch": 7.487415101877747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18741 + }, + { + "epoch": 7.487814622453056, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18742 + }, + { + "epoch": 7.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18743 + }, + { + "epoch": 7.488613663603676, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18744 + }, + { + "epoch": 7.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18745 + }, + { + "epoch": 7.489412704754295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18746 + }, + { + "epoch": 7.4898122253296044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18747 + }, + { + "epoch": 7.490211745904914, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18748 + }, + { + "epoch": 7.490611266480224, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18749 + }, + { + "epoch": 7.491010787055533, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18750 + }, + { + "epoch": 7.491410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18751 + }, + { + "epoch": 7.4918098282061525, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18752 + }, + { + "epoch": 7.492209348781462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18753 + }, + { + "epoch": 7.492608869356772, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18754 + }, + { + "epoch": 7.493008389932082, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18755 + }, + { + "epoch": 7.493407910507391, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18756 + }, + { + "epoch": 7.4938074310827005, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18757 + }, + { + "epoch": 7.49420695165801, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18758 + }, + { + "epoch": 7.49460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18759 + }, + { + "epoch": 7.49500599280863, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18760 + }, + { + "epoch": 7.49540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18761 + }, + { + "epoch": 7.495805033959249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18762 + }, + { + "epoch": 7.496204554534558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18763 + }, + { + "epoch": 7.496604075109868, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18764 + }, + { + "epoch": 7.497003595685178, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18765 + }, + { + "epoch": 7.497403116260488, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18766 + }, + { + "epoch": 7.497802636835797, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18767 + }, + { + "epoch": 7.498202157411106, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18768 + }, + { + "epoch": 7.498601677986416, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18769 + }, + { + "epoch": 7.499001198561726, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18770 + }, + { + "epoch": 7.499400719137036, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18771 + }, + { + "epoch": 7.4998002397123456, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18772 + }, + { + "epoch": 7.5001997602876544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18773 + }, + { + "epoch": 7.500599280862964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18774 + }, + { + "epoch": 7.500998801438274, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18775 + }, + { + "epoch": 7.501398322013584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18776 + }, + { + "epoch": 7.501797842588894, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18777 + }, + { + "epoch": 7.502197363164203, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18778 + }, + { + "epoch": 7.502596883739512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18779 + }, + { + "epoch": 7.502996404314822, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18780 + }, + { + "epoch": 7.503395924890132, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18781 + }, + { + "epoch": 7.503795445465442, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18782 + }, + { + "epoch": 7.504194966040751, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18783 + }, + { + "epoch": 7.50459448661606, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18784 + }, + { + "epoch": 7.50499400719137, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18785 + }, + { + "epoch": 7.50539352776668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18786 + }, + { + "epoch": 7.50579304834199, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18787 + }, + { + "epoch": 7.5061925689172995, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18788 + }, + { + "epoch": 7.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18789 + }, + { + "epoch": 7.506991610067918, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18790 + }, + { + "epoch": 7.507391130643228, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18791 + }, + { + "epoch": 7.507790651218538, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18792 + }, + { + "epoch": 7.5081901717938475, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18793 + }, + { + "epoch": 7.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18794 + }, + { + "epoch": 7.508989212944467, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18795 + }, + { + "epoch": 7.509388733519776, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18796 + }, + { + "epoch": 7.509788254095086, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18797 + }, + { + "epoch": 7.5101877746703956, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18798 + }, + { + "epoch": 7.510587295245705, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18799 + }, + { + "epoch": 7.510986815821015, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18800 + }, + { + "epoch": 7.511386336396324, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18801 + }, + { + "epoch": 7.511785856971634, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18802 + }, + { + "epoch": 7.512185377546944, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18803 + }, + { + "epoch": 7.512584898122253, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18804 + }, + { + "epoch": 7.512984418697563, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 18805 + }, + { + "epoch": 7.513383939272872, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18806 + }, + { + "epoch": 7.513783459848182, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18807 + }, + { + "epoch": 7.514182980423492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18808 + }, + { + "epoch": 7.514582500998801, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18809 + }, + { + "epoch": 7.514982021574111, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18810 + }, + { + "epoch": 7.515381542149421, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18811 + }, + { + "epoch": 7.515781062724731, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18812 + }, + { + "epoch": 7.51618058330004, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18813 + }, + { + "epoch": 7.5165801038753495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18814 + }, + { + "epoch": 7.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18815 + }, + { + "epoch": 7.517379145025969, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18816 + }, + { + "epoch": 7.517778665601279, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18817 + }, + { + "epoch": 7.518178186176588, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18818 + }, + { + "epoch": 7.5185777067518975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18819 + }, + { + "epoch": 7.518977227327207, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18820 + }, + { + "epoch": 7.519376747902517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18821 + }, + { + "epoch": 7.519776268477827, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18822 + }, + { + "epoch": 7.520175789053136, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18823 + }, + { + "epoch": 7.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18824 + }, + { + "epoch": 7.520974830203755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18825 + }, + { + "epoch": 7.521374350779065, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18826 + }, + { + "epoch": 7.521773871354375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18827 + }, + { + "epoch": 7.522173391929685, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18828 + }, + { + "epoch": 7.5225729125049945, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18829 + }, + { + "epoch": 7.522972433080303, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18830 + }, + { + "epoch": 7.523371953655613, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18831 + }, + { + "epoch": 7.523771474230923, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18832 + }, + { + "epoch": 7.524170994806233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18833 + }, + { + "epoch": 7.5245705153815425, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18834 + }, + { + "epoch": 7.524970035956851, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18835 + }, + { + "epoch": 7.525369556532161, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18836 + }, + { + "epoch": 7.525769077107471, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18837 + }, + { + "epoch": 7.526168597682781, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18838 + }, + { + "epoch": 7.526568118258091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18839 + }, + { + "epoch": 7.5269676388333995, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18840 + }, + { + "epoch": 7.527367159408709, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18841 + }, + { + "epoch": 7.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18842 + }, + { + "epoch": 7.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18843 + }, + { + "epoch": 7.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18844 + }, + { + "epoch": 7.528965241709948, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18845 + }, + { + "epoch": 7.529364762285257, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18846 + }, + { + "epoch": 7.529764282860567, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 18847 + }, + { + "epoch": 7.530163803435877, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18848 + }, + { + "epoch": 7.530563324011187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18849 + }, + { + "epoch": 7.5309628445864965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18850 + }, + { + "epoch": 7.531362365161806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18851 + }, + { + "epoch": 7.531761885737115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18852 + }, + { + "epoch": 7.532161406312425, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18853 + }, + { + "epoch": 7.532560926887735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18854 + }, + { + "epoch": 7.5329604474630445, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18855 + }, + { + "epoch": 7.533359968038354, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18856 + }, + { + "epoch": 7.533759488613663, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18857 + }, + { + "epoch": 7.534159009188973, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18858 + }, + { + "epoch": 7.534558529764283, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18859 + }, + { + "epoch": 7.5349580503395925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18860 + }, + { + "epoch": 7.535357570914902, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18861 + }, + { + "epoch": 7.535757091490212, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18862 + }, + { + "epoch": 7.536156612065521, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18863 + }, + { + "epoch": 7.536556132640831, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18864 + }, + { + "epoch": 7.536955653216141, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18865 + }, + { + "epoch": 7.53735517379145, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18866 + }, + { + "epoch": 7.53775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18867 + }, + { + "epoch": 7.53815421494207, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18868 + }, + { + "epoch": 7.538553735517379, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18869 + }, + { + "epoch": 7.538953256092689, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18870 + }, + { + "epoch": 7.539352776667998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18871 + }, + { + "epoch": 7.539752297243308, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 18872 + }, + { + "epoch": 7.540151817818618, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18873 + }, + { + "epoch": 7.540551338393927, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18874 + }, + { + "epoch": 7.540950858969237, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18875 + }, + { + "epoch": 7.5413503795445465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18876 + }, + { + "epoch": 7.541749900119856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18877 + }, + { + "epoch": 7.542149420695166, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18878 + }, + { + "epoch": 7.542548941270476, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18879 + }, + { + "epoch": 7.542948461845785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18880 + }, + { + "epoch": 7.5433479824210945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18881 + }, + { + "epoch": 7.543747502996404, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18882 + }, + { + "epoch": 7.544147023571714, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18883 + }, + { + "epoch": 7.544546544147024, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18884 + }, + { + "epoch": 7.544946064722334, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18885 + }, + { + "epoch": 7.5453455852976425, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18886 + }, + { + "epoch": 7.545745105872952, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18887 + }, + { + "epoch": 7.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18888 + }, + { + "epoch": 7.546544147023572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18889 + }, + { + "epoch": 7.546943667598882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18890 + }, + { + "epoch": 7.547343188174191, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18891 + }, + { + "epoch": 7.5477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18892 + }, + { + "epoch": 7.54814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18893 + }, + { + "epoch": 7.54854174990012, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18894 + }, + { + "epoch": 7.54894127047543, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18895 + }, + { + "epoch": 7.549340791050739, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18896 + }, + { + "epoch": 7.549740311626048, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18897 + }, + { + "epoch": 7.550139832201358, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18898 + }, + { + "epoch": 7.550539352776668, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18899 + }, + { + "epoch": 7.550938873351978, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18900 + }, + { + "epoch": 7.551338393927288, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18901 + }, + { + "epoch": 7.551737914502597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18902 + }, + { + "epoch": 7.552137435077906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18903 + }, + { + "epoch": 7.552536955653216, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18904 + }, + { + "epoch": 7.552936476228526, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18905 + }, + { + "epoch": 7.553335996803836, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18906 + }, + { + "epoch": 7.553735517379145, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18907 + }, + { + "epoch": 7.554135037954454, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18908 + }, + { + "epoch": 7.554534558529764, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18909 + }, + { + "epoch": 7.554934079105074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18910 + }, + { + "epoch": 7.555333599680384, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18911 + }, + { + "epoch": 7.555733120255693, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18912 + }, + { + "epoch": 7.556132640831002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18913 + }, + { + "epoch": 7.556532161406312, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18914 + }, + { + "epoch": 7.556931681981622, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18915 + }, + { + "epoch": 7.557331202556932, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18916 + }, + { + "epoch": 7.5577307231322415, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18917 + }, + { + "epoch": 7.558130243707551, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18918 + }, + { + "epoch": 7.558529764282861, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18919 + }, + { + "epoch": 7.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18920 + }, + { + "epoch": 7.55932880543348, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18921 + }, + { + "epoch": 7.5597283260087895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18922 + }, + { + "epoch": 7.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18923 + }, + { + "epoch": 7.560527367159409, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18924 + }, + { + "epoch": 7.560926887734718, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18925 + }, + { + "epoch": 7.561326408310028, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18926 + }, + { + "epoch": 7.561725928885338, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18927 + }, + { + "epoch": 7.562125449460647, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18928 + }, + { + "epoch": 7.562524970035957, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18929 + }, + { + "epoch": 7.562924490611266, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 18930 + }, + { + "epoch": 7.563324011186576, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18931 + }, + { + "epoch": 7.563723531761886, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18932 + }, + { + "epoch": 7.564123052337195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18933 + }, + { + "epoch": 7.564522572912505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18934 + }, + { + "epoch": 7.564922093487815, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18935 + }, + { + "epoch": 7.565321614063124, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18936 + }, + { + "epoch": 7.565721134638434, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18937 + }, + { + "epoch": 7.566120655213743, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18938 + }, + { + "epoch": 7.566520175789053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18939 + }, + { + "epoch": 7.566919696364363, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18940 + }, + { + "epoch": 7.567319216939673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18941 + }, + { + "epoch": 7.567718737514982, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 18942 + }, + { + "epoch": 7.5681182580902915, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18943 + }, + { + "epoch": 7.568517778665601, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18944 + }, + { + "epoch": 7.568917299240911, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18945 + }, + { + "epoch": 7.569316819816221, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18946 + }, + { + "epoch": 7.56971634039153, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18947 + }, + { + "epoch": 7.5701158609668395, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18948 + }, + { + "epoch": 7.570515381542149, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18949 + }, + { + "epoch": 7.570914902117459, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 18950 + }, + { + "epoch": 7.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18951 + }, + { + "epoch": 7.571713943268079, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18952 + }, + { + "epoch": 7.572113463843388, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18953 + }, + { + "epoch": 7.572512984418697, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18954 + }, + { + "epoch": 7.572912504994007, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18955 + }, + { + "epoch": 7.573312025569317, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18956 + }, + { + "epoch": 7.573711546144627, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18957 + }, + { + "epoch": 7.5741110667199365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18958 + }, + { + "epoch": 7.574510587295245, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18959 + }, + { + "epoch": 7.574910107870555, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 18960 + }, + { + "epoch": 7.575309628445865, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18961 + }, + { + "epoch": 7.575709149021175, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18962 + }, + { + "epoch": 7.5761086695964845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18963 + }, + { + "epoch": 7.5765081901717934, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18964 + }, + { + "epoch": 7.576907710747103, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18965 + }, + { + "epoch": 7.577307231322413, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18966 + }, + { + "epoch": 7.577706751897723, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 18967 + }, + { + "epoch": 7.578106272473033, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18968 + }, + { + "epoch": 7.578505793048342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18969 + }, + { + "epoch": 7.578905313623651, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18970 + }, + { + "epoch": 7.579304834198961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18971 + }, + { + "epoch": 7.579704354774271, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 18972 + }, + { + "epoch": 7.580103875349581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18973 + }, + { + "epoch": 7.58050339592489, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18974 + }, + { + "epoch": 7.5809029165002, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18975 + }, + { + "epoch": 7.581302437075509, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18976 + }, + { + "epoch": 7.581701957650819, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18977 + }, + { + "epoch": 7.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18978 + }, + { + "epoch": 7.5825009988014385, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18979 + }, + { + "epoch": 7.582900519376748, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18980 + }, + { + "epoch": 7.583300039952057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 18981 + }, + { + "epoch": 7.583699560527367, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18982 + }, + { + "epoch": 7.584099081102677, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18983 + }, + { + "epoch": 7.5844986016779865, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18984 + }, + { + "epoch": 7.584898122253296, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18985 + }, + { + "epoch": 7.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18986 + }, + { + "epoch": 7.585697163403915, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18987 + }, + { + "epoch": 7.586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18988 + }, + { + "epoch": 7.5864962045545346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18989 + }, + { + "epoch": 7.586895725129844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18990 + }, + { + "epoch": 7.587295245705154, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18991 + }, + { + "epoch": 7.587694766280464, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18992 + }, + { + "epoch": 7.588094286855773, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18993 + }, + { + "epoch": 7.588493807431083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18994 + }, + { + "epoch": 7.588893328006392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18995 + }, + { + "epoch": 7.589292848581702, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18996 + }, + { + "epoch": 7.589692369157012, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18997 + }, + { + "epoch": 7.590091889732321, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18998 + }, + { + "epoch": 7.590491410307631, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18999 + }, + { + "epoch": 7.59089093088294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19000 + }, + { + "epoch": 7.59129045145825, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19001 + }, + { + "epoch": 7.59168997203356, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19002 + }, + { + "epoch": 7.592089492608869, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 19003 + }, + { + "epoch": 7.592489013184179, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19004 + }, + { + "epoch": 7.5928885337594885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19005 + }, + { + "epoch": 7.593288054334798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19006 + }, + { + "epoch": 7.593687574910108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19007 + }, + { + "epoch": 7.594087095485418, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19008 + }, + { + "epoch": 7.594486616060728, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19009 + }, + { + "epoch": 7.5948861366360365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19010 + }, + { + "epoch": 7.595285657211346, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19011 + }, + { + "epoch": 7.595685177786656, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19012 + }, + { + "epoch": 7.596084698361966, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19013 + }, + { + "epoch": 7.596484218937276, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 19014 + }, + { + "epoch": 7.5968837395125846, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19015 + }, + { + "epoch": 7.597283260087894, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19016 + }, + { + "epoch": 7.597682780663204, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19017 + }, + { + "epoch": 7.598082301238514, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19018 + }, + { + "epoch": 7.598481821813824, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19019 + }, + { + "epoch": 7.598881342389133, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19020 + }, + { + "epoch": 7.599280862964442, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19021 + }, + { + "epoch": 7.599680383539752, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19022 + }, + { + "epoch": 7.600079904115062, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19023 + }, + { + "epoch": 7.600479424690372, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19024 + }, + { + "epoch": 7.6008789452656815, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19025 + }, + { + "epoch": 7.60127846584099, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19026 + }, + { + "epoch": 7.6016779864163, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19027 + }, + { + "epoch": 7.60207750699161, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19028 + }, + { + "epoch": 7.60247702756692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19029 + }, + { + "epoch": 7.60287654814223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19030 + }, + { + "epoch": 7.603276068717539, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19031 + }, + { + "epoch": 7.603675589292848, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19032 + }, + { + "epoch": 7.604075109868158, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19033 + }, + { + "epoch": 7.604474630443468, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19034 + }, + { + "epoch": 7.604874151018778, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19035 + }, + { + "epoch": 7.605273671594087, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19036 + }, + { + "epoch": 7.605673192169396, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19037 + }, + { + "epoch": 7.606072712744706, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19038 + }, + { + "epoch": 7.606472233320016, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19039 + }, + { + "epoch": 7.606871753895326, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19040 + }, + { + "epoch": 7.6072712744706354, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19041 + }, + { + "epoch": 7.607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19042 + }, + { + "epoch": 7.608070315621254, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19043 + }, + { + "epoch": 7.608469836196564, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19044 + }, + { + "epoch": 7.608869356771874, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19045 + }, + { + "epoch": 7.6092688773471835, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19046 + }, + { + "epoch": 7.609668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19047 + }, + { + "epoch": 7.610067918497803, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19048 + }, + { + "epoch": 7.610467439073112, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19049 + }, + { + "epoch": 7.610866959648422, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19050 + }, + { + "epoch": 7.6112664802237315, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19051 + }, + { + "epoch": 7.611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 19052 + }, + { + "epoch": 7.612065521374351, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19053 + }, + { + "epoch": 7.61246504194966, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19054 + }, + { + "epoch": 7.61286456252497, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19055 + }, + { + "epoch": 7.61326408310028, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19056 + }, + { + "epoch": 7.613663603675589, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19057 + }, + { + "epoch": 7.614063124250899, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19058 + }, + { + "epoch": 7.614462644826209, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19059 + }, + { + "epoch": 7.614862165401518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19060 + }, + { + "epoch": 7.615261685976828, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19061 + }, + { + "epoch": 7.615661206552137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19062 + }, + { + "epoch": 7.616060727127447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19063 + }, + { + "epoch": 7.616460247702757, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19064 + }, + { + "epoch": 7.616859768278067, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19065 + }, + { + "epoch": 7.617259288853376, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19066 + }, + { + "epoch": 7.6176588094286855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19067 + }, + { + "epoch": 7.618058330003995, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19068 + }, + { + "epoch": 7.618457850579305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19069 + }, + { + "epoch": 7.618857371154615, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19070 + }, + { + "epoch": 7.619256891729924, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19071 + }, + { + "epoch": 7.6196564123052335, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19072 + }, + { + "epoch": 7.620055932880543, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19073 + }, + { + "epoch": 7.620455453455853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19074 + }, + { + "epoch": 7.620854974031163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19075 + }, + { + "epoch": 7.621254494606472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19076 + }, + { + "epoch": 7.6216540151817815, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19077 + }, + { + "epoch": 7.622053535757091, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19078 + }, + { + "epoch": 7.622453056332401, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19079 + }, + { + "epoch": 7.622852576907711, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19080 + }, + { + "epoch": 7.623252097483021, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19081 + }, + { + "epoch": 7.6236516180583305, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19082 + }, + { + "epoch": 7.624051138633639, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19083 + }, + { + "epoch": 7.624450659208949, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19084 + }, + { + "epoch": 7.624850179784259, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19085 + }, + { + "epoch": 7.625249700359569, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19086 + }, + { + "epoch": 7.6256492209348785, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19087 + }, + { + "epoch": 7.626048741510187, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19088 + }, + { + "epoch": 7.626448262085497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19089 + }, + { + "epoch": 7.626847782660807, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19090 + }, + { + "epoch": 7.627247303236117, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19091 + }, + { + "epoch": 7.627646823811427, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19092 + }, + { + "epoch": 7.6280463443867355, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19093 + }, + { + "epoch": 7.628445864962045, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19094 + }, + { + "epoch": 7.628845385537355, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19095 + }, + { + "epoch": 7.629244906112665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19096 + }, + { + "epoch": 7.629644426687975, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19097 + }, + { + "epoch": 7.630043947263284, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19098 + }, + { + "epoch": 7.630443467838594, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19099 + }, + { + "epoch": 7.630842988413903, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19100 + }, + { + "epoch": 7.631242508989213, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19101 + }, + { + "epoch": 7.631642029564523, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19102 + }, + { + "epoch": 7.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19103 + }, + { + "epoch": 7.632441070715142, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19104 + }, + { + "epoch": 7.632840591290451, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19105 + }, + { + "epoch": 7.633240111865761, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19106 + }, + { + "epoch": 7.633639632441071, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19107 + }, + { + "epoch": 7.6340391530163805, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19108 + }, + { + "epoch": 7.63443867359169, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19109 + }, + { + "epoch": 7.634838194166999, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19110 + }, + { + "epoch": 7.635237714742309, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19111 + }, + { + "epoch": 7.635637235317619, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19112 + }, + { + "epoch": 7.6360367558929285, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19113 + }, + { + "epoch": 7.636436276468238, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19114 + }, + { + "epoch": 7.636835797043548, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19115 + }, + { + "epoch": 7.637235317618857, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 19116 + }, + { + "epoch": 7.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19117 + }, + { + "epoch": 7.638034358769477, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19118 + }, + { + "epoch": 7.638433879344786, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19119 + }, + { + "epoch": 7.638833399920096, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19120 + }, + { + "epoch": 7.639232920495406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19121 + }, + { + "epoch": 7.639632441070715, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19122 + }, + { + "epoch": 7.640031961646025, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19123 + }, + { + "epoch": 7.640431482221334, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19124 + }, + { + "epoch": 7.640831002796644, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19125 + }, + { + "epoch": 7.641230523371954, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19126 + }, + { + "epoch": 7.641630043947263, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 19127 + }, + { + "epoch": 7.642029564522573, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19128 + }, + { + "epoch": 7.642429085097882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19129 + }, + { + "epoch": 7.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19130 + }, + { + "epoch": 7.643228126248502, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19131 + }, + { + "epoch": 7.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19132 + }, + { + "epoch": 7.644027167399121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 19133 + }, + { + "epoch": 7.6444266879744305, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19134 + }, + { + "epoch": 7.64482620854974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19135 + }, + { + "epoch": 7.64522572912505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19136 + }, + { + "epoch": 7.64562524970036, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19137 + }, + { + "epoch": 7.64602477027567, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19138 + }, + { + "epoch": 7.6464242908509785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19139 + }, + { + "epoch": 7.646823811426288, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 19140 + }, + { + "epoch": 7.647223332001598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19141 + }, + { + "epoch": 7.647622852576908, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19142 + }, + { + "epoch": 7.648022373152218, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 19143 + }, + { + "epoch": 7.648421893727527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19144 + }, + { + "epoch": 7.648821414302836, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19145 + }, + { + "epoch": 7.649220934878146, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19146 + }, + { + "epoch": 7.649620455453456, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19147 + }, + { + "epoch": 7.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19148 + }, + { + "epoch": 7.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19149 + }, + { + "epoch": 7.650819017179384, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 19150 + }, + { + "epoch": 7.651218537754694, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19151 + }, + { + "epoch": 7.651618058330004, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19152 + }, + { + "epoch": 7.652017578905314, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19153 + }, + { + "epoch": 7.6524170994806235, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19154 + }, + { + "epoch": 7.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19155 + }, + { + "epoch": 7.653216140631242, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19156 + }, + { + "epoch": 7.653615661206552, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19157 + }, + { + "epoch": 7.654015181781862, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19158 + }, + { + "epoch": 7.654414702357172, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19159 + }, + { + "epoch": 7.654814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19160 + }, + { + "epoch": 7.65521374350779, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19161 + }, + { + "epoch": 7.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19162 + }, + { + "epoch": 7.65601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 19163 + }, + { + "epoch": 7.65641230523372, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19164 + }, + { + "epoch": 7.656811825809029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19165 + }, + { + "epoch": 7.657211346384338, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19166 + }, + { + "epoch": 7.657610866959648, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 19167 + }, + { + "epoch": 7.658010387534958, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19168 + }, + { + "epoch": 7.658409908110268, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19169 + }, + { + "epoch": 7.6588094286855775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19170 + }, + { + "epoch": 7.659208949260887, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19171 + }, + { + "epoch": 7.659608469836197, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19172 + }, + { + "epoch": 7.660007990411506, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19173 + }, + { + "epoch": 7.660407510986816, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19174 + }, + { + "epoch": 7.6608070315621255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19175 + }, + { + "epoch": 7.661206552137435, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19176 + }, + { + "epoch": 7.661606072712745, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19177 + }, + { + "epoch": 7.662005593288054, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19178 + }, + { + "epoch": 7.662405113863364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19179 + }, + { + "epoch": 7.6628046344386735, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19180 + }, + { + "epoch": 7.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19181 + }, + { + "epoch": 7.663603675589293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19182 + }, + { + "epoch": 7.664003196164602, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19183 + }, + { + "epoch": 7.664402716739912, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19184 + }, + { + "epoch": 7.664802237315222, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 19185 + }, + { + "epoch": 7.665201757890531, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19186 + }, + { + "epoch": 7.665601278465841, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 19187 + }, + { + "epoch": 7.666000799041151, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19188 + }, + { + "epoch": 7.666400319616461, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19189 + }, + { + "epoch": 7.66679984019177, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19190 + }, + { + "epoch": 7.667199360767079, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19191 + }, + { + "epoch": 7.667598881342389, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19192 + }, + { + "epoch": 7.667998401917699, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19193 + }, + { + "epoch": 7.668397922493009, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19194 + }, + { + "epoch": 7.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19195 + }, + { + "epoch": 7.6691969636436275, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19196 + }, + { + "epoch": 7.669596484218937, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19197 + }, + { + "epoch": 7.669996004794247, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19198 + }, + { + "epoch": 7.670395525369557, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19199 + }, + { + "epoch": 7.670795045944866, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19200 + }, + { + "epoch": 7.6711945665201755, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19201 + }, + { + "epoch": 7.671594087095485, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19202 + }, + { + "epoch": 7.671993607670795, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19203 + }, + { + "epoch": 7.672393128246105, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 19204 + }, + { + "epoch": 7.672792648821415, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 19205 + }, + { + "epoch": 7.6731921693967235, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19206 + }, + { + "epoch": 7.673591689972033, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19207 + }, + { + "epoch": 7.673991210547343, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19208 + }, + { + "epoch": 7.674390731122653, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19209 + }, + { + "epoch": 7.674790251697963, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19210 + }, + { + "epoch": 7.6751897722732725, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19211 + }, + { + "epoch": 7.675589292848581, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19212 + }, + { + "epoch": 7.675988813423891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19213 + }, + { + "epoch": 7.676388333999201, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19214 + }, + { + "epoch": 7.676787854574511, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19215 + }, + { + "epoch": 7.6771873751498205, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19216 + }, + { + "epoch": 7.677586895725129, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 19217 + }, + { + "epoch": 7.677986416300439, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 19218 + }, + { + "epoch": 7.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19219 + }, + { + "epoch": 7.678785457451059, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19220 + }, + { + "epoch": 7.679184978026369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19221 + }, + { + "epoch": 7.679584498601678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19222 + }, + { + "epoch": 7.679984019176987, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19223 + }, + { + "epoch": 7.680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19224 + }, + { + "epoch": 7.680783060327607, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19225 + }, + { + "epoch": 7.681182580902917, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19226 + }, + { + "epoch": 7.681582101478226, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19227 + }, + { + "epoch": 7.681981622053536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19228 + }, + { + "epoch": 7.682381142628845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19229 + }, + { + "epoch": 7.682780663204155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19230 + }, + { + "epoch": 7.683180183779465, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19231 + }, + { + "epoch": 7.683579704354774, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19232 + }, + { + "epoch": 7.683979224930084, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19233 + }, + { + "epoch": 7.684378745505393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19234 + }, + { + "epoch": 7.684778266080703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19235 + }, + { + "epoch": 7.685177786656013, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19236 + }, + { + "epoch": 7.6855773072313225, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19237 + }, + { + "epoch": 7.685976827806632, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19238 + }, + { + "epoch": 7.686376348381942, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19239 + }, + { + "epoch": 7.686775868957251, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19240 + }, + { + "epoch": 7.687175389532561, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19241 + }, + { + "epoch": 7.6875749101078705, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19242 + }, + { + "epoch": 7.68797443068318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19243 + }, + { + "epoch": 7.68837395125849, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19244 + }, + { + "epoch": 7.6887734718338, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19245 + }, + { + "epoch": 7.689172992409109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19246 + }, + { + "epoch": 7.689572512984419, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19247 + }, + { + "epoch": 7.689972033559728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 19248 + }, + { + "epoch": 7.690371554135038, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19249 + }, + { + "epoch": 7.690771074710348, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19250 + }, + { + "epoch": 7.691170595285657, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19251 + }, + { + "epoch": 7.691570115860967, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19252 + }, + { + "epoch": 7.691969636436276, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19253 + }, + { + "epoch": 7.692369157011586, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19254 + }, + { + "epoch": 7.692768677586896, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19255 + }, + { + "epoch": 7.693168198162205, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19256 + }, + { + "epoch": 7.693567718737515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19257 + }, + { + "epoch": 7.6939672393128244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19258 + }, + { + "epoch": 7.694366759888134, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19259 + }, + { + "epoch": 7.694766280463444, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19260 + }, + { + "epoch": 7.695165801038754, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19261 + }, + { + "epoch": 7.695565321614064, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19262 + }, + { + "epoch": 7.6959648421893725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19263 + }, + { + "epoch": 7.696364362764682, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 19264 + }, + { + "epoch": 7.696763883339992, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 19265 + }, + { + "epoch": 7.697163403915302, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19266 + }, + { + "epoch": 7.697562924490612, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19267 + }, + { + "epoch": 7.6979624450659205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 19268 + }, + { + "epoch": 7.69836196564123, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19269 + }, + { + "epoch": 7.69876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19270 + }, + { + "epoch": 7.69916100679185, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19271 + }, + { + "epoch": 7.69956052736716, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19272 + }, + { + "epoch": 7.699960047942469, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19273 + }, + { + "epoch": 7.700359568517778, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19274 + }, + { + "epoch": 7.700759089093088, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19275 + }, + { + "epoch": 7.701158609668398, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 19276 + }, + { + "epoch": 7.701558130243708, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19277 + }, + { + "epoch": 7.7019576508190175, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19278 + }, + { + "epoch": 7.702357171394327, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19279 + }, + { + "epoch": 7.702756691969636, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19280 + }, + { + "epoch": 7.703156212544946, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19281 + }, + { + "epoch": 7.703555733120256, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19282 + }, + { + "epoch": 7.7039552536955656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19283 + }, + { + "epoch": 7.704354774270875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19284 + }, + { + "epoch": 7.704754294846184, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19285 + }, + { + "epoch": 7.705153815421494, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19286 + }, + { + "epoch": 7.705553335996804, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19287 + }, + { + "epoch": 7.705952856572114, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19288 + }, + { + "epoch": 7.706352377147423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19289 + }, + { + "epoch": 7.706751897722732, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19290 + }, + { + "epoch": 7.707151418298042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19291 + }, + { + "epoch": 7.707550938873352, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19292 + }, + { + "epoch": 7.707950459448662, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19293 + }, + { + "epoch": 7.708349980023971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19294 + }, + { + "epoch": 7.708749500599281, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19295 + }, + { + "epoch": 7.70914902117459, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19296 + }, + { + "epoch": 7.7095485417499, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19297 + }, + { + "epoch": 7.70994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19298 + }, + { + "epoch": 7.7103475829005195, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19299 + }, + { + "epoch": 7.710747103475829, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19300 + }, + { + "epoch": 7.711146624051139, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19301 + }, + { + "epoch": 7.711546144626448, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19302 + }, + { + "epoch": 7.711945665201758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19303 + }, + { + "epoch": 7.7123451857770675, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19304 + }, + { + "epoch": 7.712744706352377, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19305 + }, + { + "epoch": 7.713144226927687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19306 + }, + { + "epoch": 7.713543747502996, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19307 + }, + { + "epoch": 7.713943268078306, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19308 + }, + { + "epoch": 7.714342788653616, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19309 + }, + { + "epoch": 7.714742309228925, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19310 + }, + { + "epoch": 7.715141829804235, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19311 + }, + { + "epoch": 7.715541350379545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19312 + }, + { + "epoch": 7.715940870954854, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19313 + }, + { + "epoch": 7.716340391530164, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19314 + }, + { + "epoch": 7.716739912105473, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 19315 + }, + { + "epoch": 7.717139432680783, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19316 + }, + { + "epoch": 7.717538953256093, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19317 + }, + { + "epoch": 7.717938473831403, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19318 + }, + { + "epoch": 7.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19319 + }, + { + "epoch": 7.718737514982021, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19320 + }, + { + "epoch": 7.719137035557331, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19321 + }, + { + "epoch": 7.719536556132641, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19322 + }, + { + "epoch": 7.719936076707951, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19323 + }, + { + "epoch": 7.72033559728326, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19324 + }, + { + "epoch": 7.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19325 + }, + { + "epoch": 7.721134638433879, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19326 + }, + { + "epoch": 7.721534159009189, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19327 + }, + { + "epoch": 7.721933679584499, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19328 + }, + { + "epoch": 7.722333200159808, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19329 + }, + { + "epoch": 7.7227327207351175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19330 + }, + { + "epoch": 7.723132241310427, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19331 + }, + { + "epoch": 7.723531761885737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 19332 + }, + { + "epoch": 7.723931282461047, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19333 + }, + { + "epoch": 7.724330803036357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19334 + }, + { + "epoch": 7.7247303236116664, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19335 + }, + { + "epoch": 7.725129844186975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19336 + }, + { + "epoch": 7.725529364762285, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19337 + }, + { + "epoch": 7.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19338 + }, + { + "epoch": 7.726328405912905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19339 + }, + { + "epoch": 7.7267279264882145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19340 + }, + { + "epoch": 7.727127447063523, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19341 + }, + { + "epoch": 7.727526967638833, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19342 + }, + { + "epoch": 7.727926488214143, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19343 + }, + { + "epoch": 7.728326008789453, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19344 + }, + { + "epoch": 7.7287255293647625, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 19345 + }, + { + "epoch": 7.729125049940071, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19346 + }, + { + "epoch": 7.729524570515381, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19347 + }, + { + "epoch": 7.729924091090691, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19348 + }, + { + "epoch": 7.730323611666001, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19349 + }, + { + "epoch": 7.730723132241311, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19350 + }, + { + "epoch": 7.73112265281662, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19351 + }, + { + "epoch": 7.73152217339193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19352 + }, + { + "epoch": 7.731921693967239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 19353 + }, + { + "epoch": 7.732321214542549, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19354 + }, + { + "epoch": 7.732720735117859, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 19355 + }, + { + "epoch": 7.733120255693168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19356 + }, + { + "epoch": 7.733519776268478, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19357 + }, + { + "epoch": 7.733919296843787, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 19358 + }, + { + "epoch": 7.734318817419097, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19359 + }, + { + "epoch": 7.734718337994407, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19360 + }, + { + "epoch": 7.7351178585697165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19361 + }, + { + "epoch": 7.735517379145026, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19362 + }, + { + "epoch": 7.735916899720335, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19363 + }, + { + "epoch": 7.736316420295645, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19364 + }, + { + "epoch": 7.736715940870955, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19365 + }, + { + "epoch": 7.7371154614462645, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 19366 + }, + { + "epoch": 7.737514982021574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19367 + }, + { + "epoch": 7.737914502596884, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19368 + }, + { + "epoch": 7.738314023172194, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19369 + }, + { + "epoch": 7.738713543747503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19370 + }, + { + "epoch": 7.7391130643228125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19371 + }, + { + "epoch": 7.739512584898122, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19372 + }, + { + "epoch": 7.739912105473432, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19373 + }, + { + "epoch": 7.740311626048742, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19374 + }, + { + "epoch": 7.740711146624051, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19375 + }, + { + "epoch": 7.741110667199361, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19376 + }, + { + "epoch": 7.74151018777467, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19377 + }, + { + "epoch": 7.74190970834998, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19378 + }, + { + "epoch": 7.74230922892529, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19379 + }, + { + "epoch": 7.742708749500599, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19380 + }, + { + "epoch": 7.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19381 + }, + { + "epoch": 7.743507790651218, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19382 + }, + { + "epoch": 7.743907311226528, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19383 + }, + { + "epoch": 7.744306831801838, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19384 + }, + { + "epoch": 7.744706352377148, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19385 + }, + { + "epoch": 7.745105872952457, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19386 + }, + { + "epoch": 7.7455053935277665, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19387 + }, + { + "epoch": 7.745904914103076, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19388 + }, + { + "epoch": 7.746304434678386, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 19389 + }, + { + "epoch": 7.746703955253696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 19390 + }, + { + "epoch": 7.747103475829006, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19391 + }, + { + "epoch": 7.7475029964043145, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19392 + }, + { + "epoch": 7.747902516979624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19393 + }, + { + "epoch": 7.748302037554934, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19394 + }, + { + "epoch": 7.748701558130244, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19395 + }, + { + "epoch": 7.749101078705554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19396 + }, + { + "epoch": 7.7495005992808625, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19397 + }, + { + "epoch": 7.749900119856172, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19398 + }, + { + "epoch": 7.750299640431482, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19399 + }, + { + "epoch": 7.750699161006792, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19400 + }, + { + "epoch": 7.751098681582102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 19401 + }, + { + "epoch": 7.7514982021574115, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 19402 + }, + { + "epoch": 7.75189772273272, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19403 + }, + { + "epoch": 7.75229724330803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19404 + }, + { + "epoch": 7.75269676388334, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19405 + }, + { + "epoch": 7.75309628445865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19406 + }, + { + "epoch": 7.7534958050339595, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19407 + }, + { + "epoch": 7.753895325609269, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 19408 + }, + { + "epoch": 7.754294846184578, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19409 + }, + { + "epoch": 7.754694366759888, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19410 + }, + { + "epoch": 7.755093887335198, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19411 + }, + { + "epoch": 7.755493407910508, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 19412 + }, + { + "epoch": 7.755892928485817, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19413 + }, + { + "epoch": 7.756292449061126, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19414 + }, + { + "epoch": 7.756691969636436, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19415 + }, + { + "epoch": 7.757091490211746, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19416 + }, + { + "epoch": 7.757491010787056, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19417 + }, + { + "epoch": 7.757890531362365, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19418 + }, + { + "epoch": 7.758290051937674, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19419 + }, + { + "epoch": 7.758689572512984, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 19420 + }, + { + "epoch": 7.759089093088294, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 19421 + }, + { + "epoch": 7.759488613663604, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19422 + }, + { + "epoch": 7.759888134238913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19423 + }, + { + "epoch": 7.760287654814223, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19424 + }, + { + "epoch": 7.760687175389533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 19425 + }, + { + "epoch": 7.761086695964842, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19426 + }, + { + "epoch": 7.761486216540152, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 19427 + }, + { + "epoch": 7.7618857371154615, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19428 + }, + { + "epoch": 7.762285257690771, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19429 + }, + { + "epoch": 7.762684778266081, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19430 + }, + { + "epoch": 7.76308429884139, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19431 + }, + { + "epoch": 7.7634838194167, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19432 + }, + { + "epoch": 7.7638833399920095, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19433 + }, + { + "epoch": 7.764282860567319, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19434 + }, + { + "epoch": 7.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19435 + }, + { + "epoch": 7.765081901717938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19436 + }, + { + "epoch": 7.765481422293248, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 19437 + }, + { + "epoch": 7.765880942868558, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19438 + }, + { + "epoch": 7.766280463443867, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19439 + }, + { + "epoch": 7.766679984019177, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19440 + }, + { + "epoch": 7.767079504594487, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19441 + }, + { + "epoch": 7.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19442 + }, + { + "epoch": 7.767878545745106, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19443 + }, + { + "epoch": 7.768278066320415, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19444 + }, + { + "epoch": 7.768677586895725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19445 + }, + { + "epoch": 7.769077107471035, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 19446 + }, + { + "epoch": 7.769476628046345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19447 + }, + { + "epoch": 7.769876148621654, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19448 + }, + { + "epoch": 7.770275669196963, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19449 + }, + { + "epoch": 7.770675189772273, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 19450 + }, + { + "epoch": 7.771074710347583, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19451 + }, + { + "epoch": 7.771474230922893, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19452 + }, + { + "epoch": 7.771873751498202, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19453 + }, + { + "epoch": 7.7722732720735115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19454 + }, + { + "epoch": 7.772672792648821, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19455 + }, + { + "epoch": 7.773072313224131, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19456 + }, + { + "epoch": 7.773471833799441, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19457 + }, + { + "epoch": 7.773871354374751, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19458 + }, + { + "epoch": 7.77427087495006, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19459 + }, + { + "epoch": 7.774670395525369, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19460 + }, + { + "epoch": 7.775069916100679, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19461 + }, + { + "epoch": 7.775469436675989, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19462 + }, + { + "epoch": 7.775868957251299, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19463 + }, + { + "epoch": 7.7762684778266085, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19464 + }, + { + "epoch": 7.776667998401917, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19465 + }, + { + "epoch": 7.777067518977227, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 19466 + }, + { + "epoch": 7.777467039552537, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19467 + }, + { + "epoch": 7.777866560127847, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19468 + }, + { + "epoch": 7.7782660807031565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19469 + }, + { + "epoch": 7.778665601278465, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19470 + }, + { + "epoch": 7.779065121853775, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 19471 + }, + { + "epoch": 7.779464642429085, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 19472 + }, + { + "epoch": 7.779864163004395, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19473 + }, + { + "epoch": 7.7802636835797045, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19474 + }, + { + "epoch": 7.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19475 + }, + { + "epoch": 7.781062724730323, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19476 + }, + { + "epoch": 7.781462245305633, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19477 + }, + { + "epoch": 7.781861765880943, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19478 + }, + { + "epoch": 7.782261286456253, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19479 + }, + { + "epoch": 7.782660807031562, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19480 + }, + { + "epoch": 7.783060327606872, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19481 + }, + { + "epoch": 7.783459848182181, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19482 + }, + { + "epoch": 7.783859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19483 + }, + { + "epoch": 7.784258889332801, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19484 + }, + { + "epoch": 7.78465840990811, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19485 + }, + { + "epoch": 7.78505793048342, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19486 + }, + { + "epoch": 7.785457451058729, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19487 + }, + { + "epoch": 7.785856971634039, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 19488 + }, + { + "epoch": 7.786256492209349, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19489 + }, + { + "epoch": 7.7866560127846585, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19490 + }, + { + "epoch": 7.787055533359968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19491 + }, + { + "epoch": 7.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19492 + }, + { + "epoch": 7.787854574510587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 19493 + }, + { + "epoch": 7.788254095085897, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19494 + }, + { + "epoch": 7.7886536156612065, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19495 + }, + { + "epoch": 7.789053136236516, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19496 + }, + { + "epoch": 7.789452656811826, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19497 + }, + { + "epoch": 7.789852177387136, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19498 + }, + { + "epoch": 7.790251697962445, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19499 + }, + { + "epoch": 7.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19500 + } + ], + "logging_steps": 1.0, + "max_steps": 250300, + "num_input_tokens_seen": 0, + "num_train_epochs": 100, + "save_steps": 250, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.919856447051006e+18, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-19500/training_args.bin b/checkpoint-19500/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2bc6e8c6a7b45b79e23a7e4b282dcb7f0e746bc4 --- /dev/null +++ b/checkpoint-19500/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9307aa53030dab795f2b0f5373487f7b4270f37adff792b3db36f71bcb8dae52 +size 5816 diff --git a/checkpoint-19750/config.json b/checkpoint-19750/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fe7dbf7d28e62328ecc2df097527bd51ec161f9d --- /dev/null +++ b/checkpoint-19750/config.json @@ -0,0 +1,73 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 6, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 512, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 64, + "intermediate_size": 8960, + "layerwise_start_idx": 21, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 7, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f128c512-mix-1.0-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-19750/generation_config.json b/checkpoint-19750/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-19750/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-19750/model-00001-of-00002.safetensors b/checkpoint-19750/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6293d4e4fd5d0e9b6a2676780a31cc2fab74ffae --- /dev/null +++ b/checkpoint-19750/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c24b2ec53694c8db7b0da1c7e723c0089b1943bf5be240ecbfe6a61152c4dee3 +size 4754815448 diff --git a/checkpoint-19750/model-00002-of-00002.safetensors b/checkpoint-19750/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c76cf00e7a9c7ce6c6d8a7a347afcbf18b661e2d --- /dev/null +++ b/checkpoint-19750/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d096d5260908edb86a73ee04889ac23d52b4054c4adabbc4a0e6ec76854d6f +size 840447856 diff --git a/checkpoint-19750/model.safetensors.index.json b/checkpoint-19750/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea3b7f98260439a7c863a0290604b077271f337 --- /dev/null +++ b/checkpoint-19750/model.safetensors.index.json @@ -0,0 +1,1076 @@ +{ + "metadata": { + "total_size": 5595140920 + }, + "weight_map": { + "embed_tokens.weight": "model-00001-of-00002.safetensors", + "img2text.bias": "model-00001-of-00002.safetensors", + "img2text.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.weight": "model-00001-of-00002.safetensors", + "layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.gate": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.22.gate": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.23.gate": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.24.gate": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.25.gate": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.26.gate": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.27.gate": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "norm.weight": "model-00001-of-00002.safetensors", + "patch_embedder.proj.bias": "model-00001-of-00002.safetensors", + "patch_embedder.proj.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.bias": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_out.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.conv_out.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.encoder.conv_in.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_in.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-19750/optimizer.pt b/checkpoint-19750/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..60c74a784b753058cc22fc76f3c890565b043fe1 --- /dev/null +++ b/checkpoint-19750/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffe37c975bc267c889d9d9139b73e7e0e05c4f9b520beec17e6033b2aca1db58 +size 5620161007 diff --git a/checkpoint-19750/rng_state_0.pth b/checkpoint-19750/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..bd0551596167940810fc379c8c42de17a936d054 --- /dev/null +++ b/checkpoint-19750/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e1f35a347eb686b2291e18e0db1d702a416a741c9ace2da6c890df09553aa0cb +size 15024 diff --git a/checkpoint-19750/rng_state_1.pth b/checkpoint-19750/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..926201758ccf07c9f283ebf601aa921e7453f595 --- /dev/null +++ b/checkpoint-19750/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:926d1772beb31047bb47e570f4dba18862d61be379ef0b3cadae6c6c5a644f9c +size 15024 diff --git a/checkpoint-19750/rng_state_2.pth b/checkpoint-19750/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..ba76b98e9007f47286765e92069bceeab0da41b9 --- /dev/null +++ b/checkpoint-19750/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:526945f32b7c061a9cfaf536859a93b109646f9df223b6c130b041004e2c88cb +size 15024 diff --git a/checkpoint-19750/rng_state_3.pth b/checkpoint-19750/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..602445836612fd718e42219e8f7ed82daa2fc186 --- /dev/null +++ b/checkpoint-19750/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e93efac603aca1058289269de0968ecee8bb583ee7e98143566e736b0965383 +size 15024 diff --git a/checkpoint-19750/scheduler.pt b/checkpoint-19750/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..249a7bbb7074bf9bb4ed38217ee72862b6c4e9ad --- /dev/null +++ b/checkpoint-19750/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0cf42ee96064c9ca1d57e07f00fc20d9087286fe5d6a6918f1c5a4f1503347be +size 1064 diff --git a/checkpoint-19750/trainer_state.json b/checkpoint-19750/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..9049a0867e4236cadfef290ead5cb01cc299cc1c --- /dev/null +++ b/checkpoint-19750/trainer_state.json @@ -0,0 +1,138283 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.890531362365162, + "eval_steps": 500, + "global_step": 19750, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00039952057530962844, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 3.1824, + "step": 1 + }, + { + "epoch": 0.0007990411506192569, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.9778, + "step": 2 + }, + { + "epoch": 0.0011985617259288853, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 2.7824, + "step": 3 + }, + { + "epoch": 0.0015980823012385138, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 2.6074, + "step": 4 + }, + { + "epoch": 0.001997602876548142, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.4209, + "step": 5 + }, + { + "epoch": 0.0023971234518577705, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 2.2482, + "step": 6 + }, + { + "epoch": 0.002796644027167399, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 2.1017, + "step": 7 + }, + { + "epoch": 0.0031961646024770275, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 2.0034, + "step": 8 + }, + { + "epoch": 0.003595685177786656, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.9567, + "step": 9 + }, + { + "epoch": 0.003995205753096284, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.9417, + "step": 10 + }, + { + "epoch": 0.0043947263284059125, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.9526, + "step": 11 + }, + { + "epoch": 0.004794246903715541, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.9613, + "step": 12 + }, + { + "epoch": 0.0051937674790251695, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.9467, + "step": 13 + }, + { + "epoch": 0.005593288054334798, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.9215, + "step": 14 + }, + { + "epoch": 0.0059928086296444265, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.9154, + "step": 15 + }, + { + "epoch": 0.006392329204954055, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.9064, + "step": 16 + }, + { + "epoch": 0.0067918497802636835, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.888, + "step": 17 + }, + { + "epoch": 0.007191370355573312, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.8729, + "step": 18 + }, + { + "epoch": 0.0075908909308829405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8425, + "step": 19 + }, + { + "epoch": 0.007990411506192568, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.8468, + "step": 20 + }, + { + "epoch": 0.008389932081502197, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.8418, + "step": 21 + }, + { + "epoch": 0.008789452656811825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.825, + "step": 22 + }, + { + "epoch": 0.009188973232121454, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8176, + "step": 23 + }, + { + "epoch": 0.009588493807431082, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.7827, + "step": 24 + }, + { + "epoch": 0.00998801438274071, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.7695, + "step": 25 + }, + { + "epoch": 0.010387534958050339, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.78, + "step": 26 + }, + { + "epoch": 0.010787055533359968, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.7588, + "step": 27 + }, + { + "epoch": 0.011186576108669596, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.743, + "step": 28 + }, + { + "epoch": 0.011586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.7357, + "step": 29 + }, + { + "epoch": 0.011985617259288853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.7271, + "step": 30 + }, + { + "epoch": 0.012385137834598482, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.7053, + "step": 31 + }, + { + "epoch": 0.01278465840990811, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 32 + }, + { + "epoch": 0.013184178985217739, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.6831, + "step": 33 + }, + { + "epoch": 0.013583699560527367, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 34 + }, + { + "epoch": 0.013983220135836995, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6682, + "step": 35 + }, + { + "epoch": 0.014382740711146624, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6573, + "step": 36 + }, + { + "epoch": 0.014782261286456252, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.6507, + "step": 37 + }, + { + "epoch": 0.015181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.6424, + "step": 38 + }, + { + "epoch": 0.01558130243707551, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.6251, + "step": 39 + }, + { + "epoch": 0.015980823012385136, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.6215, + "step": 40 + }, + { + "epoch": 0.016380343587694766, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.6226, + "step": 41 + }, + { + "epoch": 0.016779864163004393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.6005, + "step": 42 + }, + { + "epoch": 0.017179384738314023, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.6224, + "step": 43 + }, + { + "epoch": 0.01757890531362365, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5901, + "step": 44 + }, + { + "epoch": 0.01797842588893328, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5962, + "step": 45 + }, + { + "epoch": 0.018377946464242907, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 46 + }, + { + "epoch": 0.018777467039552537, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5697, + "step": 47 + }, + { + "epoch": 0.019176987614862164, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 48 + }, + { + "epoch": 0.019576508190171794, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.5531, + "step": 49 + }, + { + "epoch": 0.01997602876548142, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.554, + "step": 50 + }, + { + "epoch": 0.02037554934079105, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.5539, + "step": 51 + }, + { + "epoch": 0.020775069916100678, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.5591, + "step": 52 + }, + { + "epoch": 0.02117459049141031, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.545, + "step": 53 + }, + { + "epoch": 0.021574111066719935, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5221, + "step": 54 + }, + { + "epoch": 0.021973631642029565, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.5234, + "step": 55 + }, + { + "epoch": 0.022373152217339192, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.5186, + "step": 56 + }, + { + "epoch": 0.022772672792648822, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.5273, + "step": 57 + }, + { + "epoch": 0.02317219336795845, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5201, + "step": 58 + }, + { + "epoch": 0.02357171394326808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5108, + "step": 59 + }, + { + "epoch": 0.023971234518577706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.5078, + "step": 60 + }, + { + "epoch": 0.024370755093887336, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.5026, + "step": 61 + }, + { + "epoch": 0.024770275669196963, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.5076, + "step": 62 + }, + { + "epoch": 0.025169796244506593, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4956, + "step": 63 + }, + { + "epoch": 0.02556931681981622, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4925, + "step": 64 + }, + { + "epoch": 0.02596883739512585, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.5045, + "step": 65 + }, + { + "epoch": 0.026368357970435477, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.4843, + "step": 66 + }, + { + "epoch": 0.026767878545745107, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4807, + "step": 67 + }, + { + "epoch": 0.027167399121054734, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 68 + }, + { + "epoch": 0.027566919696364364, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4784, + "step": 69 + }, + { + "epoch": 0.02796644027167399, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4758, + "step": 70 + }, + { + "epoch": 0.02836596084698362, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4871, + "step": 71 + }, + { + "epoch": 0.028765481422293248, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4753, + "step": 72 + }, + { + "epoch": 0.029165001997602878, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4922, + "step": 73 + }, + { + "epoch": 0.029564522572912505, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4759, + "step": 74 + }, + { + "epoch": 0.02996404314822213, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.4703, + "step": 75 + }, + { + "epoch": 0.030363563723531762, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4666, + "step": 76 + }, + { + "epoch": 0.03076308429884139, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4628, + "step": 77 + }, + { + "epoch": 0.03116260487415102, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.4656, + "step": 78 + }, + { + "epoch": 0.031562125449460646, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4695, + "step": 79 + }, + { + "epoch": 0.03196164602477027, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4517, + "step": 80 + }, + { + "epoch": 0.032361166600079906, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4577, + "step": 81 + }, + { + "epoch": 0.03276068717538953, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4526, + "step": 82 + }, + { + "epoch": 0.03316020775069916, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4476, + "step": 83 + }, + { + "epoch": 0.033559728326008786, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.4499, + "step": 84 + }, + { + "epoch": 0.03395924890131842, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.4549, + "step": 85 + }, + { + "epoch": 0.03435876947662805, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4548, + "step": 86 + }, + { + "epoch": 0.034758290051937674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4525, + "step": 87 + }, + { + "epoch": 0.0351578106272473, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 88 + }, + { + "epoch": 0.035557331202556934, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4343, + "step": 89 + }, + { + "epoch": 0.03595685177786656, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4398, + "step": 90 + }, + { + "epoch": 0.03635637235317619, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4431, + "step": 91 + }, + { + "epoch": 0.036755892928485814, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.438, + "step": 92 + }, + { + "epoch": 0.03715541350379545, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4291, + "step": 93 + }, + { + "epoch": 0.037554934079105075, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.4373, + "step": 94 + }, + { + "epoch": 0.0379544546544147, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4409, + "step": 95 + }, + { + "epoch": 0.03835397522972433, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4305, + "step": 96 + }, + { + "epoch": 0.03875349580503396, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.4337, + "step": 97 + }, + { + "epoch": 0.03915301638034359, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4309, + "step": 98 + }, + { + "epoch": 0.039552536955653216, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 99 + }, + { + "epoch": 0.03995205753096284, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 100 + }, + { + "epoch": 0.040351578106272476, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.42, + "step": 101 + }, + { + "epoch": 0.0407510986815821, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4211, + "step": 102 + }, + { + "epoch": 0.04115061925689173, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4342, + "step": 103 + }, + { + "epoch": 0.041550139832201356, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4191, + "step": 104 + }, + { + "epoch": 0.04194966040751099, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.4234, + "step": 105 + }, + { + "epoch": 0.04234918098282062, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 106 + }, + { + "epoch": 0.042748701558130243, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4308, + "step": 107 + }, + { + "epoch": 0.04314822213343987, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.4189, + "step": 108 + }, + { + "epoch": 0.043547742708749504, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 109 + }, + { + "epoch": 0.04394726328405913, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4278, + "step": 110 + }, + { + "epoch": 0.04434678385936876, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 111 + }, + { + "epoch": 0.044746304434678384, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.4276, + "step": 112 + }, + { + "epoch": 0.04514582500998801, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.4203, + "step": 113 + }, + { + "epoch": 0.045545345585297645, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4313, + "step": 114 + }, + { + "epoch": 0.04594486616060727, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4186, + "step": 115 + }, + { + "epoch": 0.0463443867359169, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4215, + "step": 116 + }, + { + "epoch": 0.046743907311226525, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.4272, + "step": 117 + }, + { + "epoch": 0.04714342788653616, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4262, + "step": 118 + }, + { + "epoch": 0.047542948461845785, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 119 + }, + { + "epoch": 0.04794246903715541, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4071, + "step": 120 + }, + { + "epoch": 0.04834198961246504, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4241, + "step": 121 + }, + { + "epoch": 0.04874151018777467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 122 + }, + { + "epoch": 0.0491410307630843, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4252, + "step": 123 + }, + { + "epoch": 0.049540551338393926, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.424, + "step": 124 + }, + { + "epoch": 0.04994007191370355, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 125 + }, + { + "epoch": 0.05033959248901319, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.412, + "step": 126 + }, + { + "epoch": 0.05073911306432281, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4199, + "step": 127 + }, + { + "epoch": 0.05113863363963244, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 128 + }, + { + "epoch": 0.05153815421494207, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 129 + }, + { + "epoch": 0.0519376747902517, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 130 + }, + { + "epoch": 0.05233719536556133, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4049, + "step": 131 + }, + { + "epoch": 0.052736715940870954, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4134, + "step": 132 + }, + { + "epoch": 0.05313623651618058, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 133 + }, + { + "epoch": 0.053535757091490214, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4156, + "step": 134 + }, + { + "epoch": 0.05393527766679984, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 135 + }, + { + "epoch": 0.05433479824210947, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 136 + }, + { + "epoch": 0.054734318817419095, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 137 + }, + { + "epoch": 0.05513383939272873, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 138 + }, + { + "epoch": 0.055533359968038355, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.428, + "step": 139 + }, + { + "epoch": 0.05593288054334798, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 140 + }, + { + "epoch": 0.05633240111865761, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.4083, + "step": 141 + }, + { + "epoch": 0.05673192169396724, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 142 + }, + { + "epoch": 0.05713144226927687, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 143 + }, + { + "epoch": 0.057530962844586496, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 144 + }, + { + "epoch": 0.05793048341989612, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 145 + }, + { + "epoch": 0.058330003995205756, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 146 + }, + { + "epoch": 0.05872952457051538, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.4168, + "step": 147 + }, + { + "epoch": 0.05912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 148 + }, + { + "epoch": 0.05952856572113464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 149 + }, + { + "epoch": 0.05992808629644426, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 150 + }, + { + "epoch": 0.0603276068717539, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4171, + "step": 151 + }, + { + "epoch": 0.060727127447063524, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 152 + }, + { + "epoch": 0.06112664802237315, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 153 + }, + { + "epoch": 0.06152616859768278, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3966, + "step": 154 + }, + { + "epoch": 0.06192568917299241, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.4121, + "step": 155 + }, + { + "epoch": 0.06232520974830204, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 156 + }, + { + "epoch": 0.06272473032361167, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 157 + }, + { + "epoch": 0.06312425089892129, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 158 + }, + { + "epoch": 0.06352377147423093, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.413, + "step": 159 + }, + { + "epoch": 0.06392329204954054, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4174, + "step": 160 + }, + { + "epoch": 0.06432281262485018, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4133, + "step": 161 + }, + { + "epoch": 0.06472233320015981, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 162 + }, + { + "epoch": 0.06512185377546943, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4128, + "step": 163 + }, + { + "epoch": 0.06552137435077907, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3945, + "step": 164 + }, + { + "epoch": 0.0659208949260887, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 165 + }, + { + "epoch": 0.06632041550139832, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 166 + }, + { + "epoch": 0.06671993607670795, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4104, + "step": 167 + }, + { + "epoch": 0.06711945665201757, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.4144, + "step": 168 + }, + { + "epoch": 0.0675189772273272, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4059, + "step": 169 + }, + { + "epoch": 0.06791849780263684, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 170 + }, + { + "epoch": 0.06831801837794646, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 171 + }, + { + "epoch": 0.0687175389532561, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 172 + }, + { + "epoch": 0.06911705952856573, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 173 + }, + { + "epoch": 0.06951658010387535, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 174 + }, + { + "epoch": 0.06991610067918498, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4097, + "step": 175 + }, + { + "epoch": 0.0703156212544946, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 176 + }, + { + "epoch": 0.07071514182980423, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4052, + "step": 177 + }, + { + "epoch": 0.07111466240511387, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.417, + "step": 178 + }, + { + "epoch": 0.07151418298042349, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 179 + }, + { + "epoch": 0.07191370355573312, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 180 + }, + { + "epoch": 0.07231322413104276, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 181 + }, + { + "epoch": 0.07271274470635238, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 182 + }, + { + "epoch": 0.07311226528166201, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 183 + }, + { + "epoch": 0.07351178585697163, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 184 + }, + { + "epoch": 0.07391130643228126, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 185 + }, + { + "epoch": 0.0743108270075909, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 186 + }, + { + "epoch": 0.07471034758290052, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 187 + }, + { + "epoch": 0.07510986815821015, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 188 + }, + { + "epoch": 0.07550938873351977, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.4132, + "step": 189 + }, + { + "epoch": 0.0759089093088294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.4067, + "step": 190 + }, + { + "epoch": 0.07630842988413904, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 191 + }, + { + "epoch": 0.07670795045944866, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 192 + }, + { + "epoch": 0.07710747103475829, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3964, + "step": 193 + }, + { + "epoch": 0.07750699161006792, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 194 + }, + { + "epoch": 0.07790651218537754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3979, + "step": 195 + }, + { + "epoch": 0.07830603276068718, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 196 + }, + { + "epoch": 0.0787055533359968, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4111, + "step": 197 + }, + { + "epoch": 0.07910507391130643, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 198 + }, + { + "epoch": 0.07950459448661606, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 199 + }, + { + "epoch": 0.07990411506192568, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 200 + }, + { + "epoch": 0.08030363563723532, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 201 + }, + { + "epoch": 0.08070315621254495, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 202 + }, + { + "epoch": 0.08110267678785457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4021, + "step": 203 + }, + { + "epoch": 0.0815021973631642, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 204 + }, + { + "epoch": 0.08190171793847383, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 205 + }, + { + "epoch": 0.08230123851378346, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 206 + }, + { + "epoch": 0.08270075908909309, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.4057, + "step": 207 + }, + { + "epoch": 0.08310027966440271, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4158, + "step": 208 + }, + { + "epoch": 0.08349980023971235, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3898, + "step": 209 + }, + { + "epoch": 0.08389932081502198, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 210 + }, + { + "epoch": 0.0842988413903316, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.4193, + "step": 211 + }, + { + "epoch": 0.08469836196564123, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 212 + }, + { + "epoch": 0.08509788254095085, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 213 + }, + { + "epoch": 0.08549740311626049, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 214 + }, + { + "epoch": 0.08589692369157012, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4098, + "step": 215 + }, + { + "epoch": 0.08629644426687974, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.4015, + "step": 216 + }, + { + "epoch": 0.08669596484218937, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.4113, + "step": 217 + }, + { + "epoch": 0.08709548541749901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 218 + }, + { + "epoch": 0.08749500599280863, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 219 + }, + { + "epoch": 0.08789452656811826, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 220 + }, + { + "epoch": 0.08829404714342788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 221 + }, + { + "epoch": 0.08869356771873751, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.409, + "step": 222 + }, + { + "epoch": 0.08909308829404715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 223 + }, + { + "epoch": 0.08949260886935677, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 224 + }, + { + "epoch": 0.0898921294446664, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 225 + }, + { + "epoch": 0.09029165001997602, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 226 + }, + { + "epoch": 0.09069117059528566, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3904, + "step": 227 + }, + { + "epoch": 0.09109069117059529, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 228 + }, + { + "epoch": 0.09149021174590491, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 229 + }, + { + "epoch": 0.09188973232121454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 230 + }, + { + "epoch": 0.09228925289652418, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 231 + }, + { + "epoch": 0.0926887734718338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 232 + }, + { + "epoch": 0.09308829404714343, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 233 + }, + { + "epoch": 0.09348781462245305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 234 + }, + { + "epoch": 0.09388733519776268, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 235 + }, + { + "epoch": 0.09428685577307232, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 236 + }, + { + "epoch": 0.09468637634838194, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 237 + }, + { + "epoch": 0.09508589692369157, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 238 + }, + { + "epoch": 0.0954854174990012, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 239 + }, + { + "epoch": 0.09588493807431082, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4064, + "step": 240 + }, + { + "epoch": 0.09628445864962046, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 241 + }, + { + "epoch": 0.09668397922493008, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 242 + }, + { + "epoch": 0.09708349980023971, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 243 + }, + { + "epoch": 0.09748302037554935, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 244 + }, + { + "epoch": 0.09788254095085897, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3961, + "step": 245 + }, + { + "epoch": 0.0982820615261686, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4061, + "step": 246 + }, + { + "epoch": 0.09868158210147823, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 247 + }, + { + "epoch": 0.09908110267678785, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 248 + }, + { + "epoch": 0.09948062325209749, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 249 + }, + { + "epoch": 0.0998801438274071, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 250 + }, + { + "epoch": 0.10027966440271674, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 251 + }, + { + "epoch": 0.10067918497802637, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 252 + }, + { + "epoch": 0.10107870555333599, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.4024, + "step": 253 + }, + { + "epoch": 0.10147822612864563, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 254 + }, + { + "epoch": 0.10187774670395526, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 255 + }, + { + "epoch": 0.10227726727926488, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 256 + }, + { + "epoch": 0.10267678785457451, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 257 + }, + { + "epoch": 0.10307630842988413, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 258 + }, + { + "epoch": 0.10347582900519377, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 259 + }, + { + "epoch": 0.1038753495805034, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 260 + }, + { + "epoch": 0.10427487015581302, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 261 + }, + { + "epoch": 0.10467439073112265, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 262 + }, + { + "epoch": 0.10507391130643227, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 263 + }, + { + "epoch": 0.10547343188174191, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 264 + }, + { + "epoch": 0.10587295245705154, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3988, + "step": 265 + }, + { + "epoch": 0.10627247303236116, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 266 + }, + { + "epoch": 0.1066719936076708, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.4079, + "step": 267 + }, + { + "epoch": 0.10707151418298043, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 268 + }, + { + "epoch": 0.10747103475829005, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3954, + "step": 269 + }, + { + "epoch": 0.10787055533359968, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 270 + }, + { + "epoch": 0.1082700759089093, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 271 + }, + { + "epoch": 0.10866959648421894, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 272 + }, + { + "epoch": 0.10906911705952857, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 273 + }, + { + "epoch": 0.10946863763483819, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3899, + "step": 274 + }, + { + "epoch": 0.10986815821014782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 275 + }, + { + "epoch": 0.11026767878545746, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4012, + "step": 276 + }, + { + "epoch": 0.11066719936076708, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 277 + }, + { + "epoch": 0.11106671993607671, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3948, + "step": 278 + }, + { + "epoch": 0.11146624051138633, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4175, + "step": 279 + }, + { + "epoch": 0.11186576108669596, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 280 + }, + { + "epoch": 0.1122652816620056, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 281 + }, + { + "epoch": 0.11266480223731522, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3994, + "step": 282 + }, + { + "epoch": 0.11306432281262485, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 283 + }, + { + "epoch": 0.11346384338793448, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 284 + }, + { + "epoch": 0.1138633639632441, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 285 + }, + { + "epoch": 0.11426288453855374, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 286 + }, + { + "epoch": 0.11466240511386336, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4045, + "step": 287 + }, + { + "epoch": 0.11506192568917299, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 288 + }, + { + "epoch": 0.11546144626448263, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 289 + }, + { + "epoch": 0.11586096683979225, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 290 + }, + { + "epoch": 0.11626048741510188, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 291 + }, + { + "epoch": 0.11666000799041151, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 292 + }, + { + "epoch": 0.11705952856572113, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 293 + }, + { + "epoch": 0.11745904914103077, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3991, + "step": 294 + }, + { + "epoch": 0.11785856971634039, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 295 + }, + { + "epoch": 0.11825809029165002, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 296 + }, + { + "epoch": 0.11865761086695965, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 297 + }, + { + "epoch": 0.11905713144226927, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 298 + }, + { + "epoch": 0.11945665201757891, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 299 + }, + { + "epoch": 0.11985617259288853, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 300 + }, + { + "epoch": 0.12025569316819816, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3998, + "step": 301 + }, + { + "epoch": 0.1206552137435078, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 302 + }, + { + "epoch": 0.12105473431881741, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 303 + }, + { + "epoch": 0.12145425489412705, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 304 + }, + { + "epoch": 0.12185377546943668, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 305 + }, + { + "epoch": 0.1222532960447463, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 306 + }, + { + "epoch": 0.12265281662005593, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 307 + }, + { + "epoch": 0.12305233719536555, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 308 + }, + { + "epoch": 0.12345185777067519, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 309 + }, + { + "epoch": 0.12385137834598482, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 310 + }, + { + "epoch": 0.12425089892129444, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 311 + }, + { + "epoch": 0.12465041949660408, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 312 + }, + { + "epoch": 0.1250499400719137, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 313 + }, + { + "epoch": 0.12544946064722334, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 314 + }, + { + "epoch": 0.12584898122253296, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 315 + }, + { + "epoch": 0.12624850179784258, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 316 + }, + { + "epoch": 0.12664802237315223, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 317 + }, + { + "epoch": 0.12704754294846185, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 318 + }, + { + "epoch": 0.12744706352377147, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 319 + }, + { + "epoch": 0.1278465840990811, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 320 + }, + { + "epoch": 0.12824610467439074, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 321 + }, + { + "epoch": 0.12864562524970036, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3987, + "step": 322 + }, + { + "epoch": 0.12904514582500998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 323 + }, + { + "epoch": 0.12944466640031962, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 324 + }, + { + "epoch": 0.12984418697562924, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 325 + }, + { + "epoch": 0.13024370755093886, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 326 + }, + { + "epoch": 0.1306432281262485, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 327 + }, + { + "epoch": 0.13104274870155813, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 328 + }, + { + "epoch": 0.13144226927686775, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 329 + }, + { + "epoch": 0.1318417898521774, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 330 + }, + { + "epoch": 0.13224131042748702, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 331 + }, + { + "epoch": 0.13264083100279664, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 332 + }, + { + "epoch": 0.1330403515781063, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 333 + }, + { + "epoch": 0.1334398721534159, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 334 + }, + { + "epoch": 0.13383939272872553, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 335 + }, + { + "epoch": 0.13423891330403515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 336 + }, + { + "epoch": 0.1346384338793448, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 337 + }, + { + "epoch": 0.1350379544546544, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 338 + }, + { + "epoch": 0.13543747502996403, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 339 + }, + { + "epoch": 0.13583699560527368, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 340 + }, + { + "epoch": 0.1362365161805833, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 341 + }, + { + "epoch": 0.13663603675589292, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 342 + }, + { + "epoch": 0.13703555733120257, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3889, + "step": 343 + }, + { + "epoch": 0.1374350779065122, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3995, + "step": 344 + }, + { + "epoch": 0.1378345984818218, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 345 + }, + { + "epoch": 0.13823411905713145, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 346 + }, + { + "epoch": 0.13863363963244107, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3983, + "step": 347 + }, + { + "epoch": 0.1390331602077507, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 348 + }, + { + "epoch": 0.13943268078306031, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 349 + }, + { + "epoch": 0.13983220135836996, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 350 + }, + { + "epoch": 0.14023172193367958, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 351 + }, + { + "epoch": 0.1406312425089892, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 352 + }, + { + "epoch": 0.14103076308429885, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 353 + }, + { + "epoch": 0.14143028365960847, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.391, + "step": 354 + }, + { + "epoch": 0.1418298042349181, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3937, + "step": 355 + }, + { + "epoch": 0.14222932481022774, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3858, + "step": 356 + }, + { + "epoch": 0.14262884538553736, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 357 + }, + { + "epoch": 0.14302836596084698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 358 + }, + { + "epoch": 0.14342788653615662, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 359 + }, + { + "epoch": 0.14382740711146624, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 360 + }, + { + "epoch": 0.14422692768677586, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 361 + }, + { + "epoch": 0.1446264482620855, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 362 + }, + { + "epoch": 0.14502596883739513, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 363 + }, + { + "epoch": 0.14542548941270475, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 364 + }, + { + "epoch": 0.14582500998801437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 365 + }, + { + "epoch": 0.14622453056332402, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 366 + }, + { + "epoch": 0.14662405113863364, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 367 + }, + { + "epoch": 0.14702357171394326, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 368 + }, + { + "epoch": 0.1474230922892529, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 369 + }, + { + "epoch": 0.14782261286456252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 370 + }, + { + "epoch": 0.14822213343987214, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 371 + }, + { + "epoch": 0.1486216540151818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 372 + }, + { + "epoch": 0.1490211745904914, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 373 + }, + { + "epoch": 0.14942069516580103, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3942, + "step": 374 + }, + { + "epoch": 0.14982021574111068, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 375 + }, + { + "epoch": 0.1502197363164203, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 376 + }, + { + "epoch": 0.15061925689172992, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3903, + "step": 377 + }, + { + "epoch": 0.15101877746703954, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 378 + }, + { + "epoch": 0.1514182980423492, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 379 + }, + { + "epoch": 0.1518178186176588, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 380 + }, + { + "epoch": 0.15221733919296843, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 381 + }, + { + "epoch": 0.15261685976827807, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 382 + }, + { + "epoch": 0.1530163803435877, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 383 + }, + { + "epoch": 0.1534159009188973, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 384 + }, + { + "epoch": 0.15381542149420696, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 385 + }, + { + "epoch": 0.15421494206951658, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 386 + }, + { + "epoch": 0.1546144626448262, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 387 + }, + { + "epoch": 0.15501398322013585, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 388 + }, + { + "epoch": 0.15541350379544547, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 389 + }, + { + "epoch": 0.1558130243707551, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 390 + }, + { + "epoch": 0.15621254494606474, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 391 + }, + { + "epoch": 0.15661206552137436, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 392 + }, + { + "epoch": 0.15701158609668397, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 393 + }, + { + "epoch": 0.1574111066719936, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 394 + }, + { + "epoch": 0.15781062724730324, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 395 + }, + { + "epoch": 0.15821014782261286, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 396 + }, + { + "epoch": 0.15860966839792248, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 397 + }, + { + "epoch": 0.15900918897323213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 398 + }, + { + "epoch": 0.15940870954854175, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 399 + }, + { + "epoch": 0.15980823012385137, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3929, + "step": 400 + }, + { + "epoch": 0.16020775069916102, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3835, + "step": 401 + }, + { + "epoch": 0.16060727127447064, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 402 + }, + { + "epoch": 0.16100679184978026, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 403 + }, + { + "epoch": 0.1614063124250899, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.4032, + "step": 404 + }, + { + "epoch": 0.16180583300039952, + "grad_norm": 3.984375, + "learning_rate": 0.0002, + "loss": 1.48, + "step": 405 + }, + { + "epoch": 0.16220535357570914, + "grad_norm": 2.234375, + "learning_rate": 0.0002, + "loss": 1.4185, + "step": 406 + }, + { + "epoch": 0.1626048741510188, + "grad_norm": 3.921875, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 407 + }, + { + "epoch": 0.1630043947263284, + "grad_norm": 3.578125, + "learning_rate": 0.0002, + "loss": 1.4966, + "step": 408 + }, + { + "epoch": 0.16340391530163803, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 409 + }, + { + "epoch": 0.16380343587694765, + "grad_norm": 2.671875, + "learning_rate": 0.0002, + "loss": 1.4878, + "step": 410 + }, + { + "epoch": 0.1642029564522573, + "grad_norm": 2.34375, + "learning_rate": 0.0002, + "loss": 1.491, + "step": 411 + }, + { + "epoch": 0.16460247702756692, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 412 + }, + { + "epoch": 0.16500199760287654, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4232, + "step": 413 + }, + { + "epoch": 0.16540151817818619, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.4109, + "step": 414 + }, + { + "epoch": 0.1658010387534958, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.4187, + "step": 415 + }, + { + "epoch": 0.16620055932880543, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 416 + }, + { + "epoch": 0.16660007990411507, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 417 + }, + { + "epoch": 0.1669996004794247, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 418 + }, + { + "epoch": 0.1673991210547343, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3982, + "step": 419 + }, + { + "epoch": 0.16779864163004396, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 420 + }, + { + "epoch": 0.16819816220535358, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 421 + }, + { + "epoch": 0.1685976827806632, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 422 + }, + { + "epoch": 0.16899720335597282, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3933, + "step": 423 + }, + { + "epoch": 0.16939672393128247, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 424 + }, + { + "epoch": 0.1697962445065921, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3909, + "step": 425 + }, + { + "epoch": 0.1701957650819017, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3978, + "step": 426 + }, + { + "epoch": 0.17059528565721135, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 427 + }, + { + "epoch": 0.17099480623252097, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3955, + "step": 428 + }, + { + "epoch": 0.1713943268078306, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 429 + }, + { + "epoch": 0.17179384738314024, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 430 + }, + { + "epoch": 0.17219336795844986, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3965, + "step": 431 + }, + { + "epoch": 0.17259288853375948, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 432 + }, + { + "epoch": 0.17299240910906913, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 433 + }, + { + "epoch": 0.17339192968437875, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 434 + }, + { + "epoch": 0.17379145025968837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 435 + }, + { + "epoch": 0.17419097083499802, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 436 + }, + { + "epoch": 0.17459049141030764, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 437 + }, + { + "epoch": 0.17499001198561726, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3928, + "step": 438 + }, + { + "epoch": 0.17538953256092688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 439 + }, + { + "epoch": 0.17578905313623652, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 440 + }, + { + "epoch": 0.17618857371154614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 441 + }, + { + "epoch": 0.17658809428685576, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3912, + "step": 442 + }, + { + "epoch": 0.1769876148621654, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 443 + }, + { + "epoch": 0.17738713543747503, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 444 + }, + { + "epoch": 0.17778665601278465, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 445 + }, + { + "epoch": 0.1781861765880943, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 446 + }, + { + "epoch": 0.17858569716340392, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 447 + }, + { + "epoch": 0.17898521773871354, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 448 + }, + { + "epoch": 0.17938473831402318, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 449 + }, + { + "epoch": 0.1797842588893328, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 450 + }, + { + "epoch": 0.18018377946464242, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 451 + }, + { + "epoch": 0.18058330003995204, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 452 + }, + { + "epoch": 0.1809828206152617, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 453 + }, + { + "epoch": 0.1813823411905713, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 454 + }, + { + "epoch": 0.18178186176588093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.4001, + "step": 455 + }, + { + "epoch": 0.18218138234119058, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 456 + }, + { + "epoch": 0.1825809029165002, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 457 + }, + { + "epoch": 0.18298042349180982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 458 + }, + { + "epoch": 0.18337994406711947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 459 + }, + { + "epoch": 0.18377946464242909, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 460 + }, + { + "epoch": 0.1841789852177387, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 461 + }, + { + "epoch": 0.18457850579304835, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 462 + }, + { + "epoch": 0.18497802636835797, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 463 + }, + { + "epoch": 0.1853775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 464 + }, + { + "epoch": 0.18577706751897724, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 465 + }, + { + "epoch": 0.18617658809428686, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3796, + "step": 466 + }, + { + "epoch": 0.18657610866959648, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 467 + }, + { + "epoch": 0.1869756292449061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 468 + }, + { + "epoch": 0.18737514982021575, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 469 + }, + { + "epoch": 0.18777467039552537, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 470 + }, + { + "epoch": 0.188174190970835, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 471 + }, + { + "epoch": 0.18857371154614463, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 472 + }, + { + "epoch": 0.18897323212145425, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 473 + }, + { + "epoch": 0.18937275269676387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 474 + }, + { + "epoch": 0.18977227327207352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 475 + }, + { + "epoch": 0.19017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 476 + }, + { + "epoch": 0.19057131442269276, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 477 + }, + { + "epoch": 0.1909708349980024, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 478 + }, + { + "epoch": 0.19137035557331203, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 479 + }, + { + "epoch": 0.19176987614862165, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3807, + "step": 480 + }, + { + "epoch": 0.1921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 481 + }, + { + "epoch": 0.19256891729924092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 482 + }, + { + "epoch": 0.19296843787455054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 483 + }, + { + "epoch": 0.19336795844986016, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 484 + }, + { + "epoch": 0.1937674790251698, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 485 + }, + { + "epoch": 0.19416699960047942, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 486 + }, + { + "epoch": 0.19456652017578904, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 487 + }, + { + "epoch": 0.1949660407510987, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 488 + }, + { + "epoch": 0.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 489 + }, + { + "epoch": 0.19576508190171793, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 490 + }, + { + "epoch": 0.19616460247702758, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3816, + "step": 491 + }, + { + "epoch": 0.1965641230523372, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 492 + }, + { + "epoch": 0.19696364362764682, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3824, + "step": 493 + }, + { + "epoch": 0.19736316420295646, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.387, + "step": 494 + }, + { + "epoch": 0.19776268477826608, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 495 + }, + { + "epoch": 0.1981622053535757, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 496 + }, + { + "epoch": 0.19856172592888532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 497 + }, + { + "epoch": 0.19896124650419497, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 498 + }, + { + "epoch": 0.1993607670795046, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 499 + }, + { + "epoch": 0.1997602876548142, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 500 + }, + { + "epoch": 0.20015980823012386, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 501 + }, + { + "epoch": 0.20055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 502 + }, + { + "epoch": 0.2009588493807431, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 503 + }, + { + "epoch": 0.20135836995605275, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 504 + }, + { + "epoch": 0.20175789053136237, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 505 + }, + { + "epoch": 0.20215741110667199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 506 + }, + { + "epoch": 0.20255693168198163, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 507 + }, + { + "epoch": 0.20295645225729125, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 508 + }, + { + "epoch": 0.20335597283260087, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 509 + }, + { + "epoch": 0.20375549340791052, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 510 + }, + { + "epoch": 0.20415501398322014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.384, + "step": 511 + }, + { + "epoch": 0.20455453455852976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 512 + }, + { + "epoch": 0.20495405513383938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 513 + }, + { + "epoch": 0.20535357570914903, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 514 + }, + { + "epoch": 0.20575309628445865, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 515 + }, + { + "epoch": 0.20615261685976827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 516 + }, + { + "epoch": 0.20655213743507791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.38, + "step": 517 + }, + { + "epoch": 0.20695165801038753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 518 + }, + { + "epoch": 0.20735117858569715, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 519 + }, + { + "epoch": 0.2077506991610068, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 520 + }, + { + "epoch": 0.20815021973631642, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 521 + }, + { + "epoch": 0.20854974031162604, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 522 + }, + { + "epoch": 0.2089492608869357, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 523 + }, + { + "epoch": 0.2093487814622453, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 524 + }, + { + "epoch": 0.20974830203755493, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3845, + "step": 525 + }, + { + "epoch": 0.21014782261286455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 526 + }, + { + "epoch": 0.2105473431881742, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 527 + }, + { + "epoch": 0.21094686376348382, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3958, + "step": 528 + }, + { + "epoch": 0.21134638433879344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 529 + }, + { + "epoch": 0.21174590491410308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 530 + }, + { + "epoch": 0.2121454254894127, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 531 + }, + { + "epoch": 0.21254494606472232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 532 + }, + { + "epoch": 0.21294446664003197, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 533 + }, + { + "epoch": 0.2133439872153416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 534 + }, + { + "epoch": 0.2137435077906512, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 535 + }, + { + "epoch": 0.21414302836596086, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 536 + }, + { + "epoch": 0.21454254894127048, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 537 + }, + { + "epoch": 0.2149420695165801, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 538 + }, + { + "epoch": 0.21534159009188975, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 539 + }, + { + "epoch": 0.21574111066719937, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 540 + }, + { + "epoch": 0.21614063124250898, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 541 + }, + { + "epoch": 0.2165401518178186, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 542 + }, + { + "epoch": 0.21693967239312825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3754, + "step": 543 + }, + { + "epoch": 0.21733919296843787, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 544 + }, + { + "epoch": 0.2177387135437475, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 545 + }, + { + "epoch": 0.21813823411905714, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 546 + }, + { + "epoch": 0.21853775469436676, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3887, + "step": 547 + }, + { + "epoch": 0.21893727526967638, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 548 + }, + { + "epoch": 0.21933679584498603, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 549 + }, + { + "epoch": 0.21973631642029565, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 550 + }, + { + "epoch": 0.22013583699560527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 551 + }, + { + "epoch": 0.2205353575709149, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 552 + }, + { + "epoch": 0.22093487814622453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 553 + }, + { + "epoch": 0.22133439872153415, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 554 + }, + { + "epoch": 0.2217339192968438, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3992, + "step": 555 + }, + { + "epoch": 0.22213343987215342, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 556 + }, + { + "epoch": 0.22253296044746304, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 557 + }, + { + "epoch": 0.22293248102277266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 558 + }, + { + "epoch": 0.2233320015980823, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 559 + }, + { + "epoch": 0.22373152217339193, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 560 + }, + { + "epoch": 0.22413104274870155, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 561 + }, + { + "epoch": 0.2245305633240112, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 562 + }, + { + "epoch": 0.22493008389932082, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 563 + }, + { + "epoch": 0.22532960447463043, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 564 + }, + { + "epoch": 0.22572912504994008, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 565 + }, + { + "epoch": 0.2261286456252497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 566 + }, + { + "epoch": 0.22652816620055932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 567 + }, + { + "epoch": 0.22692768677586897, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 568 + }, + { + "epoch": 0.2273272073511786, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 569 + }, + { + "epoch": 0.2277267279264882, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 570 + }, + { + "epoch": 0.22812624850179783, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 571 + }, + { + "epoch": 0.22852576907710748, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 572 + }, + { + "epoch": 0.2289252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 573 + }, + { + "epoch": 0.22932481022772672, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 574 + }, + { + "epoch": 0.22972433080303636, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 575 + }, + { + "epoch": 0.23012385137834598, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 576 + }, + { + "epoch": 0.2305233719536556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3786, + "step": 577 + }, + { + "epoch": 0.23092289252896525, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 578 + }, + { + "epoch": 0.23132241310427487, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 579 + }, + { + "epoch": 0.2317219336795845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 580 + }, + { + "epoch": 0.23212145425489414, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 581 + }, + { + "epoch": 0.23252097483020376, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3781, + "step": 582 + }, + { + "epoch": 0.23292049540551338, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 583 + }, + { + "epoch": 0.23332001598082303, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 584 + }, + { + "epoch": 0.23371953655613265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 585 + }, + { + "epoch": 0.23411905713144227, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 586 + }, + { + "epoch": 0.23451857770675189, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 587 + }, + { + "epoch": 0.23491809828206153, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 588 + }, + { + "epoch": 0.23531761885737115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 589 + }, + { + "epoch": 0.23571713943268077, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 590 + }, + { + "epoch": 0.23611666000799042, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 591 + }, + { + "epoch": 0.23651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 592 + }, + { + "epoch": 0.23691570115860966, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 593 + }, + { + "epoch": 0.2373152217339193, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 594 + }, + { + "epoch": 0.23771474230922893, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 595 + }, + { + "epoch": 0.23811426288453855, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 596 + }, + { + "epoch": 0.2385137834598482, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3914, + "step": 597 + }, + { + "epoch": 0.23891330403515781, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 598 + }, + { + "epoch": 0.23931282461046743, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 599 + }, + { + "epoch": 0.23971234518577705, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 600 + }, + { + "epoch": 0.2401118657610867, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 601 + }, + { + "epoch": 0.24051138633639632, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 602 + }, + { + "epoch": 0.24091090691170594, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 603 + }, + { + "epoch": 0.2413104274870156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 604 + }, + { + "epoch": 0.2417099480623252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 605 + }, + { + "epoch": 0.24210946863763483, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 606 + }, + { + "epoch": 0.24250898921294448, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 607 + }, + { + "epoch": 0.2429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 608 + }, + { + "epoch": 0.24330803036356372, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 609 + }, + { + "epoch": 0.24370755093887336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 610 + }, + { + "epoch": 0.24410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 611 + }, + { + "epoch": 0.2445065920894926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 612 + }, + { + "epoch": 0.24490611266480225, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 613 + }, + { + "epoch": 0.24530563324011187, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 614 + }, + { + "epoch": 0.2457051538154215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 615 + }, + { + "epoch": 0.2461046743907311, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3783, + "step": 616 + }, + { + "epoch": 0.24650419496604076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 617 + }, + { + "epoch": 0.24690371554135038, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 618 + }, + { + "epoch": 0.24730323611666, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 619 + }, + { + "epoch": 0.24770275669196964, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 620 + }, + { + "epoch": 0.24810227726727926, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 621 + }, + { + "epoch": 0.24850179784258888, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 622 + }, + { + "epoch": 0.24890131841789853, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 623 + }, + { + "epoch": 0.24930083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 624 + }, + { + "epoch": 0.24970035956851777, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 625 + }, + { + "epoch": 0.2500998801438274, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 626 + }, + { + "epoch": 0.250499400719137, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3748, + "step": 627 + }, + { + "epoch": 0.2508989212944467, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 628 + }, + { + "epoch": 0.2512984418697563, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.377, + "step": 629 + }, + { + "epoch": 0.2516979624450659, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 630 + }, + { + "epoch": 0.25209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 631 + }, + { + "epoch": 0.25249700359568517, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 632 + }, + { + "epoch": 0.2528965241709948, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 633 + }, + { + "epoch": 0.25329604474630446, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 634 + }, + { + "epoch": 0.2536955653216141, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 635 + }, + { + "epoch": 0.2540950858969237, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 636 + }, + { + "epoch": 0.2544946064722333, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 637 + }, + { + "epoch": 0.25489412704754294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 638 + }, + { + "epoch": 0.25529364762285256, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 639 + }, + { + "epoch": 0.2556931681981622, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 640 + }, + { + "epoch": 0.25609268877347185, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 641 + }, + { + "epoch": 0.2564922093487815, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 642 + }, + { + "epoch": 0.2568917299240911, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 643 + }, + { + "epoch": 0.2572912504994007, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 644 + }, + { + "epoch": 0.25769077107471033, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 645 + }, + { + "epoch": 0.25809029165001995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 646 + }, + { + "epoch": 0.25848981222532963, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 647 + }, + { + "epoch": 0.25888933280063925, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 648 + }, + { + "epoch": 0.25928885337594887, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 649 + }, + { + "epoch": 0.2596883739512585, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 650 + }, + { + "epoch": 0.2600878945265681, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 651 + }, + { + "epoch": 0.26048741510187773, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 652 + }, + { + "epoch": 0.26088693567718735, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3866, + "step": 653 + }, + { + "epoch": 0.261286456252497, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 654 + }, + { + "epoch": 0.26168597682780664, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 655 + }, + { + "epoch": 0.26208549740311626, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 656 + }, + { + "epoch": 0.2624850179784259, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 657 + }, + { + "epoch": 0.2628845385537355, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3802, + "step": 658 + }, + { + "epoch": 0.2632840591290451, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 659 + }, + { + "epoch": 0.2636835797043548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 660 + }, + { + "epoch": 0.2640831002796644, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 661 + }, + { + "epoch": 0.26448262085497404, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 662 + }, + { + "epoch": 0.26488214143028366, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 663 + }, + { + "epoch": 0.2652816620055933, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 664 + }, + { + "epoch": 0.2656811825809029, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 665 + }, + { + "epoch": 0.2660807031562126, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 666 + }, + { + "epoch": 0.2664802237315222, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 667 + }, + { + "epoch": 0.2668797443068318, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 668 + }, + { + "epoch": 0.26727926488214143, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 669 + }, + { + "epoch": 0.26767878545745105, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 670 + }, + { + "epoch": 0.26807830603276067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 671 + }, + { + "epoch": 0.2684778266080703, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 672 + }, + { + "epoch": 0.26887734718337997, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 673 + }, + { + "epoch": 0.2692768677586896, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 674 + }, + { + "epoch": 0.2696763883339992, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 675 + }, + { + "epoch": 0.2700759089093088, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 676 + }, + { + "epoch": 0.27047542948461845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 677 + }, + { + "epoch": 0.27087495005992807, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 678 + }, + { + "epoch": 0.27127447063523774, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 679 + }, + { + "epoch": 0.27167399121054736, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 680 + }, + { + "epoch": 0.272073511785857, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 681 + }, + { + "epoch": 0.2724730323611666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 682 + }, + { + "epoch": 0.2728725529364762, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 683 + }, + { + "epoch": 0.27327207351178584, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 684 + }, + { + "epoch": 0.27367159408709546, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 685 + }, + { + "epoch": 0.27407111466240514, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3747, + "step": 686 + }, + { + "epoch": 0.27447063523771476, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 687 + }, + { + "epoch": 0.2748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 688 + }, + { + "epoch": 0.275269676388334, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 689 + }, + { + "epoch": 0.2756691969636436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 690 + }, + { + "epoch": 0.27606871753895323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 691 + }, + { + "epoch": 0.2764682381142629, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 692 + }, + { + "epoch": 0.27686775868957253, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 693 + }, + { + "epoch": 0.27726727926488215, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 694 + }, + { + "epoch": 0.27766679984019177, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3842, + "step": 695 + }, + { + "epoch": 0.2780663204155014, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 696 + }, + { + "epoch": 0.278465840990811, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 697 + }, + { + "epoch": 0.27886536156612063, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 698 + }, + { + "epoch": 0.2792648821414303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 699 + }, + { + "epoch": 0.2796644027167399, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 700 + }, + { + "epoch": 0.28006392329204954, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3742, + "step": 701 + }, + { + "epoch": 0.28046344386735916, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 702 + }, + { + "epoch": 0.2808629644426688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 703 + }, + { + "epoch": 0.2812624850179784, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 704 + }, + { + "epoch": 0.2816620055932881, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 705 + }, + { + "epoch": 0.2820615261685977, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 706 + }, + { + "epoch": 0.2824610467439073, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 707 + }, + { + "epoch": 0.28286056731921694, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 708 + }, + { + "epoch": 0.28326008789452656, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 709 + }, + { + "epoch": 0.2836596084698362, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 710 + }, + { + "epoch": 0.2840591290451458, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 711 + }, + { + "epoch": 0.2844586496204555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 712 + }, + { + "epoch": 0.2848581701957651, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 713 + }, + { + "epoch": 0.2852576907710747, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 714 + }, + { + "epoch": 0.28565721134638433, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 715 + }, + { + "epoch": 0.28605673192169395, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 716 + }, + { + "epoch": 0.28645625249700357, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 717 + }, + { + "epoch": 0.28685577307231325, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 718 + }, + { + "epoch": 0.28725529364762287, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 719 + }, + { + "epoch": 0.2876548142229325, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 720 + }, + { + "epoch": 0.2880543347982421, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3826, + "step": 721 + }, + { + "epoch": 0.2884538553735517, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 722 + }, + { + "epoch": 0.28885337594886135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 723 + }, + { + "epoch": 0.289252896524171, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 724 + }, + { + "epoch": 0.28965241709948064, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 725 + }, + { + "epoch": 0.29005193767479026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 726 + }, + { + "epoch": 0.2904514582500999, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 727 + }, + { + "epoch": 0.2908509788254095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 728 + }, + { + "epoch": 0.2912504994007191, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3854, + "step": 729 + }, + { + "epoch": 0.29165001997602874, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 730 + }, + { + "epoch": 0.2920495405513384, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 731 + }, + { + "epoch": 0.29244906112664804, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 732 + }, + { + "epoch": 0.29284858170195766, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 733 + }, + { + "epoch": 0.2932481022772673, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 734 + }, + { + "epoch": 0.2936476228525769, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 735 + }, + { + "epoch": 0.2940471434278865, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 736 + }, + { + "epoch": 0.2944466640031962, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 737 + }, + { + "epoch": 0.2948461845785058, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 738 + }, + { + "epoch": 0.29524570515381543, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 739 + }, + { + "epoch": 0.29564522572912505, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 740 + }, + { + "epoch": 0.29604474630443467, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 741 + }, + { + "epoch": 0.2964442668797443, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 742 + }, + { + "epoch": 0.2968437874550539, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 743 + }, + { + "epoch": 0.2972433080303636, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 744 + }, + { + "epoch": 0.2976428286056732, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 745 + }, + { + "epoch": 0.2980423491809828, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 746 + }, + { + "epoch": 0.29844186975629244, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 747 + }, + { + "epoch": 0.29884139033160206, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 748 + }, + { + "epoch": 0.2992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 749 + }, + { + "epoch": 0.29964043148222136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.382, + "step": 750 + }, + { + "epoch": 0.300039952057531, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 751 + }, + { + "epoch": 0.3004394726328406, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 752 + }, + { + "epoch": 0.3008389932081502, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 753 + }, + { + "epoch": 0.30123851378345984, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 754 + }, + { + "epoch": 0.30163803435876946, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 755 + }, + { + "epoch": 0.3020375549340791, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 756 + }, + { + "epoch": 0.30243707550938875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 757 + }, + { + "epoch": 0.3028365960846984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 758 + }, + { + "epoch": 0.303236116660008, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 759 + }, + { + "epoch": 0.3036356372353176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3698, + "step": 760 + }, + { + "epoch": 0.30403515781062723, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 761 + }, + { + "epoch": 0.30443467838593685, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3761, + "step": 762 + }, + { + "epoch": 0.3048341989612465, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 763 + }, + { + "epoch": 0.30523371953655615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 764 + }, + { + "epoch": 0.30563324011186577, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 765 + }, + { + "epoch": 0.3060327606871754, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 766 + }, + { + "epoch": 0.306432281262485, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 767 + }, + { + "epoch": 0.3068318018377946, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 768 + }, + { + "epoch": 0.3072313224131043, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3694, + "step": 769 + }, + { + "epoch": 0.3076308429884139, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 770 + }, + { + "epoch": 0.30803036356372354, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 771 + }, + { + "epoch": 0.30842988413903316, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 772 + }, + { + "epoch": 0.3088294047143428, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 773 + }, + { + "epoch": 0.3092289252896524, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 774 + }, + { + "epoch": 0.309628445864962, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3775, + "step": 775 + }, + { + "epoch": 0.3100279664402717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 776 + }, + { + "epoch": 0.3104274870155813, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 777 + }, + { + "epoch": 0.31082700759089094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 778 + }, + { + "epoch": 0.31122652816620056, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 779 + }, + { + "epoch": 0.3116260487415102, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 780 + }, + { + "epoch": 0.3120255693168198, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 781 + }, + { + "epoch": 0.31242508989212947, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 782 + }, + { + "epoch": 0.3128246104674391, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 783 + }, + { + "epoch": 0.3132241310427487, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 784 + }, + { + "epoch": 0.31362365161805833, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 785 + }, + { + "epoch": 0.31402317219336795, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 786 + }, + { + "epoch": 0.31442269276867757, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 787 + }, + { + "epoch": 0.3148222133439872, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 788 + }, + { + "epoch": 0.31522173391929686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 789 + }, + { + "epoch": 0.3156212544946065, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 790 + }, + { + "epoch": 0.3160207750699161, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 791 + }, + { + "epoch": 0.3164202956452257, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3728, + "step": 792 + }, + { + "epoch": 0.31681981622053534, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 793 + }, + { + "epoch": 0.31721933679584496, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 794 + }, + { + "epoch": 0.31761885737115464, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 795 + }, + { + "epoch": 0.31801837794646426, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 796 + }, + { + "epoch": 0.3184178985217739, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 797 + }, + { + "epoch": 0.3188174190970835, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 798 + }, + { + "epoch": 0.3192169396723931, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 799 + }, + { + "epoch": 0.31961646024770274, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 800 + }, + { + "epoch": 0.32001598082301236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 801 + }, + { + "epoch": 0.32041550139832203, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 802 + }, + { + "epoch": 0.32081502197363165, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 803 + }, + { + "epoch": 0.3212145425489413, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 804 + }, + { + "epoch": 0.3216140631242509, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 805 + }, + { + "epoch": 0.3220135836995605, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 806 + }, + { + "epoch": 0.32241310427487013, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 807 + }, + { + "epoch": 0.3228126248501798, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 808 + }, + { + "epoch": 0.3232121454254894, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 809 + }, + { + "epoch": 0.32361166600079905, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 810 + }, + { + "epoch": 0.32401118657610867, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 811 + }, + { + "epoch": 0.3244107071514183, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 812 + }, + { + "epoch": 0.3248102277267279, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 813 + }, + { + "epoch": 0.3252097483020376, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.373, + "step": 814 + }, + { + "epoch": 0.3256092688773472, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 815 + }, + { + "epoch": 0.3260087894526568, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 816 + }, + { + "epoch": 0.32640831002796644, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 817 + }, + { + "epoch": 0.32680783060327606, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 818 + }, + { + "epoch": 0.3272073511785857, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 819 + }, + { + "epoch": 0.3276068717538953, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 820 + }, + { + "epoch": 0.328006392329205, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 821 + }, + { + "epoch": 0.3284059129045146, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 822 + }, + { + "epoch": 0.3288054334798242, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 823 + }, + { + "epoch": 0.32920495405513384, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 824 + }, + { + "epoch": 0.32960447463044346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 825 + }, + { + "epoch": 0.3300039952057531, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 826 + }, + { + "epoch": 0.33040351578106275, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 827 + }, + { + "epoch": 0.33080303635637237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 828 + }, + { + "epoch": 0.331202556931682, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 829 + }, + { + "epoch": 0.3316020775069916, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 830 + }, + { + "epoch": 0.33200159808230123, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 831 + }, + { + "epoch": 0.33240111865761085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 832 + }, + { + "epoch": 0.33280063923292047, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 833 + }, + { + "epoch": 0.33320015980823015, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 834 + }, + { + "epoch": 0.33359968038353977, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 835 + }, + { + "epoch": 0.3339992009588494, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 836 + }, + { + "epoch": 0.334398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 837 + }, + { + "epoch": 0.3347982421094686, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 838 + }, + { + "epoch": 0.33519776268477824, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 839 + }, + { + "epoch": 0.3355972832600879, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 840 + }, + { + "epoch": 0.33599680383539754, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 841 + }, + { + "epoch": 0.33639632441070716, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 842 + }, + { + "epoch": 0.3367958449860168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 843 + }, + { + "epoch": 0.3371953655613264, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 844 + }, + { + "epoch": 0.337594886136636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 845 + }, + { + "epoch": 0.33799440671194564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 846 + }, + { + "epoch": 0.3383939272872553, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 847 + }, + { + "epoch": 0.33879344786256493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 848 + }, + { + "epoch": 0.33919296843787455, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 849 + }, + { + "epoch": 0.3395924890131842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 850 + }, + { + "epoch": 0.3399920095884938, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 851 + }, + { + "epoch": 0.3403915301638034, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 852 + }, + { + "epoch": 0.3407910507391131, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 853 + }, + { + "epoch": 0.3411905713144227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 854 + }, + { + "epoch": 0.34159009188973233, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 855 + }, + { + "epoch": 0.34198961246504195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 856 + }, + { + "epoch": 0.34238913304035157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 857 + }, + { + "epoch": 0.3427886536156612, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 858 + }, + { + "epoch": 0.34318817419097086, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 859 + }, + { + "epoch": 0.3435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 860 + }, + { + "epoch": 0.3439872153415901, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 861 + }, + { + "epoch": 0.3443867359168997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 862 + }, + { + "epoch": 0.34478625649220934, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 863 + }, + { + "epoch": 0.34518577706751896, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 864 + }, + { + "epoch": 0.3455852976428286, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 865 + }, + { + "epoch": 0.34598481821813826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 866 + }, + { + "epoch": 0.3463843387934479, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3776, + "step": 867 + }, + { + "epoch": 0.3467838593687575, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 868 + }, + { + "epoch": 0.3471833799440671, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 869 + }, + { + "epoch": 0.34758290051937674, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 870 + }, + { + "epoch": 0.34798242109468636, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 871 + }, + { + "epoch": 0.34838194166999603, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 872 + }, + { + "epoch": 0.34878146224530565, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 873 + }, + { + "epoch": 0.34918098282061527, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 874 + }, + { + "epoch": 0.3495805033959249, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 875 + }, + { + "epoch": 0.3499800239712345, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 876 + }, + { + "epoch": 0.35037954454654413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 877 + }, + { + "epoch": 0.35077906512185375, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 878 + }, + { + "epoch": 0.3511785856971634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 879 + }, + { + "epoch": 0.35157810627247305, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 880 + }, + { + "epoch": 0.35197762684778267, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 881 + }, + { + "epoch": 0.3523771474230923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 882 + }, + { + "epoch": 0.3527766679984019, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 883 + }, + { + "epoch": 0.3531761885737115, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 884 + }, + { + "epoch": 0.3535757091490212, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 885 + }, + { + "epoch": 0.3539752297243308, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 886 + }, + { + "epoch": 0.35437475029964044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 887 + }, + { + "epoch": 0.35477427087495006, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3751, + "step": 888 + }, + { + "epoch": 0.3551737914502597, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 889 + }, + { + "epoch": 0.3555733120255693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 890 + }, + { + "epoch": 0.3559728326008789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 891 + }, + { + "epoch": 0.3563723531761886, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 892 + }, + { + "epoch": 0.3567718737514982, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 893 + }, + { + "epoch": 0.35717139432680783, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 894 + }, + { + "epoch": 0.35757091490211745, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 895 + }, + { + "epoch": 0.3579704354774271, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3758, + "step": 896 + }, + { + "epoch": 0.3583699560527367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 897 + }, + { + "epoch": 0.35876947662804637, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 898 + }, + { + "epoch": 0.359168997203356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 899 + }, + { + "epoch": 0.3595685177786656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 900 + }, + { + "epoch": 0.35996803835397523, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 901 + }, + { + "epoch": 0.36036755892928485, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3774, + "step": 902 + }, + { + "epoch": 0.36076707950459447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 903 + }, + { + "epoch": 0.3611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 904 + }, + { + "epoch": 0.36156612065521376, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 905 + }, + { + "epoch": 0.3619656412305234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 906 + }, + { + "epoch": 0.362365161805833, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 907 + }, + { + "epoch": 0.3627646823811426, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 908 + }, + { + "epoch": 0.36316420295645224, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 909 + }, + { + "epoch": 0.36356372353176186, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 910 + }, + { + "epoch": 0.36396324410707154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 911 + }, + { + "epoch": 0.36436276468238116, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 912 + }, + { + "epoch": 0.3647622852576908, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 913 + }, + { + "epoch": 0.3651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 914 + }, + { + "epoch": 0.36556132640831, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 915 + }, + { + "epoch": 0.36596084698361964, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 916 + }, + { + "epoch": 0.3663603675589293, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 917 + }, + { + "epoch": 0.36675988813423893, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 918 + }, + { + "epoch": 0.36715940870954855, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 919 + }, + { + "epoch": 0.36755892928485817, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 920 + }, + { + "epoch": 0.3679584498601678, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 921 + }, + { + "epoch": 0.3683579704354774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 922 + }, + { + "epoch": 0.36875749101078703, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 923 + }, + { + "epoch": 0.3691570115860967, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 924 + }, + { + "epoch": 0.3695565321614063, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 925 + }, + { + "epoch": 0.36995605273671595, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 926 + }, + { + "epoch": 0.37035557331202557, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 927 + }, + { + "epoch": 0.3707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 928 + }, + { + "epoch": 0.3711546144626448, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 929 + }, + { + "epoch": 0.3715541350379545, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 930 + }, + { + "epoch": 0.3719536556132641, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 931 + }, + { + "epoch": 0.3723531761885737, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 932 + }, + { + "epoch": 0.37275269676388334, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 933 + }, + { + "epoch": 0.37315221733919296, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 934 + }, + { + "epoch": 0.3735517379145026, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 935 + }, + { + "epoch": 0.3739512584898122, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 936 + }, + { + "epoch": 0.3743507790651219, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 937 + }, + { + "epoch": 0.3747502996404315, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 938 + }, + { + "epoch": 0.3751498202157411, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 939 + }, + { + "epoch": 0.37554934079105073, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 940 + }, + { + "epoch": 0.37594886136636035, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 941 + }, + { + "epoch": 0.37634838194167, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 942 + }, + { + "epoch": 0.37674790251697965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 943 + }, + { + "epoch": 0.37714742309228927, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 944 + }, + { + "epoch": 0.3775469436675989, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 945 + }, + { + "epoch": 0.3779464642429085, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 946 + }, + { + "epoch": 0.37834598481821813, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 947 + }, + { + "epoch": 0.37874550539352775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 948 + }, + { + "epoch": 0.37914502596883737, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 949 + }, + { + "epoch": 0.37954454654414704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 950 + }, + { + "epoch": 0.37994406711945666, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 951 + }, + { + "epoch": 0.3803435876947663, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 952 + }, + { + "epoch": 0.3807431082700759, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 953 + }, + { + "epoch": 0.3811426288453855, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 954 + }, + { + "epoch": 0.38154214942069514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 955 + }, + { + "epoch": 0.3819416699960048, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 956 + }, + { + "epoch": 0.38234119057131444, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 957 + }, + { + "epoch": 0.38274071114662406, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 958 + }, + { + "epoch": 0.3831402317219337, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 959 + }, + { + "epoch": 0.3835397522972433, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 960 + }, + { + "epoch": 0.3839392728725529, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 961 + }, + { + "epoch": 0.3843387934478626, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3661, + "step": 962 + }, + { + "epoch": 0.3847383140231722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 963 + }, + { + "epoch": 0.38513783459848183, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 964 + }, + { + "epoch": 0.38553735517379145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 965 + }, + { + "epoch": 0.38593687574910107, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 966 + }, + { + "epoch": 0.3863363963244107, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 967 + }, + { + "epoch": 0.3867359168997203, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 968 + }, + { + "epoch": 0.38713543747503, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 969 + }, + { + "epoch": 0.3875349580503396, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 970 + }, + { + "epoch": 0.3879344786256492, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 971 + }, + { + "epoch": 0.38833399920095885, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 972 + }, + { + "epoch": 0.38873351977626847, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 973 + }, + { + "epoch": 0.3891330403515781, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 974 + }, + { + "epoch": 0.38953256092688776, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 975 + }, + { + "epoch": 0.3899320815021974, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 976 + }, + { + "epoch": 0.390331602077507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 977 + }, + { + "epoch": 0.3907311226528166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 978 + }, + { + "epoch": 0.39113064322812624, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 979 + }, + { + "epoch": 0.39153016380343586, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 980 + }, + { + "epoch": 0.3919296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 981 + }, + { + "epoch": 0.39232920495405516, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 982 + }, + { + "epoch": 0.3927287255293648, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 983 + }, + { + "epoch": 0.3931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 984 + }, + { + "epoch": 0.393527766679984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 985 + }, + { + "epoch": 0.39392728725529363, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 986 + }, + { + "epoch": 0.39432680783060325, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 987 + }, + { + "epoch": 0.39472632840591293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 988 + }, + { + "epoch": 0.39512584898122255, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 989 + }, + { + "epoch": 0.39552536955653217, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 990 + }, + { + "epoch": 0.3959248901318418, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 991 + }, + { + "epoch": 0.3963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 992 + }, + { + "epoch": 0.39672393128246103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 993 + }, + { + "epoch": 0.39712345185777065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 994 + }, + { + "epoch": 0.3975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 995 + }, + { + "epoch": 0.39792249300838994, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 996 + }, + { + "epoch": 0.39832201358369956, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 997 + }, + { + "epoch": 0.3987215341590092, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 998 + }, + { + "epoch": 0.3991210547343188, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 999 + }, + { + "epoch": 0.3995205753096284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1000 + }, + { + "epoch": 0.3999200958849381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1001 + }, + { + "epoch": 0.4003196164602477, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1002 + }, + { + "epoch": 0.40071913703555734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1003 + }, + { + "epoch": 0.40111865761086696, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1004 + }, + { + "epoch": 0.4015181781861766, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 1005 + }, + { + "epoch": 0.4019176987614862, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1006 + }, + { + "epoch": 0.4023172193367959, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1007 + }, + { + "epoch": 0.4027167399121055, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1008 + }, + { + "epoch": 0.4031162604874151, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1009 + }, + { + "epoch": 0.40351578106272473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1010 + }, + { + "epoch": 0.40391530163803435, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1011 + }, + { + "epoch": 0.40431482221334397, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1012 + }, + { + "epoch": 0.4047143427886536, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 1013 + }, + { + "epoch": 0.40511386336396327, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1014 + }, + { + "epoch": 0.4055133839392729, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1015 + }, + { + "epoch": 0.4059129045145825, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1016 + }, + { + "epoch": 0.4063124250898921, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1017 + }, + { + "epoch": 0.40671194566520175, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 1018 + }, + { + "epoch": 0.40711146624051137, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1019 + }, + { + "epoch": 0.40751098681582104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1020 + }, + { + "epoch": 0.40791050739113066, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1021 + }, + { + "epoch": 0.4083100279664403, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1022 + }, + { + "epoch": 0.4087095485417499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1023 + }, + { + "epoch": 0.4091090691170595, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 1024 + }, + { + "epoch": 0.40950858969236914, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1025 + }, + { + "epoch": 0.40990811026767876, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1026 + }, + { + "epoch": 0.41030763084298844, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1027 + }, + { + "epoch": 0.41070715141829806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1028 + }, + { + "epoch": 0.4111066719936077, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1029 + }, + { + "epoch": 0.4115061925689173, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1030 + }, + { + "epoch": 0.4119057131442269, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1031 + }, + { + "epoch": 0.41230523371953653, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1032 + }, + { + "epoch": 0.4127047542948462, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1033 + }, + { + "epoch": 0.41310427487015583, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1034 + }, + { + "epoch": 0.41350379544546545, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1035 + }, + { + "epoch": 0.41390331602077507, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1036 + }, + { + "epoch": 0.4143028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1037 + }, + { + "epoch": 0.4147023571713943, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1038 + }, + { + "epoch": 0.41510187774670393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1039 + }, + { + "epoch": 0.4155013983220136, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1040 + }, + { + "epoch": 0.4159009188973232, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1041 + }, + { + "epoch": 0.41630043947263284, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1042 + }, + { + "epoch": 0.41669996004794246, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1043 + }, + { + "epoch": 0.4170994806232521, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1044 + }, + { + "epoch": 0.4174990011985617, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1045 + }, + { + "epoch": 0.4178985217738714, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 1046 + }, + { + "epoch": 0.418298042349181, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1047 + }, + { + "epoch": 0.4186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1048 + }, + { + "epoch": 0.41909708349980024, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3711, + "step": 1049 + }, + { + "epoch": 0.41949660407510986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1050 + }, + { + "epoch": 0.4198961246504195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1051 + }, + { + "epoch": 0.4202956452257291, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1052 + }, + { + "epoch": 0.4206951658010388, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1053 + }, + { + "epoch": 0.4210946863763484, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1054 + }, + { + "epoch": 0.421494206951658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1055 + }, + { + "epoch": 0.42189372752696763, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1056 + }, + { + "epoch": 0.42229324810227725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 1057 + }, + { + "epoch": 0.42269276867758687, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1058 + }, + { + "epoch": 0.42309228925289655, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 1059 + }, + { + "epoch": 0.42349180982820617, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1060 + }, + { + "epoch": 0.4238913304035158, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1061 + }, + { + "epoch": 0.4242908509788254, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1062 + }, + { + "epoch": 0.424690371554135, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1063 + }, + { + "epoch": 0.42508989212944465, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1064 + }, + { + "epoch": 0.4254894127047543, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 1065 + }, + { + "epoch": 0.42588893328006394, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1066 + }, + { + "epoch": 0.42628845385537356, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1067 + }, + { + "epoch": 0.4266879744306832, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 1068 + }, + { + "epoch": 0.4270874950059928, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1069 + }, + { + "epoch": 0.4274870155813024, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1070 + }, + { + "epoch": 0.42788653615661204, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1071 + }, + { + "epoch": 0.4282860567319217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1072 + }, + { + "epoch": 0.42868557730723134, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1073 + }, + { + "epoch": 0.42908509788254096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1074 + }, + { + "epoch": 0.4294846184578506, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1075 + }, + { + "epoch": 0.4298841390331602, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1076 + }, + { + "epoch": 0.4302836596084698, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1077 + }, + { + "epoch": 0.4306831801837795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1078 + }, + { + "epoch": 0.4310827007590891, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1079 + }, + { + "epoch": 0.43148222133439873, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1080 + }, + { + "epoch": 0.43188174190970835, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1081 + }, + { + "epoch": 0.43228126248501797, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1082 + }, + { + "epoch": 0.4326807830603276, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 1083 + }, + { + "epoch": 0.4330803036356372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1084 + }, + { + "epoch": 0.4334798242109469, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1085 + }, + { + "epoch": 0.4338793447862565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1086 + }, + { + "epoch": 0.4342788653615661, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 1087 + }, + { + "epoch": 0.43467838593687574, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1088 + }, + { + "epoch": 0.43507790651218536, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1089 + }, + { + "epoch": 0.435477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1090 + }, + { + "epoch": 0.43587694766280466, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1091 + }, + { + "epoch": 0.4362764682381143, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1092 + }, + { + "epoch": 0.4366759888134239, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1093 + }, + { + "epoch": 0.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1094 + }, + { + "epoch": 0.43747502996404314, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1095 + }, + { + "epoch": 0.43787455053935276, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1096 + }, + { + "epoch": 0.4382740711146624, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1097 + }, + { + "epoch": 0.43867359168997205, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1098 + }, + { + "epoch": 0.4390731122652817, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1099 + }, + { + "epoch": 0.4394726328405913, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1100 + }, + { + "epoch": 0.4398721534159009, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1101 + }, + { + "epoch": 0.44027167399121053, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1102 + }, + { + "epoch": 0.44067119456652015, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 1103 + }, + { + "epoch": 0.4410707151418298, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1104 + }, + { + "epoch": 0.44147023571713945, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1105 + }, + { + "epoch": 0.44186975629244907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1106 + }, + { + "epoch": 0.4422692768677587, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1107 + }, + { + "epoch": 0.4426687974430683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1108 + }, + { + "epoch": 0.4430683180183779, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1109 + }, + { + "epoch": 0.4434678385936876, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 1110 + }, + { + "epoch": 0.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1111 + }, + { + "epoch": 0.44426687974430684, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1112 + }, + { + "epoch": 0.44466640031961646, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1113 + }, + { + "epoch": 0.4450659208949261, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1114 + }, + { + "epoch": 0.4454654414702357, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1115 + }, + { + "epoch": 0.4458649620455453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1116 + }, + { + "epoch": 0.446264482620855, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 1117 + }, + { + "epoch": 0.4466640031961646, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1118 + }, + { + "epoch": 0.44706352377147424, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1119 + }, + { + "epoch": 0.44746304434678386, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1120 + }, + { + "epoch": 0.4478625649220935, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1121 + }, + { + "epoch": 0.4482620854974031, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1122 + }, + { + "epoch": 0.44866160607271277, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1123 + }, + { + "epoch": 0.4490611266480224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1124 + }, + { + "epoch": 0.449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 1125 + }, + { + "epoch": 0.44986016779864163, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1126 + }, + { + "epoch": 0.45025968837395125, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1127 + }, + { + "epoch": 0.45065920894926087, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1128 + }, + { + "epoch": 0.4510587295245705, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1129 + }, + { + "epoch": 0.45145825009988017, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1130 + }, + { + "epoch": 0.4518577706751898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1131 + }, + { + "epoch": 0.4522572912504994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1132 + }, + { + "epoch": 0.452656811825809, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 1133 + }, + { + "epoch": 0.45305633240111864, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1134 + }, + { + "epoch": 0.45345585297642826, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1135 + }, + { + "epoch": 0.45385537355173794, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1136 + }, + { + "epoch": 0.45425489412704756, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1137 + }, + { + "epoch": 0.4546544147023572, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1138 + }, + { + "epoch": 0.4550539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 1139 + }, + { + "epoch": 0.4554534558529764, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1140 + }, + { + "epoch": 0.45585297642828604, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 1141 + }, + { + "epoch": 0.45625249700359566, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1142 + }, + { + "epoch": 0.45665201757890533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1143 + }, + { + "epoch": 0.45705153815421495, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1144 + }, + { + "epoch": 0.4574510587295246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1145 + }, + { + "epoch": 0.4578505793048342, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1146 + }, + { + "epoch": 0.4582500998801438, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1147 + }, + { + "epoch": 0.45864962045545343, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1148 + }, + { + "epoch": 0.4590491410307631, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1149 + }, + { + "epoch": 0.45944866160607273, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1150 + }, + { + "epoch": 0.45984818218138235, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3724, + "step": 1151 + }, + { + "epoch": 0.46024770275669197, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1152 + }, + { + "epoch": 0.4606472233320016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1153 + }, + { + "epoch": 0.4610467439073112, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1154 + }, + { + "epoch": 0.4614462644826209, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1155 + }, + { + "epoch": 0.4618457850579305, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1156 + }, + { + "epoch": 0.4622453056332401, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1157 + }, + { + "epoch": 0.46264482620854974, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1158 + }, + { + "epoch": 0.46304434678385936, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1159 + }, + { + "epoch": 0.463443867359169, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 1160 + }, + { + "epoch": 0.4638433879344786, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1161 + }, + { + "epoch": 0.4642429085097883, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1162 + }, + { + "epoch": 0.4646424290850979, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1163 + }, + { + "epoch": 0.4650419496604075, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 1164 + }, + { + "epoch": 0.46544147023571714, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1165 + }, + { + "epoch": 0.46584099081102676, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1166 + }, + { + "epoch": 0.4662405113863364, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1167 + }, + { + "epoch": 0.46664003196164605, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1168 + }, + { + "epoch": 0.46703955253695567, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1169 + }, + { + "epoch": 0.4674390731122653, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1170 + }, + { + "epoch": 0.4678385936875749, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1171 + }, + { + "epoch": 0.46823811426288453, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1172 + }, + { + "epoch": 0.46863763483819415, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1173 + }, + { + "epoch": 0.46903715541350377, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1174 + }, + { + "epoch": 0.46943667598881345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1175 + }, + { + "epoch": 0.46983619656412307, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1176 + }, + { + "epoch": 0.4702357171394327, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1177 + }, + { + "epoch": 0.4706352377147423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 1178 + }, + { + "epoch": 0.4710347582900519, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1179 + }, + { + "epoch": 0.47143427886536154, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1180 + }, + { + "epoch": 0.4718337994406712, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 1181 + }, + { + "epoch": 0.47223332001598084, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1182 + }, + { + "epoch": 0.47263284059129046, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1183 + }, + { + "epoch": 0.4730323611666001, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1184 + }, + { + "epoch": 0.4734318817419097, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1185 + }, + { + "epoch": 0.4738314023172193, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1186 + }, + { + "epoch": 0.47423092289252894, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1187 + }, + { + "epoch": 0.4746304434678386, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1188 + }, + { + "epoch": 0.47502996404314823, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1189 + }, + { + "epoch": 0.47542948461845785, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 1190 + }, + { + "epoch": 0.4758290051937675, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1191 + }, + { + "epoch": 0.4762285257690771, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1192 + }, + { + "epoch": 0.4766280463443867, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1193 + }, + { + "epoch": 0.4770275669196964, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1194 + }, + { + "epoch": 0.477427087495006, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1195 + }, + { + "epoch": 0.47782660807031563, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1196 + }, + { + "epoch": 0.47822612864562525, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1197 + }, + { + "epoch": 0.47862564922093487, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1198 + }, + { + "epoch": 0.4790251697962445, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1199 + }, + { + "epoch": 0.4794246903715541, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1200 + }, + { + "epoch": 0.4798242109468638, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1201 + }, + { + "epoch": 0.4802237315221734, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1202 + }, + { + "epoch": 0.480623252097483, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1203 + }, + { + "epoch": 0.48102277267279264, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1204 + }, + { + "epoch": 0.48142229324810226, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1205 + }, + { + "epoch": 0.4818218138234119, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1206 + }, + { + "epoch": 0.48222133439872156, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1207 + }, + { + "epoch": 0.4826208549740312, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1208 + }, + { + "epoch": 0.4830203755493408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1209 + }, + { + "epoch": 0.4834198961246504, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1210 + }, + { + "epoch": 0.48381941669996004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1211 + }, + { + "epoch": 0.48421893727526966, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1212 + }, + { + "epoch": 0.48461845785057933, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1213 + }, + { + "epoch": 0.48501797842588895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1214 + }, + { + "epoch": 0.48541749900119857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1215 + }, + { + "epoch": 0.4858170195765082, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1216 + }, + { + "epoch": 0.4862165401518178, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1217 + }, + { + "epoch": 0.48661606072712743, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 1218 + }, + { + "epoch": 0.48701558130243705, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1219 + }, + { + "epoch": 0.4874151018777467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1220 + }, + { + "epoch": 0.48781462245305635, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1221 + }, + { + "epoch": 0.48821414302836597, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1222 + }, + { + "epoch": 0.4886136636036756, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1223 + }, + { + "epoch": 0.4890131841789852, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1224 + }, + { + "epoch": 0.4894127047542948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1225 + }, + { + "epoch": 0.4898122253296045, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1226 + }, + { + "epoch": 0.4902117459049141, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1227 + }, + { + "epoch": 0.49061126648022374, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1228 + }, + { + "epoch": 0.49101078705553336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1229 + }, + { + "epoch": 0.491410307630843, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 1230 + }, + { + "epoch": 0.4918098282061526, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1231 + }, + { + "epoch": 0.4922093487814622, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1232 + }, + { + "epoch": 0.4926088693567719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1233 + }, + { + "epoch": 0.4930083899320815, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1234 + }, + { + "epoch": 0.49340791050739113, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1235 + }, + { + "epoch": 0.49380743108270075, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1236 + }, + { + "epoch": 0.4942069516580104, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 1237 + }, + { + "epoch": 0.49460647223332, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1238 + }, + { + "epoch": 0.49500599280862967, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1239 + }, + { + "epoch": 0.4954055133839393, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1240 + }, + { + "epoch": 0.4958050339592489, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1241 + }, + { + "epoch": 0.49620455453455853, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1242 + }, + { + "epoch": 0.49660407510986815, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1243 + }, + { + "epoch": 0.49700359568517777, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1244 + }, + { + "epoch": 0.4974031162604874, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1245 + }, + { + "epoch": 0.49780263683579706, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1246 + }, + { + "epoch": 0.4982021574111067, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1247 + }, + { + "epoch": 0.4986016779864163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1248 + }, + { + "epoch": 0.4990011985617259, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1249 + }, + { + "epoch": 0.49940071913703554, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1250 + }, + { + "epoch": 0.49980023971234516, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1251 + }, + { + "epoch": 0.5001997602876548, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1252 + }, + { + "epoch": 0.5005992808629645, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1253 + }, + { + "epoch": 0.500998801438274, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1254 + }, + { + "epoch": 0.5013983220135837, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1255 + }, + { + "epoch": 0.5017978425888934, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1256 + }, + { + "epoch": 0.5021973631642029, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1257 + }, + { + "epoch": 0.5025968837395126, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1258 + }, + { + "epoch": 0.5029964043148222, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1259 + }, + { + "epoch": 0.5033959248901319, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1260 + }, + { + "epoch": 0.5037954454654414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1261 + }, + { + "epoch": 0.5041949660407511, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1262 + }, + { + "epoch": 0.5045944866160608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 1263 + }, + { + "epoch": 0.5049940071913703, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1264 + }, + { + "epoch": 0.50539352776668, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1265 + }, + { + "epoch": 0.5057930483419896, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1266 + }, + { + "epoch": 0.5061925689172992, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1267 + }, + { + "epoch": 0.5065920894926089, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3678, + "step": 1268 + }, + { + "epoch": 0.5069916100679185, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.365, + "step": 1269 + }, + { + "epoch": 0.5073911306432282, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1270 + }, + { + "epoch": 0.5077906512185377, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1271 + }, + { + "epoch": 0.5081901717938474, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1272 + }, + { + "epoch": 0.508589692369157, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1273 + }, + { + "epoch": 0.5089892129444666, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1274 + }, + { + "epoch": 0.5093887335197763, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1275 + }, + { + "epoch": 0.5097882540950859, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1276 + }, + { + "epoch": 0.5101877746703956, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1277 + }, + { + "epoch": 0.5105872952457051, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 1278 + }, + { + "epoch": 0.5109868158210148, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1279 + }, + { + "epoch": 0.5113863363963244, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1280 + }, + { + "epoch": 0.511785856971634, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1281 + }, + { + "epoch": 0.5121853775469437, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1282 + }, + { + "epoch": 0.5125848981222533, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1283 + }, + { + "epoch": 0.512984418697563, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1284 + }, + { + "epoch": 0.5133839392728725, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1285 + }, + { + "epoch": 0.5137834598481822, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.368, + "step": 1286 + }, + { + "epoch": 0.5141829804234919, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1287 + }, + { + "epoch": 0.5145825009988014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1288 + }, + { + "epoch": 0.5149820215741111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1289 + }, + { + "epoch": 0.5153815421494207, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1290 + }, + { + "epoch": 0.5157810627247303, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1291 + }, + { + "epoch": 0.5161805833000399, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 1292 + }, + { + "epoch": 0.5165801038753496, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1293 + }, + { + "epoch": 0.5169796244506593, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1294 + }, + { + "epoch": 0.5173791450259688, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1295 + }, + { + "epoch": 0.5177786656012785, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 1296 + }, + { + "epoch": 0.5181781861765881, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1297 + }, + { + "epoch": 0.5185777067518977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1298 + }, + { + "epoch": 0.5189772273272073, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1299 + }, + { + "epoch": 0.519376747902517, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1300 + }, + { + "epoch": 0.5197762684778267, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1301 + }, + { + "epoch": 0.5201757890531362, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1302 + }, + { + "epoch": 0.5205753096284459, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1303 + }, + { + "epoch": 0.5209748302037555, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1304 + }, + { + "epoch": 0.5213743507790651, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1305 + }, + { + "epoch": 0.5217738713543747, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1306 + }, + { + "epoch": 0.5221733919296844, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1307 + }, + { + "epoch": 0.522572912504994, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1308 + }, + { + "epoch": 0.5229724330803036, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1309 + }, + { + "epoch": 0.5233719536556133, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1310 + }, + { + "epoch": 0.5237714742309229, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1311 + }, + { + "epoch": 0.5241709948062325, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1312 + }, + { + "epoch": 0.5245705153815422, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1313 + }, + { + "epoch": 0.5249700359568518, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1314 + }, + { + "epoch": 0.5253695565321614, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1315 + }, + { + "epoch": 0.525769077107471, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1316 + }, + { + "epoch": 0.5261685976827807, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1317 + }, + { + "epoch": 0.5265681182580902, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1318 + }, + { + "epoch": 0.5269676388333999, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 1319 + }, + { + "epoch": 0.5273671594087096, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1320 + }, + { + "epoch": 0.5277666799840192, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1321 + }, + { + "epoch": 0.5281662005593288, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1322 + }, + { + "epoch": 0.5285657211346384, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1323 + }, + { + "epoch": 0.5289652417099481, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1324 + }, + { + "epoch": 0.5293647622852576, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1325 + }, + { + "epoch": 0.5297642828605673, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1326 + }, + { + "epoch": 0.530163803435877, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1327 + }, + { + "epoch": 0.5305633240111866, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 1328 + }, + { + "epoch": 0.5309628445864962, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1329 + }, + { + "epoch": 0.5313623651618058, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1330 + }, + { + "epoch": 0.5317618857371155, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1331 + }, + { + "epoch": 0.5321614063124251, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1332 + }, + { + "epoch": 0.5325609268877347, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1333 + }, + { + "epoch": 0.5329604474630444, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1334 + }, + { + "epoch": 0.533359968038354, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1335 + }, + { + "epoch": 0.5337594886136636, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1336 + }, + { + "epoch": 0.5341590091889732, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1337 + }, + { + "epoch": 0.5345585297642829, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1338 + }, + { + "epoch": 0.5349580503395925, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1339 + }, + { + "epoch": 0.5353575709149021, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1340 + }, + { + "epoch": 0.5357570914902118, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1341 + }, + { + "epoch": 0.5361566120655213, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1342 + }, + { + "epoch": 0.536556132640831, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1343 + }, + { + "epoch": 0.5369556532161406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1344 + }, + { + "epoch": 0.5373551737914503, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1345 + }, + { + "epoch": 0.5377546943667599, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1346 + }, + { + "epoch": 0.5381542149420695, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1347 + }, + { + "epoch": 0.5385537355173792, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1348 + }, + { + "epoch": 0.5389532560926887, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 1349 + }, + { + "epoch": 0.5393527766679984, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1350 + }, + { + "epoch": 0.539752297243308, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1351 + }, + { + "epoch": 0.5401518178186177, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1352 + }, + { + "epoch": 0.5405513383939273, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1353 + }, + { + "epoch": 0.5409508589692369, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 1354 + }, + { + "epoch": 0.5413503795445466, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1355 + }, + { + "epoch": 0.5417499001198561, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1356 + }, + { + "epoch": 0.5421494206951658, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1357 + }, + { + "epoch": 0.5425489412704755, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1358 + }, + { + "epoch": 0.542948461845785, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1359 + }, + { + "epoch": 0.5433479824210947, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1360 + }, + { + "epoch": 0.5437475029964043, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1361 + }, + { + "epoch": 0.544147023571714, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1362 + }, + { + "epoch": 0.5445465441470235, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1363 + }, + { + "epoch": 0.5449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1364 + }, + { + "epoch": 0.5453455852976429, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1365 + }, + { + "epoch": 0.5457451058729524, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1366 + }, + { + "epoch": 0.5461446264482621, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1367 + }, + { + "epoch": 0.5465441470235717, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 1368 + }, + { + "epoch": 0.5469436675988814, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1369 + }, + { + "epoch": 0.5473431881741909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1370 + }, + { + "epoch": 0.5477427087495006, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1371 + }, + { + "epoch": 0.5481422293248103, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 1372 + }, + { + "epoch": 0.5485417499001198, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1373 + }, + { + "epoch": 0.5489412704754295, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1374 + }, + { + "epoch": 0.5493407910507391, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1375 + }, + { + "epoch": 0.5497403116260487, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1376 + }, + { + "epoch": 0.5501398322013584, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1377 + }, + { + "epoch": 0.550539352776668, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1378 + }, + { + "epoch": 0.5509388733519777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1379 + }, + { + "epoch": 0.5513383939272872, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1380 + }, + { + "epoch": 0.5517379145025969, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1381 + }, + { + "epoch": 0.5521374350779065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1382 + }, + { + "epoch": 0.5525369556532161, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1383 + }, + { + "epoch": 0.5529364762285258, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1384 + }, + { + "epoch": 0.5533359968038354, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1385 + }, + { + "epoch": 0.5537355173791451, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1386 + }, + { + "epoch": 0.5541350379544546, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1387 + }, + { + "epoch": 0.5545345585297643, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1388 + }, + { + "epoch": 0.5549340791050739, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1389 + }, + { + "epoch": 0.5553335996803835, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1390 + }, + { + "epoch": 0.5557331202556932, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1391 + }, + { + "epoch": 0.5561326408310028, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1392 + }, + { + "epoch": 0.5565321614063125, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1393 + }, + { + "epoch": 0.556931681981622, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 1394 + }, + { + "epoch": 0.5573312025569317, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1395 + }, + { + "epoch": 0.5577307231322413, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1396 + }, + { + "epoch": 0.5581302437075509, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1397 + }, + { + "epoch": 0.5585297642828606, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1398 + }, + { + "epoch": 0.5589292848581702, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1399 + }, + { + "epoch": 0.5593288054334798, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1400 + }, + { + "epoch": 0.5597283260087894, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1401 + }, + { + "epoch": 0.5601278465840991, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1402 + }, + { + "epoch": 0.5605273671594088, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1403 + }, + { + "epoch": 0.5609268877347183, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1404 + }, + { + "epoch": 0.561326408310028, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1405 + }, + { + "epoch": 0.5617259288853376, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1406 + }, + { + "epoch": 0.5621254494606472, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 1407 + }, + { + "epoch": 0.5625249700359568, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1408 + }, + { + "epoch": 0.5629244906112665, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1409 + }, + { + "epoch": 0.5633240111865762, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1410 + }, + { + "epoch": 0.5637235317618857, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1411 + }, + { + "epoch": 0.5641230523371954, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1412 + }, + { + "epoch": 0.564522572912505, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1413 + }, + { + "epoch": 0.5649220934878146, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1414 + }, + { + "epoch": 0.5653216140631242, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1415 + }, + { + "epoch": 0.5657211346384339, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1416 + }, + { + "epoch": 0.5661206552137436, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1417 + }, + { + "epoch": 0.5665201757890531, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1418 + }, + { + "epoch": 0.5669196963643628, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1419 + }, + { + "epoch": 0.5673192169396724, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1420 + }, + { + "epoch": 0.567718737514982, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1421 + }, + { + "epoch": 0.5681182580902916, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1422 + }, + { + "epoch": 0.5685177786656013, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1423 + }, + { + "epoch": 0.568917299240911, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1424 + }, + { + "epoch": 0.5693168198162205, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1425 + }, + { + "epoch": 0.5697163403915302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1426 + }, + { + "epoch": 0.5701158609668397, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1427 + }, + { + "epoch": 0.5705153815421494, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1428 + }, + { + "epoch": 0.5709149021174591, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1429 + }, + { + "epoch": 0.5713144226927687, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1430 + }, + { + "epoch": 0.5717139432680783, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1431 + }, + { + "epoch": 0.5721134638433879, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1432 + }, + { + "epoch": 0.5725129844186976, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1433 + }, + { + "epoch": 0.5729125049940071, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1434 + }, + { + "epoch": 0.5733120255693168, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1435 + }, + { + "epoch": 0.5737115461446265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1436 + }, + { + "epoch": 0.5741110667199361, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1437 + }, + { + "epoch": 0.5745105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1438 + }, + { + "epoch": 0.5749101078705553, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1439 + }, + { + "epoch": 0.575309628445865, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1440 + }, + { + "epoch": 0.5757091490211745, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1441 + }, + { + "epoch": 0.5761086695964842, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1442 + }, + { + "epoch": 0.5765081901717939, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1443 + }, + { + "epoch": 0.5769077107471035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1444 + }, + { + "epoch": 0.5773072313224131, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1445 + }, + { + "epoch": 0.5777067518977227, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1446 + }, + { + "epoch": 0.5781062724730324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1447 + }, + { + "epoch": 0.578505793048342, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1448 + }, + { + "epoch": 0.5789053136236516, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1449 + }, + { + "epoch": 0.5793048341989613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 1450 + }, + { + "epoch": 0.5797043547742708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1451 + }, + { + "epoch": 0.5801038753495805, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1452 + }, + { + "epoch": 0.5805033959248901, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1453 + }, + { + "epoch": 0.5809029165001998, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1454 + }, + { + "epoch": 0.5813024370755094, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1455 + }, + { + "epoch": 0.581701957650819, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1456 + }, + { + "epoch": 0.5821014782261287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1457 + }, + { + "epoch": 0.5825009988014382, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1458 + }, + { + "epoch": 0.5829005193767479, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1459 + }, + { + "epoch": 0.5833000399520575, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1460 + }, + { + "epoch": 0.5836995605273672, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1461 + }, + { + "epoch": 0.5840990811026768, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1462 + }, + { + "epoch": 0.5844986016779864, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1463 + }, + { + "epoch": 0.5848981222532961, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1464 + }, + { + "epoch": 0.5852976428286056, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1465 + }, + { + "epoch": 0.5856971634039153, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1466 + }, + { + "epoch": 0.5860966839792249, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1467 + }, + { + "epoch": 0.5864962045545346, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1468 + }, + { + "epoch": 0.5868957251298442, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1469 + }, + { + "epoch": 0.5872952457051538, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1470 + }, + { + "epoch": 0.5876947662804635, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1471 + }, + { + "epoch": 0.588094286855773, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1472 + }, + { + "epoch": 0.5884938074310827, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1473 + }, + { + "epoch": 0.5888933280063924, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1474 + }, + { + "epoch": 0.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1475 + }, + { + "epoch": 0.5896923691570116, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1476 + }, + { + "epoch": 0.5900918897323212, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1477 + }, + { + "epoch": 0.5904914103076309, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1478 + }, + { + "epoch": 0.5908909308829404, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1479 + }, + { + "epoch": 0.5912904514582501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1480 + }, + { + "epoch": 0.5916899720335598, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1481 + }, + { + "epoch": 0.5920894926088693, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1482 + }, + { + "epoch": 0.592489013184179, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1483 + }, + { + "epoch": 0.5928885337594886, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1484 + }, + { + "epoch": 0.5932880543347983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1485 + }, + { + "epoch": 0.5936875749101078, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1486 + }, + { + "epoch": 0.5940870954854175, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1487 + }, + { + "epoch": 0.5944866160607272, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1488 + }, + { + "epoch": 0.5948861366360367, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1489 + }, + { + "epoch": 0.5952856572113464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1490 + }, + { + "epoch": 0.595685177786656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1491 + }, + { + "epoch": 0.5960846983619656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1492 + }, + { + "epoch": 0.5964842189372753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1493 + }, + { + "epoch": 0.5968837395125849, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1494 + }, + { + "epoch": 0.5972832600878946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1495 + }, + { + "epoch": 0.5976827806632041, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1496 + }, + { + "epoch": 0.5980823012385138, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1497 + }, + { + "epoch": 0.5984818218138234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 1498 + }, + { + "epoch": 0.598881342389133, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1499 + }, + { + "epoch": 0.5992808629644427, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1500 + }, + { + "epoch": 0.5996803835397523, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1501 + }, + { + "epoch": 0.600079904115062, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1502 + }, + { + "epoch": 0.6004794246903715, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1503 + }, + { + "epoch": 0.6008789452656812, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1504 + }, + { + "epoch": 0.6012784658409908, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1505 + }, + { + "epoch": 0.6016779864163004, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1506 + }, + { + "epoch": 0.6020775069916101, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1507 + }, + { + "epoch": 0.6024770275669197, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1508 + }, + { + "epoch": 0.6028765481422294, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1509 + }, + { + "epoch": 0.6032760687175389, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1510 + }, + { + "epoch": 0.6036755892928486, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1511 + }, + { + "epoch": 0.6040751098681582, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1512 + }, + { + "epoch": 0.6044746304434678, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3669, + "step": 1513 + }, + { + "epoch": 0.6048741510187775, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1514 + }, + { + "epoch": 0.6052736715940871, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1515 + }, + { + "epoch": 0.6056731921693967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1516 + }, + { + "epoch": 0.6060727127447063, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1517 + }, + { + "epoch": 0.606472233320016, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1518 + }, + { + "epoch": 0.6068717538953257, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1519 + }, + { + "epoch": 0.6072712744706352, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1520 + }, + { + "epoch": 0.6076707950459449, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1521 + }, + { + "epoch": 0.6080703156212545, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1522 + }, + { + "epoch": 0.6084698361965641, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1523 + }, + { + "epoch": 0.6088693567718737, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1524 + }, + { + "epoch": 0.6092688773471834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1525 + }, + { + "epoch": 0.609668397922493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1526 + }, + { + "epoch": 0.6100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1527 + }, + { + "epoch": 0.6104674390731123, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1528 + }, + { + "epoch": 0.6108669596484219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1529 + }, + { + "epoch": 0.6112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1530 + }, + { + "epoch": 0.6116660007990411, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1531 + }, + { + "epoch": 0.6120655213743508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1532 + }, + { + "epoch": 0.6124650419496604, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1533 + }, + { + "epoch": 0.61286456252497, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1534 + }, + { + "epoch": 0.6132640831002797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1535 + }, + { + "epoch": 0.6136636036755893, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1536 + }, + { + "epoch": 0.6140631242508989, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1537 + }, + { + "epoch": 0.6144626448262086, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1538 + }, + { + "epoch": 0.6148621654015182, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1539 + }, + { + "epoch": 0.6152616859768278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1540 + }, + { + "epoch": 0.6156612065521374, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1541 + }, + { + "epoch": 0.6160607271274471, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1542 + }, + { + "epoch": 0.6164602477027566, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1543 + }, + { + "epoch": 0.6168597682780663, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1544 + }, + { + "epoch": 0.617259288853376, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1545 + }, + { + "epoch": 0.6176588094286856, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1546 + }, + { + "epoch": 0.6180583300039952, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1547 + }, + { + "epoch": 0.6184578505793048, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1548 + }, + { + "epoch": 0.6188573711546145, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1549 + }, + { + "epoch": 0.619256891729924, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1550 + }, + { + "epoch": 0.6196564123052337, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1551 + }, + { + "epoch": 0.6200559328805434, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1552 + }, + { + "epoch": 0.620455453455853, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1553 + }, + { + "epoch": 0.6208549740311626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 1554 + }, + { + "epoch": 0.6212544946064722, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1555 + }, + { + "epoch": 0.6216540151817819, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1556 + }, + { + "epoch": 0.6220535357570914, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1557 + }, + { + "epoch": 0.6224530563324011, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1558 + }, + { + "epoch": 0.6228525769077108, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1559 + }, + { + "epoch": 0.6232520974830204, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1560 + }, + { + "epoch": 0.62365161805833, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1561 + }, + { + "epoch": 0.6240511386336396, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1562 + }, + { + "epoch": 0.6244506592089493, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1563 + }, + { + "epoch": 0.6248501797842589, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1564 + }, + { + "epoch": 0.6252497003595685, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1565 + }, + { + "epoch": 0.6256492209348782, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1566 + }, + { + "epoch": 0.6260487415101877, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1567 + }, + { + "epoch": 0.6264482620854974, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1568 + }, + { + "epoch": 0.626847782660807, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1569 + }, + { + "epoch": 0.6272473032361167, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1570 + }, + { + "epoch": 0.6276468238114263, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1571 + }, + { + "epoch": 0.6280463443867359, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1572 + }, + { + "epoch": 0.6284458649620456, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1573 + }, + { + "epoch": 0.6288453855373551, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1574 + }, + { + "epoch": 0.6292449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1575 + }, + { + "epoch": 0.6296444266879744, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1576 + }, + { + "epoch": 0.630043947263284, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1577 + }, + { + "epoch": 0.6304434678385937, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1578 + }, + { + "epoch": 0.6308429884139033, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1579 + }, + { + "epoch": 0.631242508989213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1580 + }, + { + "epoch": 0.6316420295645225, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1581 + }, + { + "epoch": 0.6320415501398322, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1582 + }, + { + "epoch": 0.6324410707151419, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1583 + }, + { + "epoch": 0.6328405912904514, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1584 + }, + { + "epoch": 0.6332401118657611, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1585 + }, + { + "epoch": 0.6336396324410707, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1586 + }, + { + "epoch": 0.6340391530163804, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1587 + }, + { + "epoch": 0.6344386735916899, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1588 + }, + { + "epoch": 0.6348381941669996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1589 + }, + { + "epoch": 0.6352377147423093, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1590 + }, + { + "epoch": 0.6356372353176188, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1591 + }, + { + "epoch": 0.6360367558929285, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1592 + }, + { + "epoch": 0.6364362764682381, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1593 + }, + { + "epoch": 0.6368357970435478, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1594 + }, + { + "epoch": 0.6372353176188573, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1595 + }, + { + "epoch": 0.637634838194167, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1596 + }, + { + "epoch": 0.6380343587694767, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1597 + }, + { + "epoch": 0.6384338793447862, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1598 + }, + { + "epoch": 0.6388333999200959, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1599 + }, + { + "epoch": 0.6392329204954055, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1600 + }, + { + "epoch": 0.6396324410707152, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1601 + }, + { + "epoch": 0.6400319616460247, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1602 + }, + { + "epoch": 0.6404314822213344, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1603 + }, + { + "epoch": 0.6408310027966441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1604 + }, + { + "epoch": 0.6412305233719536, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1605 + }, + { + "epoch": 0.6416300439472633, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1606 + }, + { + "epoch": 0.6420295645225729, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1607 + }, + { + "epoch": 0.6424290850978825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1608 + }, + { + "epoch": 0.6428286056731922, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1609 + }, + { + "epoch": 0.6432281262485018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1610 + }, + { + "epoch": 0.6436276468238115, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1611 + }, + { + "epoch": 0.644027167399121, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1612 + }, + { + "epoch": 0.6444266879744307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1613 + }, + { + "epoch": 0.6448262085497403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1614 + }, + { + "epoch": 0.6452257291250499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1615 + }, + { + "epoch": 0.6456252497003596, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 1616 + }, + { + "epoch": 0.6460247702756692, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1617 + }, + { + "epoch": 0.6464242908509789, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 1618 + }, + { + "epoch": 0.6468238114262884, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1619 + }, + { + "epoch": 0.6472233320015981, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1620 + }, + { + "epoch": 0.6476228525769077, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1621 + }, + { + "epoch": 0.6480223731522173, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1622 + }, + { + "epoch": 0.648421893727527, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1623 + }, + { + "epoch": 0.6488214143028366, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1624 + }, + { + "epoch": 0.6492209348781462, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1625 + }, + { + "epoch": 0.6496204554534558, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1626 + }, + { + "epoch": 0.6500199760287655, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1627 + }, + { + "epoch": 0.6504194966040752, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1628 + }, + { + "epoch": 0.6508190171793847, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1629 + }, + { + "epoch": 0.6512185377546944, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1630 + }, + { + "epoch": 0.651618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1631 + }, + { + "epoch": 0.6520175789053136, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1632 + }, + { + "epoch": 0.6524170994806232, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1633 + }, + { + "epoch": 0.6528166200559329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1634 + }, + { + "epoch": 0.6532161406312426, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1635 + }, + { + "epoch": 0.6536156612065521, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 1636 + }, + { + "epoch": 0.6540151817818618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1637 + }, + { + "epoch": 0.6544147023571714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1638 + }, + { + "epoch": 0.654814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1639 + }, + { + "epoch": 0.6552137435077906, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1640 + }, + { + "epoch": 0.6556132640831003, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1641 + }, + { + "epoch": 0.65601278465841, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1642 + }, + { + "epoch": 0.6564123052337195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1643 + }, + { + "epoch": 0.6568118258090292, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1644 + }, + { + "epoch": 0.6572113463843388, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1645 + }, + { + "epoch": 0.6576108669596484, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1646 + }, + { + "epoch": 0.658010387534958, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1647 + }, + { + "epoch": 0.6584099081102677, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1648 + }, + { + "epoch": 0.6588094286855773, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1649 + }, + { + "epoch": 0.6592089492608869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 1650 + }, + { + "epoch": 0.6596084698361966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 1651 + }, + { + "epoch": 0.6600079904115062, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1652 + }, + { + "epoch": 0.6604075109868158, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1653 + }, + { + "epoch": 0.6608070315621255, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1654 + }, + { + "epoch": 0.6612065521374351, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1655 + }, + { + "epoch": 0.6616060727127447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1656 + }, + { + "epoch": 0.6620055932880543, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1657 + }, + { + "epoch": 0.662405113863364, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1658 + }, + { + "epoch": 0.6628046344386735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1659 + }, + { + "epoch": 0.6632041550139832, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1660 + }, + { + "epoch": 0.6636036755892929, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1661 + }, + { + "epoch": 0.6640031961646025, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1662 + }, + { + "epoch": 0.6644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1663 + }, + { + "epoch": 0.6648022373152217, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1664 + }, + { + "epoch": 0.6652017578905314, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1665 + }, + { + "epoch": 0.6656012784658409, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1666 + }, + { + "epoch": 0.6660007990411506, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1667 + }, + { + "epoch": 0.6664003196164603, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1668 + }, + { + "epoch": 0.6667998401917699, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1669 + }, + { + "epoch": 0.6671993607670795, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1670 + }, + { + "epoch": 0.6675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1671 + }, + { + "epoch": 0.6679984019176988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1672 + }, + { + "epoch": 0.6683979224930084, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1673 + }, + { + "epoch": 0.668797443068318, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1674 + }, + { + "epoch": 0.6691969636436277, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1675 + }, + { + "epoch": 0.6695964842189372, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1676 + }, + { + "epoch": 0.6699960047942469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 1677 + }, + { + "epoch": 0.6703955253695565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1678 + }, + { + "epoch": 0.6707950459448662, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1679 + }, + { + "epoch": 0.6711945665201758, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1680 + }, + { + "epoch": 0.6715940870954854, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1681 + }, + { + "epoch": 0.6719936076707951, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1682 + }, + { + "epoch": 0.6723931282461046, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1683 + }, + { + "epoch": 0.6727926488214143, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1684 + }, + { + "epoch": 0.6731921693967239, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1685 + }, + { + "epoch": 0.6735916899720336, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1686 + }, + { + "epoch": 0.6739912105473432, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1687 + }, + { + "epoch": 0.6743907311226528, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1688 + }, + { + "epoch": 0.6747902516979625, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1689 + }, + { + "epoch": 0.675189772273272, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1690 + }, + { + "epoch": 0.6755892928485817, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1691 + }, + { + "epoch": 0.6759888134238913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1692 + }, + { + "epoch": 0.676388333999201, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1693 + }, + { + "epoch": 0.6767878545745106, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1694 + }, + { + "epoch": 0.6771873751498202, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1695 + }, + { + "epoch": 0.6775868957251299, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1696 + }, + { + "epoch": 0.6779864163004394, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1697 + }, + { + "epoch": 0.6783859368757491, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1698 + }, + { + "epoch": 0.6787854574510588, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1699 + }, + { + "epoch": 0.6791849780263683, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 1700 + }, + { + "epoch": 0.679584498601678, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1701 + }, + { + "epoch": 0.6799840191769876, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1702 + }, + { + "epoch": 0.6803835397522973, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1703 + }, + { + "epoch": 0.6807830603276068, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1704 + }, + { + "epoch": 0.6811825809029165, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1705 + }, + { + "epoch": 0.6815821014782262, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1706 + }, + { + "epoch": 0.6819816220535357, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1707 + }, + { + "epoch": 0.6823811426288454, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1708 + }, + { + "epoch": 0.682780663204155, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1709 + }, + { + "epoch": 0.6831801837794647, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1710 + }, + { + "epoch": 0.6835797043547742, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1711 + }, + { + "epoch": 0.6839792249300839, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1712 + }, + { + "epoch": 0.6843787455053936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1713 + }, + { + "epoch": 0.6847782660807031, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1714 + }, + { + "epoch": 0.6851777866560128, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1715 + }, + { + "epoch": 0.6855773072313224, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1716 + }, + { + "epoch": 0.685976827806632, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1717 + }, + { + "epoch": 0.6863763483819417, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1718 + }, + { + "epoch": 0.6867758689572513, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1719 + }, + { + "epoch": 0.687175389532561, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1720 + }, + { + "epoch": 0.6875749101078705, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1721 + }, + { + "epoch": 0.6879744306831802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1722 + }, + { + "epoch": 0.6883739512584898, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1723 + }, + { + "epoch": 0.6887734718337994, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1724 + }, + { + "epoch": 0.6891729924091091, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1725 + }, + { + "epoch": 0.6895725129844187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1726 + }, + { + "epoch": 0.6899720335597284, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1727 + }, + { + "epoch": 0.6903715541350379, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1728 + }, + { + "epoch": 0.6907710747103476, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1729 + }, + { + "epoch": 0.6911705952856572, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1730 + }, + { + "epoch": 0.6915701158609668, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1731 + }, + { + "epoch": 0.6919696364362765, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1732 + }, + { + "epoch": 0.6923691570115861, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1733 + }, + { + "epoch": 0.6927686775868958, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1734 + }, + { + "epoch": 0.6931681981622053, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1735 + }, + { + "epoch": 0.693567718737515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1736 + }, + { + "epoch": 0.6939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1737 + }, + { + "epoch": 0.6943667598881342, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1738 + }, + { + "epoch": 0.6947662804634439, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 1739 + }, + { + "epoch": 0.6951658010387535, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1740 + }, + { + "epoch": 0.6955653216140631, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1741 + }, + { + "epoch": 0.6959648421893727, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 1742 + }, + { + "epoch": 0.6963643627646824, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1743 + }, + { + "epoch": 0.6967638833399921, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1744 + }, + { + "epoch": 0.6971634039153016, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1745 + }, + { + "epoch": 0.6975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1746 + }, + { + "epoch": 0.6979624450659209, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1747 + }, + { + "epoch": 0.6983619656412305, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1748 + }, + { + "epoch": 0.6987614862165401, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1749 + }, + { + "epoch": 0.6991610067918498, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1750 + }, + { + "epoch": 0.6995605273671595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 1751 + }, + { + "epoch": 0.699960047942469, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1752 + }, + { + "epoch": 0.7003595685177787, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1753 + }, + { + "epoch": 0.7007590890930883, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1754 + }, + { + "epoch": 0.7011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1755 + }, + { + "epoch": 0.7015581302437075, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1756 + }, + { + "epoch": 0.7019576508190172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1757 + }, + { + "epoch": 0.7023571713943269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1758 + }, + { + "epoch": 0.7027566919696364, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1759 + }, + { + "epoch": 0.7031562125449461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1760 + }, + { + "epoch": 0.7035557331202557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1761 + }, + { + "epoch": 0.7039552536955653, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1762 + }, + { + "epoch": 0.7043547742708749, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1763 + }, + { + "epoch": 0.7047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1764 + }, + { + "epoch": 0.7051538154214942, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1765 + }, + { + "epoch": 0.7055533359968038, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1766 + }, + { + "epoch": 0.7059528565721135, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1767 + }, + { + "epoch": 0.706352377147423, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1768 + }, + { + "epoch": 0.7067518977227327, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1769 + }, + { + "epoch": 0.7071514182980424, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 1770 + }, + { + "epoch": 0.707550938873352, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1771 + }, + { + "epoch": 0.7079504594486616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1772 + }, + { + "epoch": 0.7083499800239712, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1773 + }, + { + "epoch": 0.7087495005992809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1774 + }, + { + "epoch": 0.7091490211745904, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1775 + }, + { + "epoch": 0.7095485417499001, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 1776 + }, + { + "epoch": 0.7099480623252098, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1777 + }, + { + "epoch": 0.7103475829005194, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1778 + }, + { + "epoch": 0.710747103475829, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1779 + }, + { + "epoch": 0.7111466240511386, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1780 + }, + { + "epoch": 0.7115461446264483, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1781 + }, + { + "epoch": 0.7119456652017578, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1782 + }, + { + "epoch": 0.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1783 + }, + { + "epoch": 0.7127447063523772, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1784 + }, + { + "epoch": 0.7131442269276868, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1785 + }, + { + "epoch": 0.7135437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1786 + }, + { + "epoch": 0.713943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1787 + }, + { + "epoch": 0.7143427886536157, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1788 + }, + { + "epoch": 0.7147423092289253, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1789 + }, + { + "epoch": 0.7151418298042349, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1790 + }, + { + "epoch": 0.7155413503795446, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1791 + }, + { + "epoch": 0.7159408709548541, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1792 + }, + { + "epoch": 0.7163403915301638, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1793 + }, + { + "epoch": 0.7167399121054734, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1794 + }, + { + "epoch": 0.7171394326807831, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1795 + }, + { + "epoch": 0.7175389532560927, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1796 + }, + { + "epoch": 0.7179384738314023, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1797 + }, + { + "epoch": 0.718337994406712, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1798 + }, + { + "epoch": 0.7187375149820215, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1799 + }, + { + "epoch": 0.7191370355573312, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 1800 + }, + { + "epoch": 0.7195365561326408, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1801 + }, + { + "epoch": 0.7199360767079505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1802 + }, + { + "epoch": 0.7203355972832601, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 1803 + }, + { + "epoch": 0.7207351178585697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1804 + }, + { + "epoch": 0.7211346384338794, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1805 + }, + { + "epoch": 0.7215341590091889, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1806 + }, + { + "epoch": 0.7219336795844986, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1807 + }, + { + "epoch": 0.7223332001598082, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1808 + }, + { + "epoch": 0.7227327207351179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1809 + }, + { + "epoch": 0.7231322413104275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1810 + }, + { + "epoch": 0.7235317618857371, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1811 + }, + { + "epoch": 0.7239312824610468, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1812 + }, + { + "epoch": 0.7243308030363563, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1813 + }, + { + "epoch": 0.724730323611666, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1814 + }, + { + "epoch": 0.7251298441869757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1815 + }, + { + "epoch": 0.7255293647622852, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1816 + }, + { + "epoch": 0.7259288853375949, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1817 + }, + { + "epoch": 0.7263284059129045, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1818 + }, + { + "epoch": 0.7267279264882142, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1819 + }, + { + "epoch": 0.7271274470635237, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1820 + }, + { + "epoch": 0.7275269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1821 + }, + { + "epoch": 0.7279264882141431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1822 + }, + { + "epoch": 0.7283260087894526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1823 + }, + { + "epoch": 0.7287255293647623, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1824 + }, + { + "epoch": 0.7291250499400719, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 1825 + }, + { + "epoch": 0.7295245705153816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1826 + }, + { + "epoch": 0.7299240910906911, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1827 + }, + { + "epoch": 0.7303236116660008, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1828 + }, + { + "epoch": 0.7307231322413105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 1829 + }, + { + "epoch": 0.73112265281662, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1830 + }, + { + "epoch": 0.7315221733919297, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1831 + }, + { + "epoch": 0.7319216939672393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1832 + }, + { + "epoch": 0.732321214542549, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1833 + }, + { + "epoch": 0.7327207351178586, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1834 + }, + { + "epoch": 0.7331202556931682, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1835 + }, + { + "epoch": 0.7335197762684779, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1836 + }, + { + "epoch": 0.7339192968437874, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1837 + }, + { + "epoch": 0.7343188174190971, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1838 + }, + { + "epoch": 0.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1839 + }, + { + "epoch": 0.7351178585697163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1840 + }, + { + "epoch": 0.735517379145026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1841 + }, + { + "epoch": 0.7359168997203356, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1842 + }, + { + "epoch": 0.7363164202956453, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 1843 + }, + { + "epoch": 0.7367159408709548, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1844 + }, + { + "epoch": 0.7371154614462645, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1845 + }, + { + "epoch": 0.7375149820215741, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1846 + }, + { + "epoch": 0.7379145025968837, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1847 + }, + { + "epoch": 0.7383140231721934, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1848 + }, + { + "epoch": 0.738713543747503, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 1849 + }, + { + "epoch": 0.7391130643228127, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1850 + }, + { + "epoch": 0.7395125848981222, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1851 + }, + { + "epoch": 0.7399121054734319, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1852 + }, + { + "epoch": 0.7403116260487415, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1853 + }, + { + "epoch": 0.7407111466240511, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 1854 + }, + { + "epoch": 0.7411106671993608, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1855 + }, + { + "epoch": 0.7415101877746704, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1856 + }, + { + "epoch": 0.74190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1857 + }, + { + "epoch": 0.7423092289252896, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1858 + }, + { + "epoch": 0.7427087495005993, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1859 + }, + { + "epoch": 0.743108270075909, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1860 + }, + { + "epoch": 0.7435077906512185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1861 + }, + { + "epoch": 0.7439073112265282, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1862 + }, + { + "epoch": 0.7443068318018378, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 1863 + }, + { + "epoch": 0.7447063523771474, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1864 + }, + { + "epoch": 0.745105872952457, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1865 + }, + { + "epoch": 0.7455053935277667, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1866 + }, + { + "epoch": 0.7459049141030764, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1867 + }, + { + "epoch": 0.7463044346783859, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1868 + }, + { + "epoch": 0.7467039552536956, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1869 + }, + { + "epoch": 0.7471034758290052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 1870 + }, + { + "epoch": 0.7475029964043148, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1871 + }, + { + "epoch": 0.7479025169796244, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1872 + }, + { + "epoch": 0.7483020375549341, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1873 + }, + { + "epoch": 0.7487015581302437, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1874 + }, + { + "epoch": 0.7491010787055533, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1875 + }, + { + "epoch": 0.749500599280863, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1876 + }, + { + "epoch": 0.7499001198561726, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1877 + }, + { + "epoch": 0.7502996404314822, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1878 + }, + { + "epoch": 0.7506991610067919, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1879 + }, + { + "epoch": 0.7510986815821015, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1880 + }, + { + "epoch": 0.7514982021574111, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 1881 + }, + { + "epoch": 0.7518977227327207, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1882 + }, + { + "epoch": 0.7522972433080304, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1883 + }, + { + "epoch": 0.75269676388334, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1884 + }, + { + "epoch": 0.7530962844586496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1885 + }, + { + "epoch": 0.7534958050339593, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1886 + }, + { + "epoch": 0.7538953256092689, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1887 + }, + { + "epoch": 0.7542948461845785, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1888 + }, + { + "epoch": 0.7546943667598881, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1889 + }, + { + "epoch": 0.7550938873351978, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 1890 + }, + { + "epoch": 0.7554934079105073, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1891 + }, + { + "epoch": 0.755892928485817, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1892 + }, + { + "epoch": 0.7562924490611267, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1893 + }, + { + "epoch": 0.7566919696364363, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1894 + }, + { + "epoch": 0.7570914902117459, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1895 + }, + { + "epoch": 0.7574910107870555, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1896 + }, + { + "epoch": 0.7578905313623652, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1897 + }, + { + "epoch": 0.7582900519376747, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 1898 + }, + { + "epoch": 0.7586895725129844, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1899 + }, + { + "epoch": 0.7590890930882941, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1900 + }, + { + "epoch": 0.7594886136636037, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1901 + }, + { + "epoch": 0.7598881342389133, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1902 + }, + { + "epoch": 0.7602876548142229, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1903 + }, + { + "epoch": 0.7606871753895326, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1904 + }, + { + "epoch": 0.7610866959648422, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 1905 + }, + { + "epoch": 0.7614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1906 + }, + { + "epoch": 0.7618857371154615, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1907 + }, + { + "epoch": 0.762285257690771, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1908 + }, + { + "epoch": 0.7626847782660807, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1909 + }, + { + "epoch": 0.7630842988413903, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1910 + }, + { + "epoch": 0.7634838194167, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1911 + }, + { + "epoch": 0.7638833399920096, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1912 + }, + { + "epoch": 0.7642828605673192, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1913 + }, + { + "epoch": 0.7646823811426289, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1914 + }, + { + "epoch": 0.7650819017179384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1915 + }, + { + "epoch": 0.7654814222932481, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1916 + }, + { + "epoch": 0.7658809428685577, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1917 + }, + { + "epoch": 0.7662804634438674, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1918 + }, + { + "epoch": 0.766679984019177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1919 + }, + { + "epoch": 0.7670795045944866, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1920 + }, + { + "epoch": 0.7674790251697963, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1921 + }, + { + "epoch": 0.7678785457451058, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1922 + }, + { + "epoch": 0.7682780663204155, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1923 + }, + { + "epoch": 0.7686775868957252, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1924 + }, + { + "epoch": 0.7690771074710347, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1925 + }, + { + "epoch": 0.7694766280463444, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 1926 + }, + { + "epoch": 0.769876148621654, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1927 + }, + { + "epoch": 0.7702756691969637, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1928 + }, + { + "epoch": 0.7706751897722732, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1929 + }, + { + "epoch": 0.7710747103475829, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1930 + }, + { + "epoch": 0.7714742309228926, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1931 + }, + { + "epoch": 0.7718737514982021, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1932 + }, + { + "epoch": 0.7722732720735118, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1933 + }, + { + "epoch": 0.7726727926488214, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1934 + }, + { + "epoch": 0.7730723132241311, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1935 + }, + { + "epoch": 0.7734718337994406, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1936 + }, + { + "epoch": 0.7738713543747503, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1937 + }, + { + "epoch": 0.77427087495006, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1938 + }, + { + "epoch": 0.7746703955253695, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 1939 + }, + { + "epoch": 0.7750699161006792, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1940 + }, + { + "epoch": 0.7754694366759888, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1941 + }, + { + "epoch": 0.7758689572512985, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 1942 + }, + { + "epoch": 0.776268477826608, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 1943 + }, + { + "epoch": 0.7766679984019177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1944 + }, + { + "epoch": 0.7770675189772274, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 1945 + }, + { + "epoch": 0.7774670395525369, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1946 + }, + { + "epoch": 0.7778665601278466, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1947 + }, + { + "epoch": 0.7782660807031562, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1948 + }, + { + "epoch": 0.7786656012784658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1949 + }, + { + "epoch": 0.7790651218537755, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 1950 + }, + { + "epoch": 0.7794646424290851, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1951 + }, + { + "epoch": 0.7798641630043948, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1952 + }, + { + "epoch": 0.7802636835797043, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 1953 + }, + { + "epoch": 0.780663204155014, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1954 + }, + { + "epoch": 0.7810627247303236, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1955 + }, + { + "epoch": 0.7814622453056332, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1956 + }, + { + "epoch": 0.7818617658809429, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1957 + }, + { + "epoch": 0.7822612864562525, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1958 + }, + { + "epoch": 0.7826608070315622, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1959 + }, + { + "epoch": 0.7830603276068717, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1960 + }, + { + "epoch": 0.7834598481821814, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1961 + }, + { + "epoch": 0.783859368757491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1962 + }, + { + "epoch": 0.7842588893328006, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1963 + }, + { + "epoch": 0.7846584099081103, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1964 + }, + { + "epoch": 0.7850579304834199, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1965 + }, + { + "epoch": 0.7854574510587295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1966 + }, + { + "epoch": 0.7858569716340391, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 1967 + }, + { + "epoch": 0.7862564922093488, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1968 + }, + { + "epoch": 0.7866560127846585, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 1969 + }, + { + "epoch": 0.787055533359968, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1970 + }, + { + "epoch": 0.7874550539352777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 1971 + }, + { + "epoch": 0.7878545745105873, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1972 + }, + { + "epoch": 0.7882540950858969, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1973 + }, + { + "epoch": 0.7886536156612065, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1974 + }, + { + "epoch": 0.7890531362365162, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1975 + }, + { + "epoch": 0.7894526568118259, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1976 + }, + { + "epoch": 0.7898521773871354, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1977 + }, + { + "epoch": 0.7902516979624451, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1978 + }, + { + "epoch": 0.7906512185377547, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 1979 + }, + { + "epoch": 0.7910507391130643, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1980 + }, + { + "epoch": 0.7914502596883739, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1981 + }, + { + "epoch": 0.7918497802636836, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 1982 + }, + { + "epoch": 0.7922493008389933, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 1983 + }, + { + "epoch": 0.7926488214143028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1984 + }, + { + "epoch": 0.7930483419896125, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1985 + }, + { + "epoch": 0.7934478625649221, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1986 + }, + { + "epoch": 0.7938473831402317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1987 + }, + { + "epoch": 0.7942469037155413, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1988 + }, + { + "epoch": 0.794646424290851, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 1989 + }, + { + "epoch": 0.7950459448661606, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1990 + }, + { + "epoch": 0.7954454654414702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 1991 + }, + { + "epoch": 0.7958449860167799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1992 + }, + { + "epoch": 0.7962445065920895, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1993 + }, + { + "epoch": 0.7966440271673991, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1994 + }, + { + "epoch": 0.7970435477427088, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1995 + }, + { + "epoch": 0.7974430683180184, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 1996 + }, + { + "epoch": 0.797842588893328, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1997 + }, + { + "epoch": 0.7982421094686376, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1998 + }, + { + "epoch": 0.7986416300439473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1999 + }, + { + "epoch": 0.7990411506192568, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2000 + }, + { + "epoch": 0.7994406711945665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2001 + }, + { + "epoch": 0.7998401917698762, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2002 + }, + { + "epoch": 0.8002397123451858, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2003 + }, + { + "epoch": 0.8006392329204954, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2004 + }, + { + "epoch": 0.801038753495805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 2005 + }, + { + "epoch": 0.8014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2006 + }, + { + "epoch": 0.8018377946464242, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2007 + }, + { + "epoch": 0.8022373152217339, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2008 + }, + { + "epoch": 0.8026368357970436, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2009 + }, + { + "epoch": 0.8030363563723532, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 2010 + }, + { + "epoch": 0.8034358769476628, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2011 + }, + { + "epoch": 0.8038353975229724, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2012 + }, + { + "epoch": 0.8042349180982821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2013 + }, + { + "epoch": 0.8046344386735917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 2014 + }, + { + "epoch": 0.8050339592489013, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 2015 + }, + { + "epoch": 0.805433479824211, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2016 + }, + { + "epoch": 0.8058330003995205, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2017 + }, + { + "epoch": 0.8062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2018 + }, + { + "epoch": 0.8066320415501398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2019 + }, + { + "epoch": 0.8070315621254495, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2020 + }, + { + "epoch": 0.8074310827007591, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2021 + }, + { + "epoch": 0.8078306032760687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2022 + }, + { + "epoch": 0.8082301238513784, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2023 + }, + { + "epoch": 0.8086296444266879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2024 + }, + { + "epoch": 0.8090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2025 + }, + { + "epoch": 0.8094286855773072, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 2026 + }, + { + "epoch": 0.8098282061526169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2027 + }, + { + "epoch": 0.8102277267279265, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2028 + }, + { + "epoch": 0.8106272473032361, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2029 + }, + { + "epoch": 0.8110267678785458, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2030 + }, + { + "epoch": 0.8114262884538553, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2031 + }, + { + "epoch": 0.811825809029165, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 2032 + }, + { + "epoch": 0.8122253296044746, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2033 + }, + { + "epoch": 0.8126248501797843, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2034 + }, + { + "epoch": 0.8130243707550939, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 2035 + }, + { + "epoch": 0.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2036 + }, + { + "epoch": 0.8138234119057132, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 2037 + }, + { + "epoch": 0.8142229324810227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2038 + }, + { + "epoch": 0.8146224530563324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2039 + }, + { + "epoch": 0.8150219736316421, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2040 + }, + { + "epoch": 0.8154214942069516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2041 + }, + { + "epoch": 0.8158210147822613, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2042 + }, + { + "epoch": 0.8162205353575709, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2043 + }, + { + "epoch": 0.8166200559328806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2044 + }, + { + "epoch": 0.8170195765081901, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2045 + }, + { + "epoch": 0.8174190970834998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2046 + }, + { + "epoch": 0.8178186176588095, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2047 + }, + { + "epoch": 0.818218138234119, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2048 + }, + { + "epoch": 0.8186176588094287, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 2049 + }, + { + "epoch": 0.8190171793847383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2050 + }, + { + "epoch": 0.819416699960048, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2051 + }, + { + "epoch": 0.8198162205353575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2052 + }, + { + "epoch": 0.8202157411106672, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2053 + }, + { + "epoch": 0.8206152616859769, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2054 + }, + { + "epoch": 0.8210147822612864, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2055 + }, + { + "epoch": 0.8214143028365961, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2056 + }, + { + "epoch": 0.8218138234119057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2057 + }, + { + "epoch": 0.8222133439872154, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2058 + }, + { + "epoch": 0.8226128645625249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2059 + }, + { + "epoch": 0.8230123851378346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2060 + }, + { + "epoch": 0.8234119057131443, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2061 + }, + { + "epoch": 0.8238114262884538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2062 + }, + { + "epoch": 0.8242109468637635, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2063 + }, + { + "epoch": 0.8246104674390731, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2064 + }, + { + "epoch": 0.8250099880143827, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2065 + }, + { + "epoch": 0.8254095085896924, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2066 + }, + { + "epoch": 0.825809029165002, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2067 + }, + { + "epoch": 0.8262085497403117, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2068 + }, + { + "epoch": 0.8266080703156212, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2069 + }, + { + "epoch": 0.8270075908909309, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2070 + }, + { + "epoch": 0.8274071114662405, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2071 + }, + { + "epoch": 0.8278066320415501, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2072 + }, + { + "epoch": 0.8282061526168598, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 2073 + }, + { + "epoch": 0.8286056731921694, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 2074 + }, + { + "epoch": 0.829005193767479, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2075 + }, + { + "epoch": 0.8294047143427886, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2076 + }, + { + "epoch": 0.8298042349180983, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2077 + }, + { + "epoch": 0.8302037554934079, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2078 + }, + { + "epoch": 0.8306032760687175, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2079 + }, + { + "epoch": 0.8310027966440272, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2080 + }, + { + "epoch": 0.8314023172193368, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2081 + }, + { + "epoch": 0.8318018377946464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2082 + }, + { + "epoch": 0.832201358369956, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2083 + }, + { + "epoch": 0.8326008789452657, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2084 + }, + { + "epoch": 0.8330003995205754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2085 + }, + { + "epoch": 0.8333999200958849, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2086 + }, + { + "epoch": 0.8337994406711946, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2087 + }, + { + "epoch": 0.8341989612465042, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2088 + }, + { + "epoch": 0.8345984818218138, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2089 + }, + { + "epoch": 0.8349980023971234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 2090 + }, + { + "epoch": 0.8353975229724331, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 2091 + }, + { + "epoch": 0.8357970435477428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2092 + }, + { + "epoch": 0.8361965641230523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2093 + }, + { + "epoch": 0.836596084698362, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2094 + }, + { + "epoch": 0.8369956052736716, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2095 + }, + { + "epoch": 0.8373951258489812, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2096 + }, + { + "epoch": 0.8377946464242908, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2097 + }, + { + "epoch": 0.8381941669996005, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2098 + }, + { + "epoch": 0.8385936875749102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2099 + }, + { + "epoch": 0.8389932081502197, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2100 + }, + { + "epoch": 0.8393927287255294, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 2101 + }, + { + "epoch": 0.839792249300839, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2102 + }, + { + "epoch": 0.8401917698761486, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2103 + }, + { + "epoch": 0.8405912904514582, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 2104 + }, + { + "epoch": 0.8409908110267679, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2105 + }, + { + "epoch": 0.8413903316020775, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2106 + }, + { + "epoch": 0.8417898521773871, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2107 + }, + { + "epoch": 0.8421893727526968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2108 + }, + { + "epoch": 0.8425888933280063, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2109 + }, + { + "epoch": 0.842988413903316, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2110 + }, + { + "epoch": 0.8433879344786257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2111 + }, + { + "epoch": 0.8437874550539353, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2112 + }, + { + "epoch": 0.8441869756292449, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2113 + }, + { + "epoch": 0.8445864962045545, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2114 + }, + { + "epoch": 0.8449860167798642, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2115 + }, + { + "epoch": 0.8453855373551737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2116 + }, + { + "epoch": 0.8457850579304834, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2117 + }, + { + "epoch": 0.8461845785057931, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2118 + }, + { + "epoch": 0.8465840990811027, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2119 + }, + { + "epoch": 0.8469836196564123, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2120 + }, + { + "epoch": 0.8473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2121 + }, + { + "epoch": 0.8477826608070316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2122 + }, + { + "epoch": 0.8481821813823411, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 2123 + }, + { + "epoch": 0.8485817019576508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2124 + }, + { + "epoch": 0.8489812225329605, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 2125 + }, + { + "epoch": 0.84938074310827, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2126 + }, + { + "epoch": 0.8497802636835797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2127 + }, + { + "epoch": 0.8501797842588893, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2128 + }, + { + "epoch": 0.850579304834199, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2129 + }, + { + "epoch": 0.8509788254095086, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2130 + }, + { + "epoch": 0.8513783459848182, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2131 + }, + { + "epoch": 0.8517778665601279, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2132 + }, + { + "epoch": 0.8521773871354374, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 0.8525769077107471, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2134 + }, + { + "epoch": 0.8529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2135 + }, + { + "epoch": 0.8533759488613664, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2136 + }, + { + "epoch": 0.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2137 + }, + { + "epoch": 0.8541749900119856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2138 + }, + { + "epoch": 0.8545745105872953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2139 + }, + { + "epoch": 0.8549740311626048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2140 + }, + { + "epoch": 0.8553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2141 + }, + { + "epoch": 0.8557730723132241, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2142 + }, + { + "epoch": 0.8561725928885338, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 2143 + }, + { + "epoch": 0.8565721134638434, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2144 + }, + { + "epoch": 0.856971634039153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2145 + }, + { + "epoch": 0.8573711546144627, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2146 + }, + { + "epoch": 0.8577706751897722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2147 + }, + { + "epoch": 0.8581701957650819, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 2148 + }, + { + "epoch": 0.8585697163403915, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2149 + }, + { + "epoch": 0.8589692369157012, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2150 + }, + { + "epoch": 0.8593687574910108, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2151 + }, + { + "epoch": 0.8597682780663204, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2152 + }, + { + "epoch": 0.8601677986416301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2153 + }, + { + "epoch": 0.8605673192169396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 2154 + }, + { + "epoch": 0.8609668397922493, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2155 + }, + { + "epoch": 0.861366360367559, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2156 + }, + { + "epoch": 0.8617658809428685, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2157 + }, + { + "epoch": 0.8621654015181782, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2158 + }, + { + "epoch": 0.8625649220934878, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2159 + }, + { + "epoch": 0.8629644426687975, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2160 + }, + { + "epoch": 0.863363963244107, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2161 + }, + { + "epoch": 0.8637634838194167, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2162 + }, + { + "epoch": 0.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2163 + }, + { + "epoch": 0.8645625249700359, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2164 + }, + { + "epoch": 0.8649620455453456, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2165 + }, + { + "epoch": 0.8653615661206552, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2166 + }, + { + "epoch": 0.8657610866959649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2167 + }, + { + "epoch": 0.8661606072712744, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2168 + }, + { + "epoch": 0.8665601278465841, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2169 + }, + { + "epoch": 0.8669596484218938, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2170 + }, + { + "epoch": 0.8673591689972033, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2171 + }, + { + "epoch": 0.867758689572513, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2172 + }, + { + "epoch": 0.8681582101478226, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2173 + }, + { + "epoch": 0.8685577307231322, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2174 + }, + { + "epoch": 0.8689572512984419, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2175 + }, + { + "epoch": 0.8693567718737515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2176 + }, + { + "epoch": 0.8697562924490612, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2177 + }, + { + "epoch": 0.8701558130243707, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2178 + }, + { + "epoch": 0.8705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 2179 + }, + { + "epoch": 0.87095485417499, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2180 + }, + { + "epoch": 0.8713543747502996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2181 + }, + { + "epoch": 0.8717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2182 + }, + { + "epoch": 0.8721534159009189, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2183 + }, + { + "epoch": 0.8725529364762286, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2184 + }, + { + "epoch": 0.8729524570515381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2185 + }, + { + "epoch": 0.8733519776268478, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2186 + }, + { + "epoch": 0.8737514982021574, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2187 + }, + { + "epoch": 0.874151018777467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2188 + }, + { + "epoch": 0.8745505393527767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2189 + }, + { + "epoch": 0.8749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2190 + }, + { + "epoch": 0.875349580503396, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2191 + }, + { + "epoch": 0.8757491010787055, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2192 + }, + { + "epoch": 0.8761486216540152, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2193 + }, + { + "epoch": 0.8765481422293248, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2194 + }, + { + "epoch": 0.8769476628046344, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2195 + }, + { + "epoch": 0.8773471833799441, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2196 + }, + { + "epoch": 0.8777467039552537, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2197 + }, + { + "epoch": 0.8781462245305633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2198 + }, + { + "epoch": 0.8785457451058729, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 2199 + }, + { + "epoch": 0.8789452656811826, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2200 + }, + { + "epoch": 0.8793447862564923, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2201 + }, + { + "epoch": 0.8797443068318018, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2202 + }, + { + "epoch": 0.8801438274071115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2203 + }, + { + "epoch": 0.8805433479824211, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2204 + }, + { + "epoch": 0.8809428685577307, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2205 + }, + { + "epoch": 0.8813423891330403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2206 + }, + { + "epoch": 0.88174190970835, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2207 + }, + { + "epoch": 0.8821414302836597, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2208 + }, + { + "epoch": 0.8825409508589692, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2209 + }, + { + "epoch": 0.8829404714342789, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2210 + }, + { + "epoch": 0.8833399920095885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2211 + }, + { + "epoch": 0.8837395125848981, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 2212 + }, + { + "epoch": 0.8841390331602077, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2213 + }, + { + "epoch": 0.8845385537355174, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2214 + }, + { + "epoch": 0.884938074310827, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2215 + }, + { + "epoch": 0.8853375948861366, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2216 + }, + { + "epoch": 0.8857371154614463, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2217 + }, + { + "epoch": 0.8861366360367559, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2218 + }, + { + "epoch": 0.8865361566120655, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2219 + }, + { + "epoch": 0.8869356771873752, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2220 + }, + { + "epoch": 0.8873351977626848, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2221 + }, + { + "epoch": 0.8877347183379944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2222 + }, + { + "epoch": 0.888134238913304, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2223 + }, + { + "epoch": 0.8885337594886137, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2224 + }, + { + "epoch": 0.8889332800639232, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2225 + }, + { + "epoch": 0.8893328006392329, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2226 + }, + { + "epoch": 0.8897323212145426, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2227 + }, + { + "epoch": 0.8901318417898522, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2228 + }, + { + "epoch": 0.8905313623651618, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2229 + }, + { + "epoch": 0.8909308829404714, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2230 + }, + { + "epoch": 0.8913304035157811, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2231 + }, + { + "epoch": 0.8917299240910906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2232 + }, + { + "epoch": 0.8921294446664003, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2233 + }, + { + "epoch": 0.89252896524171, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2234 + }, + { + "epoch": 0.8929284858170196, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2235 + }, + { + "epoch": 0.8933280063923292, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2236 + }, + { + "epoch": 0.8937275269676388, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2237 + }, + { + "epoch": 0.8941270475429485, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2238 + }, + { + "epoch": 0.894526568118258, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2239 + }, + { + "epoch": 0.8949260886935677, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2240 + }, + { + "epoch": 0.8953256092688774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2241 + }, + { + "epoch": 0.895725129844187, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2242 + }, + { + "epoch": 0.8961246504194966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2243 + }, + { + "epoch": 0.8965241709948062, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2244 + }, + { + "epoch": 0.8969236915701159, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2245 + }, + { + "epoch": 0.8973232121454255, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2246 + }, + { + "epoch": 0.8977227327207351, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2247 + }, + { + "epoch": 0.8981222532960448, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2248 + }, + { + "epoch": 0.8985217738713543, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2249 + }, + { + "epoch": 0.898921294446664, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2250 + }, + { + "epoch": 0.8993208150219736, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2251 + }, + { + "epoch": 0.8997203355972833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2252 + }, + { + "epoch": 0.9001198561725929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2253 + }, + { + "epoch": 0.9005193767479025, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2254 + }, + { + "epoch": 0.9009188973232122, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2255 + }, + { + "epoch": 0.9013184178985217, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2256 + }, + { + "epoch": 0.9017179384738314, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2257 + }, + { + "epoch": 0.902117459049141, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2258 + }, + { + "epoch": 0.9025169796244507, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2259 + }, + { + "epoch": 0.9029165001997603, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2260 + }, + { + "epoch": 0.9033160207750699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2261 + }, + { + "epoch": 0.9037155413503796, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2262 + }, + { + "epoch": 0.9041150619256891, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2263 + }, + { + "epoch": 0.9045145825009988, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2264 + }, + { + "epoch": 0.9049141030763085, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2265 + }, + { + "epoch": 0.905313623651618, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2266 + }, + { + "epoch": 0.9057131442269277, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2267 + }, + { + "epoch": 0.9061126648022373, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2268 + }, + { + "epoch": 0.906512185377547, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 2269 + }, + { + "epoch": 0.9069117059528565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2270 + }, + { + "epoch": 0.9073112265281662, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2271 + }, + { + "epoch": 0.9077107471034759, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2272 + }, + { + "epoch": 0.9081102676787854, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2273 + }, + { + "epoch": 0.9085097882540951, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2274 + }, + { + "epoch": 0.9089093088294047, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2275 + }, + { + "epoch": 0.9093088294047144, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2276 + }, + { + "epoch": 0.9097083499800239, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2277 + }, + { + "epoch": 0.9101078705553336, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2278 + }, + { + "epoch": 0.9105073911306433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2279 + }, + { + "epoch": 0.9109069117059528, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2280 + }, + { + "epoch": 0.9113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2281 + }, + { + "epoch": 0.9117059528565721, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2282 + }, + { + "epoch": 0.9121054734318818, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2283 + }, + { + "epoch": 0.9125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2284 + }, + { + "epoch": 0.912904514582501, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2285 + }, + { + "epoch": 0.9133040351578107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2286 + }, + { + "epoch": 0.9137035557331202, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2287 + }, + { + "epoch": 0.9141030763084299, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2288 + }, + { + "epoch": 0.9145025968837395, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2289 + }, + { + "epoch": 0.9149021174590491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2290 + }, + { + "epoch": 0.9153016380343588, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 2291 + }, + { + "epoch": 0.9157011586096684, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2292 + }, + { + "epoch": 0.9161006791849781, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2293 + }, + { + "epoch": 0.9165001997602876, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2294 + }, + { + "epoch": 0.9168997203355973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2295 + }, + { + "epoch": 0.9172992409109069, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2296 + }, + { + "epoch": 0.9176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2297 + }, + { + "epoch": 0.9180982820615262, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2298 + }, + { + "epoch": 0.9184978026368358, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2299 + }, + { + "epoch": 0.9188973232121455, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2300 + }, + { + "epoch": 0.919296843787455, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2301 + }, + { + "epoch": 0.9196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2302 + }, + { + "epoch": 0.9200958849380743, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 2303 + }, + { + "epoch": 0.9204954055133839, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2304 + }, + { + "epoch": 0.9208949260886936, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2305 + }, + { + "epoch": 0.9212944466640032, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2306 + }, + { + "epoch": 0.9216939672393129, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2307 + }, + { + "epoch": 0.9220934878146224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2308 + }, + { + "epoch": 0.9224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2309 + }, + { + "epoch": 0.9228925289652418, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2310 + }, + { + "epoch": 0.9232920495405513, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2311 + }, + { + "epoch": 0.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2312 + }, + { + "epoch": 0.9240910906911706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2313 + }, + { + "epoch": 0.9244906112664802, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2314 + }, + { + "epoch": 0.9248901318417898, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2315 + }, + { + "epoch": 0.9252896524170995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 2316 + }, + { + "epoch": 0.9256891729924092, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2317 + }, + { + "epoch": 0.9260886935677187, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2318 + }, + { + "epoch": 0.9264882141430284, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2319 + }, + { + "epoch": 0.926887734718338, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2320 + }, + { + "epoch": 0.9272872552936476, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2321 + }, + { + "epoch": 0.9276867758689572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2322 + }, + { + "epoch": 0.9280862964442669, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2323 + }, + { + "epoch": 0.9284858170195766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2324 + }, + { + "epoch": 0.9288853375948861, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2325 + }, + { + "epoch": 0.9292848581701958, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2326 + }, + { + "epoch": 0.9296843787455054, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2327 + }, + { + "epoch": 0.930083899320815, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2328 + }, + { + "epoch": 0.9304834198961246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 2329 + }, + { + "epoch": 0.9308829404714343, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2330 + }, + { + "epoch": 0.931282461046744, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2331 + }, + { + "epoch": 0.9316819816220535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 2332 + }, + { + "epoch": 0.9320815021973632, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2333 + }, + { + "epoch": 0.9324810227726728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2334 + }, + { + "epoch": 0.9328805433479824, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2335 + }, + { + "epoch": 0.9332800639232921, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2336 + }, + { + "epoch": 0.9336795844986017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2337 + }, + { + "epoch": 0.9340791050739113, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2338 + }, + { + "epoch": 0.9344786256492209, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2339 + }, + { + "epoch": 0.9348781462245306, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2340 + }, + { + "epoch": 0.9352776667998401, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2341 + }, + { + "epoch": 0.9356771873751498, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2342 + }, + { + "epoch": 0.9360767079504595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2343 + }, + { + "epoch": 0.9364762285257691, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 2344 + }, + { + "epoch": 0.9368757491010787, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2345 + }, + { + "epoch": 0.9372752696763883, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2346 + }, + { + "epoch": 0.937674790251698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2347 + }, + { + "epoch": 0.9380743108270075, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2348 + }, + { + "epoch": 0.9384738314023172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2349 + }, + { + "epoch": 0.9388733519776269, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2350 + }, + { + "epoch": 0.9392728725529365, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2351 + }, + { + "epoch": 0.9396723931282461, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2352 + }, + { + "epoch": 0.9400719137035557, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2353 + }, + { + "epoch": 0.9404714342788654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2354 + }, + { + "epoch": 0.9408709548541749, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2355 + }, + { + "epoch": 0.9412704754294846, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2356 + }, + { + "epoch": 0.9416699960047943, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2357 + }, + { + "epoch": 0.9420695165801038, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2358 + }, + { + "epoch": 0.9424690371554135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 2359 + }, + { + "epoch": 0.9428685577307231, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2360 + }, + { + "epoch": 0.9432680783060328, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 2361 + }, + { + "epoch": 0.9436675988813424, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 2362 + }, + { + "epoch": 0.944067119456652, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 2363 + }, + { + "epoch": 0.9444666400319617, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2364 + }, + { + "epoch": 0.9448661606072712, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 2365 + }, + { + "epoch": 0.9452656811825809, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2366 + }, + { + "epoch": 0.9456652017578905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2367 + }, + { + "epoch": 0.9460647223332002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2368 + }, + { + "epoch": 0.9464642429085098, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2369 + }, + { + "epoch": 0.9468637634838194, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2370 + }, + { + "epoch": 0.9472632840591291, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2371 + }, + { + "epoch": 0.9476628046344386, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2372 + }, + { + "epoch": 0.9480623252097483, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2373 + }, + { + "epoch": 0.9484618457850579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2374 + }, + { + "epoch": 0.9488613663603676, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2375 + }, + { + "epoch": 0.9492608869356772, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2376 + }, + { + "epoch": 0.9496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2377 + }, + { + "epoch": 0.9500599280862965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2378 + }, + { + "epoch": 0.950459448661606, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2379 + }, + { + "epoch": 0.9508589692369157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2380 + }, + { + "epoch": 0.9512584898122254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2381 + }, + { + "epoch": 0.951658010387535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 2382 + }, + { + "epoch": 0.9520575309628446, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2383 + }, + { + "epoch": 0.9524570515381542, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2384 + }, + { + "epoch": 0.9528565721134639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2385 + }, + { + "epoch": 0.9532560926887734, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2386 + }, + { + "epoch": 0.9536556132640831, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2387 + }, + { + "epoch": 0.9540551338393928, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2388 + }, + { + "epoch": 0.9544546544147023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2389 + }, + { + "epoch": 0.954854174990012, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 2390 + }, + { + "epoch": 0.9552536955653216, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2391 + }, + { + "epoch": 0.9556532161406313, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 2392 + }, + { + "epoch": 0.9560527367159408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2393 + }, + { + "epoch": 0.9564522572912505, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2394 + }, + { + "epoch": 0.9568517778665602, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2395 + }, + { + "epoch": 0.9572512984418697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2396 + }, + { + "epoch": 0.9576508190171794, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2397 + }, + { + "epoch": 0.958050339592489, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2398 + }, + { + "epoch": 0.9584498601677987, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2399 + }, + { + "epoch": 0.9588493807431082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2400 + }, + { + "epoch": 0.9592489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2401 + }, + { + "epoch": 0.9596484218937276, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2402 + }, + { + "epoch": 0.9600479424690371, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2403 + }, + { + "epoch": 0.9604474630443468, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2404 + }, + { + "epoch": 0.9608469836196564, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2405 + }, + { + "epoch": 0.961246504194966, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2406 + }, + { + "epoch": 0.9616460247702757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2407 + }, + { + "epoch": 0.9620455453455853, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 2408 + }, + { + "epoch": 0.962445065920895, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2409 + }, + { + "epoch": 0.9628445864962045, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2410 + }, + { + "epoch": 0.9632441070715142, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2411 + }, + { + "epoch": 0.9636436276468238, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2412 + }, + { + "epoch": 0.9640431482221334, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2413 + }, + { + "epoch": 0.9644426687974431, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2414 + }, + { + "epoch": 0.9648421893727527, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2415 + }, + { + "epoch": 0.9652417099480624, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2416 + }, + { + "epoch": 0.9656412305233719, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2417 + }, + { + "epoch": 0.9660407510986816, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2418 + }, + { + "epoch": 0.9664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2419 + }, + { + "epoch": 0.9668397922493008, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2420 + }, + { + "epoch": 0.9672393128246105, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2421 + }, + { + "epoch": 0.9676388333999201, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2422 + }, + { + "epoch": 0.9680383539752297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2423 + }, + { + "epoch": 0.9684378745505393, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2424 + }, + { + "epoch": 0.968837395125849, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 2425 + }, + { + "epoch": 0.9692369157011587, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2426 + }, + { + "epoch": 0.9696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 2427 + }, + { + "epoch": 0.9700359568517779, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2428 + }, + { + "epoch": 0.9704354774270875, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2429 + }, + { + "epoch": 0.9708349980023971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2430 + }, + { + "epoch": 0.9712345185777067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2431 + }, + { + "epoch": 0.9716340391530164, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2432 + }, + { + "epoch": 0.9720335597283261, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2433 + }, + { + "epoch": 0.9724330803036356, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2434 + }, + { + "epoch": 0.9728326008789453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2435 + }, + { + "epoch": 0.9732321214542549, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2436 + }, + { + "epoch": 0.9736316420295645, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2437 + }, + { + "epoch": 0.9740311626048741, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2438 + }, + { + "epoch": 0.9744306831801838, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2439 + }, + { + "epoch": 0.9748302037554935, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2440 + }, + { + "epoch": 0.975229724330803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2441 + }, + { + "epoch": 0.9756292449061127, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2442 + }, + { + "epoch": 0.9760287654814223, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 2443 + }, + { + "epoch": 0.9764282860567319, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2444 + }, + { + "epoch": 0.9768278066320415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2445 + }, + { + "epoch": 0.9772273272073512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2446 + }, + { + "epoch": 0.9776268477826608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2447 + }, + { + "epoch": 0.9780263683579704, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2448 + }, + { + "epoch": 0.9784258889332801, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2449 + }, + { + "epoch": 0.9788254095085897, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2450 + }, + { + "epoch": 0.9792249300838993, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2451 + }, + { + "epoch": 0.979624450659209, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2452 + }, + { + "epoch": 0.9800239712345186, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2453 + }, + { + "epoch": 0.9804234918098282, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2454 + }, + { + "epoch": 0.9808230123851378, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2455 + }, + { + "epoch": 0.9812225329604475, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2456 + }, + { + "epoch": 0.981622053535757, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2457 + }, + { + "epoch": 0.9820215741110667, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2458 + }, + { + "epoch": 0.9824210946863764, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2459 + }, + { + "epoch": 0.982820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2460 + }, + { + "epoch": 0.9832201358369956, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2461 + }, + { + "epoch": 0.9836196564123052, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2462 + }, + { + "epoch": 0.9840191769876149, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2463 + }, + { + "epoch": 0.9844186975629244, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2464 + }, + { + "epoch": 0.9848182181382341, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2465 + }, + { + "epoch": 0.9852177387135438, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2466 + }, + { + "epoch": 0.9856172592888534, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2467 + }, + { + "epoch": 0.986016779864163, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2468 + }, + { + "epoch": 0.9864163004394726, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2469 + }, + { + "epoch": 0.9868158210147823, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2470 + }, + { + "epoch": 0.9872153415900919, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2471 + }, + { + "epoch": 0.9876148621654015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2472 + }, + { + "epoch": 0.9880143827407112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2473 + }, + { + "epoch": 0.9884139033160207, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2474 + }, + { + "epoch": 0.9888134238913304, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2475 + }, + { + "epoch": 0.98921294446664, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2476 + }, + { + "epoch": 0.9896124650419497, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2477 + }, + { + "epoch": 0.9900119856172593, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2478 + }, + { + "epoch": 0.9904115061925689, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2479 + }, + { + "epoch": 0.9908110267678786, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2480 + }, + { + "epoch": 0.9912105473431881, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2481 + }, + { + "epoch": 0.9916100679184978, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2482 + }, + { + "epoch": 0.9920095884938074, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2483 + }, + { + "epoch": 0.9924091090691171, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2484 + }, + { + "epoch": 0.9928086296444267, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2485 + }, + { + "epoch": 0.9932081502197363, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2486 + }, + { + "epoch": 0.993607670795046, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2487 + }, + { + "epoch": 0.9940071913703555, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 2488 + }, + { + "epoch": 0.9944067119456652, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2489 + }, + { + "epoch": 0.9948062325209748, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2490 + }, + { + "epoch": 0.9952057530962845, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2491 + }, + { + "epoch": 0.9956052736715941, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2492 + }, + { + "epoch": 0.9960047942469037, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2493 + }, + { + "epoch": 0.9964043148222134, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2494 + }, + { + "epoch": 0.9968038353975229, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2495 + }, + { + "epoch": 0.9972033559728326, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2496 + }, + { + "epoch": 0.9976028765481423, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2497 + }, + { + "epoch": 0.9980023971234518, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2498 + }, + { + "epoch": 0.9984019176987615, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2499 + }, + { + "epoch": 0.9988014382740711, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2500 + }, + { + "epoch": 0.9992009588493808, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2501 + }, + { + "epoch": 0.9996004794246903, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2502 + }, + { + "epoch": 1.0, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2503 + }, + { + "epoch": 1.0003995205753096, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2504 + }, + { + "epoch": 1.0007990411506194, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2505 + }, + { + "epoch": 1.001198561725929, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2506 + }, + { + "epoch": 1.0015980823012385, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2507 + }, + { + "epoch": 1.001997602876548, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2508 + }, + { + "epoch": 1.0023971234518578, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2509 + }, + { + "epoch": 1.0027966440271674, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2510 + }, + { + "epoch": 1.003196164602477, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2511 + }, + { + "epoch": 1.0035956851777867, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 2512 + }, + { + "epoch": 1.0039952057530963, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2513 + }, + { + "epoch": 1.0043947263284059, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2514 + }, + { + "epoch": 1.0047942469037154, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2515 + }, + { + "epoch": 1.0051937674790252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2516 + }, + { + "epoch": 1.0055932880543348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2517 + }, + { + "epoch": 1.0059928086296444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2518 + }, + { + "epoch": 1.0063923292049541, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2519 + }, + { + "epoch": 1.0067918497802637, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2520 + }, + { + "epoch": 1.0071913703555733, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2521 + }, + { + "epoch": 1.0075908909308828, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2522 + }, + { + "epoch": 1.0079904115061926, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2523 + }, + { + "epoch": 1.0083899320815022, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2524 + }, + { + "epoch": 1.0087894526568117, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 2525 + }, + { + "epoch": 1.0091889732321215, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2526 + }, + { + "epoch": 1.009588493807431, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2527 + }, + { + "epoch": 1.0099880143827407, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2528 + }, + { + "epoch": 1.0103875349580504, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2529 + }, + { + "epoch": 1.01078705553336, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2530 + }, + { + "epoch": 1.0111865761086696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2531 + }, + { + "epoch": 1.0115860966839791, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2532 + }, + { + "epoch": 1.011985617259289, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2533 + }, + { + "epoch": 1.0123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2534 + }, + { + "epoch": 1.012784658409908, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2535 + }, + { + "epoch": 1.0131841789852178, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 2536 + }, + { + "epoch": 1.0135836995605274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2537 + }, + { + "epoch": 1.013983220135837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2538 + }, + { + "epoch": 1.0143827407111465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2539 + }, + { + "epoch": 1.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2540 + }, + { + "epoch": 1.0151817818617659, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 2541 + }, + { + "epoch": 1.0155813024370755, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2542 + }, + { + "epoch": 1.0159808230123852, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2543 + }, + { + "epoch": 1.0163803435876948, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 2544 + }, + { + "epoch": 1.0167798641630044, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2545 + }, + { + "epoch": 1.017179384738314, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2546 + }, + { + "epoch": 1.0175789053136237, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2547 + }, + { + "epoch": 1.0179784258889333, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2548 + }, + { + "epoch": 1.0183779464642428, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2549 + }, + { + "epoch": 1.0187774670395526, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2550 + }, + { + "epoch": 1.0191769876148622, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2551 + }, + { + "epoch": 1.0195765081901718, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 2552 + }, + { + "epoch": 1.0199760287654813, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2553 + }, + { + "epoch": 1.020375549340791, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2554 + }, + { + "epoch": 1.0207750699161007, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2555 + }, + { + "epoch": 1.0211745904914102, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2556 + }, + { + "epoch": 1.02157411106672, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 2557 + }, + { + "epoch": 1.0219736316420296, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2558 + }, + { + "epoch": 1.0223731522173392, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2559 + }, + { + "epoch": 1.0227726727926487, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2560 + }, + { + "epoch": 1.0231721933679585, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2561 + }, + { + "epoch": 1.023571713943268, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 2562 + }, + { + "epoch": 1.0239712345185776, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2563 + }, + { + "epoch": 1.0243707550938874, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2564 + }, + { + "epoch": 1.024770275669197, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2565 + }, + { + "epoch": 1.0251697962445065, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2566 + }, + { + "epoch": 1.0255693168198161, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2567 + }, + { + "epoch": 1.025968837395126, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2568 + }, + { + "epoch": 1.0263683579704355, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2569 + }, + { + "epoch": 1.026767878545745, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2570 + }, + { + "epoch": 1.0271673991210548, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2571 + }, + { + "epoch": 1.0275669196963644, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2572 + }, + { + "epoch": 1.027966440271674, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2573 + }, + { + "epoch": 1.0283659608469837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2574 + }, + { + "epoch": 1.0287654814222933, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2575 + }, + { + "epoch": 1.0291650019976029, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 2576 + }, + { + "epoch": 1.0295645225729124, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2577 + }, + { + "epoch": 1.0299640431482222, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2578 + }, + { + "epoch": 1.0303635637235318, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 2579 + }, + { + "epoch": 1.0307630842988413, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2580 + }, + { + "epoch": 1.0311626048741511, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2581 + }, + { + "epoch": 1.0315621254494607, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2582 + }, + { + "epoch": 1.0319616460247703, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2583 + }, + { + "epoch": 1.0323611666000798, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2584 + }, + { + "epoch": 1.0327606871753896, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2585 + }, + { + "epoch": 1.0331602077506992, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2586 + }, + { + "epoch": 1.0335597283260087, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2587 + }, + { + "epoch": 1.0339592489013185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2588 + }, + { + "epoch": 1.034358769476628, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2589 + }, + { + "epoch": 1.0347582900519376, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2590 + }, + { + "epoch": 1.0351578106272472, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2591 + }, + { + "epoch": 1.035557331202557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 2592 + }, + { + "epoch": 1.0359568517778666, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2593 + }, + { + "epoch": 1.0363563723531761, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2594 + }, + { + "epoch": 1.036755892928486, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2595 + }, + { + "epoch": 1.0371554135037955, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2596 + }, + { + "epoch": 1.037554934079105, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 2597 + }, + { + "epoch": 1.0379544546544146, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2598 + }, + { + "epoch": 1.0383539752297244, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2599 + }, + { + "epoch": 1.038753495805034, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2600 + }, + { + "epoch": 1.0391530163803435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2601 + }, + { + "epoch": 1.0395525369556533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2602 + }, + { + "epoch": 1.0399520575309629, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2603 + }, + { + "epoch": 1.0403515781062724, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 2604 + }, + { + "epoch": 1.040751098681582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2605 + }, + { + "epoch": 1.0411506192568918, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2606 + }, + { + "epoch": 1.0415501398322013, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2607 + }, + { + "epoch": 1.041949660407511, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2608 + }, + { + "epoch": 1.0423491809828207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 2609 + }, + { + "epoch": 1.0427487015581303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2610 + }, + { + "epoch": 1.0431482221334398, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2611 + }, + { + "epoch": 1.0435477427087494, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2612 + }, + { + "epoch": 1.0439472632840592, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2613 + }, + { + "epoch": 1.0443467838593687, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2614 + }, + { + "epoch": 1.0447463044346783, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2615 + }, + { + "epoch": 1.045145825009988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2616 + }, + { + "epoch": 1.0455453455852977, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2617 + }, + { + "epoch": 1.0459448661606072, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2618 + }, + { + "epoch": 1.046344386735917, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2619 + }, + { + "epoch": 1.0467439073112266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2620 + }, + { + "epoch": 1.0471434278865361, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2621 + }, + { + "epoch": 1.0475429484618457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2622 + }, + { + "epoch": 1.0479424690371555, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2623 + }, + { + "epoch": 1.048341989612465, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2624 + }, + { + "epoch": 1.0487415101877746, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2625 + }, + { + "epoch": 1.0491410307630844, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 2626 + }, + { + "epoch": 1.049540551338394, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2627 + }, + { + "epoch": 1.0499400719137035, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2628 + }, + { + "epoch": 1.050339592489013, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2629 + }, + { + "epoch": 1.0507391130643229, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2630 + }, + { + "epoch": 1.0511386336396324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2631 + }, + { + "epoch": 1.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2632 + }, + { + "epoch": 1.0519376747902518, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2633 + }, + { + "epoch": 1.0523371953655614, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2634 + }, + { + "epoch": 1.052736715940871, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2635 + }, + { + "epoch": 1.0531362365161805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2636 + }, + { + "epoch": 1.0535357570914903, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2637 + }, + { + "epoch": 1.0539352776667998, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2638 + }, + { + "epoch": 1.0543347982421094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2639 + }, + { + "epoch": 1.0547343188174192, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2640 + }, + { + "epoch": 1.0551338393927288, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2641 + }, + { + "epoch": 1.0555333599680383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2642 + }, + { + "epoch": 1.0559328805433479, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2643 + }, + { + "epoch": 1.0563324011186577, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2644 + }, + { + "epoch": 1.0567319216939672, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2645 + }, + { + "epoch": 1.0571314422692768, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2646 + }, + { + "epoch": 1.0575309628445866, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2647 + }, + { + "epoch": 1.0579304834198962, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2648 + }, + { + "epoch": 1.0583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2649 + }, + { + "epoch": 1.0587295245705153, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2650 + }, + { + "epoch": 1.059129045145825, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 2651 + }, + { + "epoch": 1.0595285657211346, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2652 + }, + { + "epoch": 1.0599280862964442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2653 + }, + { + "epoch": 1.060327606871754, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2654 + }, + { + "epoch": 1.0607271274470635, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2655 + }, + { + "epoch": 1.061126648022373, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2656 + }, + { + "epoch": 1.0615261685976827, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2657 + }, + { + "epoch": 1.0619256891729925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2658 + }, + { + "epoch": 1.062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2659 + }, + { + "epoch": 1.0627247303236116, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2660 + }, + { + "epoch": 1.0631242508989214, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2661 + }, + { + "epoch": 1.063523771474231, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2662 + }, + { + "epoch": 1.0639232920495405, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2663 + }, + { + "epoch": 1.0643228126248503, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2664 + }, + { + "epoch": 1.0647223332001599, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2665 + }, + { + "epoch": 1.0651218537754694, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2666 + }, + { + "epoch": 1.065521374350779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2667 + }, + { + "epoch": 1.0659208949260888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2668 + }, + { + "epoch": 1.0663204155013983, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 2669 + }, + { + "epoch": 1.066719936076708, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2670 + }, + { + "epoch": 1.0671194566520175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2671 + }, + { + "epoch": 1.0675189772273272, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2672 + }, + { + "epoch": 1.0679184978026368, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2673 + }, + { + "epoch": 1.0683180183779464, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2674 + }, + { + "epoch": 1.0687175389532562, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2675 + }, + { + "epoch": 1.0691170595285657, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2676 + }, + { + "epoch": 1.0695165801038753, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2677 + }, + { + "epoch": 1.069916100679185, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 2678 + }, + { + "epoch": 1.0703156212544946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2679 + }, + { + "epoch": 1.0707151418298042, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2680 + }, + { + "epoch": 1.0711146624051138, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2681 + }, + { + "epoch": 1.0715141829804236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2682 + }, + { + "epoch": 1.0719137035557331, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2683 + }, + { + "epoch": 1.0723132241310427, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2684 + }, + { + "epoch": 1.0727127447063525, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2685 + }, + { + "epoch": 1.073112265281662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2686 + }, + { + "epoch": 1.0735117858569716, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2687 + }, + { + "epoch": 1.0739113064322812, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2688 + }, + { + "epoch": 1.074310827007591, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2689 + }, + { + "epoch": 1.0747103475829005, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2690 + }, + { + "epoch": 1.07510986815821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2691 + }, + { + "epoch": 1.0755093887335199, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2692 + }, + { + "epoch": 1.0759089093088294, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2693 + }, + { + "epoch": 1.076308429884139, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2694 + }, + { + "epoch": 1.0767079504594486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2695 + }, + { + "epoch": 1.0771074710347583, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2696 + }, + { + "epoch": 1.077506991610068, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2697 + }, + { + "epoch": 1.0779065121853775, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2698 + }, + { + "epoch": 1.0783060327606873, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 1.0787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2700 + }, + { + "epoch": 1.0791050739113064, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2701 + }, + { + "epoch": 1.079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 2702 + }, + { + "epoch": 1.0799041150619257, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2703 + }, + { + "epoch": 1.0803036356372353, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2704 + }, + { + "epoch": 1.0807031562125449, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2705 + }, + { + "epoch": 1.0811026767878547, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2706 + }, + { + "epoch": 1.0815021973631642, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2707 + }, + { + "epoch": 1.0819017179384738, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2708 + }, + { + "epoch": 1.0823012385137836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2709 + }, + { + "epoch": 1.0827007590890931, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2710 + }, + { + "epoch": 1.0831002796644027, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 2711 + }, + { + "epoch": 1.0834998002397123, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2712 + }, + { + "epoch": 1.083899320815022, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2713 + }, + { + "epoch": 1.0842988413903316, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2714 + }, + { + "epoch": 1.0846983619656412, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2715 + }, + { + "epoch": 1.0850978825409507, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2716 + }, + { + "epoch": 1.0854974031162605, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2717 + }, + { + "epoch": 1.08589692369157, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2718 + }, + { + "epoch": 1.0862964442668797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2719 + }, + { + "epoch": 1.0866959648421894, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2720 + }, + { + "epoch": 1.087095485417499, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2721 + }, + { + "epoch": 1.0874950059928086, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2722 + }, + { + "epoch": 1.0878945265681184, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2723 + }, + { + "epoch": 1.088294047143428, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2724 + }, + { + "epoch": 1.0886935677187375, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2725 + }, + { + "epoch": 1.089093088294047, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2726 + }, + { + "epoch": 1.0894926088693568, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2727 + }, + { + "epoch": 1.0898921294446664, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2728 + }, + { + "epoch": 1.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2729 + }, + { + "epoch": 1.0906911705952858, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2730 + }, + { + "epoch": 1.0910906911705953, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2731 + }, + { + "epoch": 1.0914902117459049, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2732 + }, + { + "epoch": 1.0918897323212144, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2733 + }, + { + "epoch": 1.0922892528965242, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2734 + }, + { + "epoch": 1.0926887734718338, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2735 + }, + { + "epoch": 1.0930882940471434, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2736 + }, + { + "epoch": 1.0934878146224531, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2737 + }, + { + "epoch": 1.0938873351977627, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2738 + }, + { + "epoch": 1.0942868557730723, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2739 + }, + { + "epoch": 1.0946863763483818, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2740 + }, + { + "epoch": 1.0950858969236916, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2741 + }, + { + "epoch": 1.0954854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2742 + }, + { + "epoch": 1.0958849380743108, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2743 + }, + { + "epoch": 1.0962844586496205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2744 + }, + { + "epoch": 1.09668397922493, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2745 + }, + { + "epoch": 1.0970834998002397, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2746 + }, + { + "epoch": 1.0974830203755492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2747 + }, + { + "epoch": 1.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2748 + }, + { + "epoch": 1.0982820615261686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2749 + }, + { + "epoch": 1.0986815821014781, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2750 + }, + { + "epoch": 1.099081102676788, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2751 + }, + { + "epoch": 1.0994806232520975, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2752 + }, + { + "epoch": 1.099880143827407, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2753 + }, + { + "epoch": 1.1002796644027169, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2754 + }, + { + "epoch": 1.1006791849780264, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2755 + }, + { + "epoch": 1.101078705553336, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 2756 + }, + { + "epoch": 1.1014782261286455, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2757 + }, + { + "epoch": 1.1018777467039553, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2758 + }, + { + "epoch": 1.102277267279265, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2759 + }, + { + "epoch": 1.1026767878545745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2760 + }, + { + "epoch": 1.103076308429884, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2761 + }, + { + "epoch": 1.1034758290051938, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2762 + }, + { + "epoch": 1.1038753495805034, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 2763 + }, + { + "epoch": 1.104274870155813, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2764 + }, + { + "epoch": 1.1046743907311227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2765 + }, + { + "epoch": 1.1050739113064323, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2766 + }, + { + "epoch": 1.1054734318817419, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2767 + }, + { + "epoch": 1.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2768 + }, + { + "epoch": 1.1062724730323612, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2769 + }, + { + "epoch": 1.1066719936076708, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2770 + }, + { + "epoch": 1.1070715141829803, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2771 + }, + { + "epoch": 1.1074710347582901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2772 + }, + { + "epoch": 1.1078705553335997, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2773 + }, + { + "epoch": 1.1082700759089092, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 2774 + }, + { + "epoch": 1.108669596484219, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2775 + }, + { + "epoch": 1.1090691170595286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2776 + }, + { + "epoch": 1.1094686376348382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2777 + }, + { + "epoch": 1.1098681582101477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2778 + }, + { + "epoch": 1.1102676787854575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2779 + }, + { + "epoch": 1.110667199360767, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2780 + }, + { + "epoch": 1.1110667199360766, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2781 + }, + { + "epoch": 1.1114662405113864, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2782 + }, + { + "epoch": 1.111865761086696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2783 + }, + { + "epoch": 1.1122652816620056, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2784 + }, + { + "epoch": 1.1126648022373151, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2785 + }, + { + "epoch": 1.113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2786 + }, + { + "epoch": 1.1134638433879345, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2787 + }, + { + "epoch": 1.113863363963244, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2788 + }, + { + "epoch": 1.1142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2789 + }, + { + "epoch": 1.1146624051138634, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2790 + }, + { + "epoch": 1.115061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2791 + }, + { + "epoch": 1.1154614462644825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2792 + }, + { + "epoch": 1.1158609668397923, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2793 + }, + { + "epoch": 1.1162604874151019, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2794 + }, + { + "epoch": 1.1166600079904114, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2795 + }, + { + "epoch": 1.1170595285657212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2796 + }, + { + "epoch": 1.1174590491410308, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2797 + }, + { + "epoch": 1.1178585697163403, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2798 + }, + { + "epoch": 1.1182580902916501, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2799 + }, + { + "epoch": 1.1186576108669597, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2800 + }, + { + "epoch": 1.1190571314422693, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2801 + }, + { + "epoch": 1.1194566520175788, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2802 + }, + { + "epoch": 1.1198561725928886, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2803 + }, + { + "epoch": 1.1202556931681982, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2804 + }, + { + "epoch": 1.1206552137435077, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2805 + }, + { + "epoch": 1.1210547343188173, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2806 + }, + { + "epoch": 1.121454254894127, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2807 + }, + { + "epoch": 1.1218537754694367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2808 + }, + { + "epoch": 1.1222532960447462, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2809 + }, + { + "epoch": 1.122652816620056, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2810 + }, + { + "epoch": 1.1230523371953656, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2811 + }, + { + "epoch": 1.1234518577706751, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2812 + }, + { + "epoch": 1.123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2813 + }, + { + "epoch": 1.1242508989212945, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2814 + }, + { + "epoch": 1.124650419496604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2815 + }, + { + "epoch": 1.1250499400719136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2816 + }, + { + "epoch": 1.1254494606472234, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2817 + }, + { + "epoch": 1.125848981222533, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2818 + }, + { + "epoch": 1.1262485017978425, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2819 + }, + { + "epoch": 1.1266480223731523, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2820 + }, + { + "epoch": 1.1270475429484619, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2821 + }, + { + "epoch": 1.1274470635237714, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2822 + }, + { + "epoch": 1.127846584099081, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 2823 + }, + { + "epoch": 1.1282461046743908, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 2824 + }, + { + "epoch": 1.1286456252497004, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2825 + }, + { + "epoch": 1.12904514582501, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2826 + }, + { + "epoch": 1.1294446664003197, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2827 + }, + { + "epoch": 1.1298441869756293, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2828 + }, + { + "epoch": 1.1302437075509388, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2829 + }, + { + "epoch": 1.1306432281262486, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2830 + }, + { + "epoch": 1.1310427487015582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2831 + }, + { + "epoch": 1.1314422692768678, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2832 + }, + { + "epoch": 1.1318417898521773, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2833 + }, + { + "epoch": 1.132241310427487, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2834 + }, + { + "epoch": 1.1326408310027967, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2835 + }, + { + "epoch": 1.1330403515781062, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2836 + }, + { + "epoch": 1.1334398721534158, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2837 + }, + { + "epoch": 1.1338393927287256, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2838 + }, + { + "epoch": 1.1342389133040351, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2839 + }, + { + "epoch": 1.1346384338793447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2840 + }, + { + "epoch": 1.1350379544546545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2841 + }, + { + "epoch": 1.135437475029964, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2842 + }, + { + "epoch": 1.1358369956052736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2843 + }, + { + "epoch": 1.1362365161805834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2844 + }, + { + "epoch": 1.136636036755893, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2845 + }, + { + "epoch": 1.1370355573312025, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2846 + }, + { + "epoch": 1.137435077906512, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2847 + }, + { + "epoch": 1.137834598481822, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2848 + }, + { + "epoch": 1.1382341190571315, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2849 + }, + { + "epoch": 1.138633639632441, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2850 + }, + { + "epoch": 1.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2851 + }, + { + "epoch": 1.1394326807830604, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2852 + }, + { + "epoch": 1.13983220135837, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2853 + }, + { + "epoch": 1.1402317219336795, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2854 + }, + { + "epoch": 1.1406312425089893, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2855 + }, + { + "epoch": 1.1410307630842988, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2856 + }, + { + "epoch": 1.1414302836596084, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2857 + }, + { + "epoch": 1.1418298042349182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2858 + }, + { + "epoch": 1.1422293248102278, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2859 + }, + { + "epoch": 1.1426288453855373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2860 + }, + { + "epoch": 1.143028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2861 + }, + { + "epoch": 1.1434278865361567, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2862 + }, + { + "epoch": 1.1438274071114662, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2863 + }, + { + "epoch": 1.1442269276867758, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2864 + }, + { + "epoch": 1.1446264482620856, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 2865 + }, + { + "epoch": 1.1450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2866 + }, + { + "epoch": 1.1454254894127047, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2867 + }, + { + "epoch": 1.1458250099880143, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 2868 + }, + { + "epoch": 1.146224530563324, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2869 + }, + { + "epoch": 1.1466240511386336, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2870 + }, + { + "epoch": 1.1470235717139432, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2871 + }, + { + "epoch": 1.147423092289253, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 2872 + }, + { + "epoch": 1.1478226128645626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2873 + }, + { + "epoch": 1.1482221334398721, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2874 + }, + { + "epoch": 1.148621654015182, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2875 + }, + { + "epoch": 1.1490211745904915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2876 + }, + { + "epoch": 1.149420695165801, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2877 + }, + { + "epoch": 1.1498202157411106, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 2878 + }, + { + "epoch": 1.1502197363164204, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2879 + }, + { + "epoch": 1.15061925689173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2880 + }, + { + "epoch": 1.1510187774670395, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2881 + }, + { + "epoch": 1.151418298042349, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2882 + }, + { + "epoch": 1.1518178186176589, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2883 + }, + { + "epoch": 1.1522173391929684, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2884 + }, + { + "epoch": 1.152616859768278, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2885 + }, + { + "epoch": 1.1530163803435878, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2886 + }, + { + "epoch": 1.1534159009188973, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2887 + }, + { + "epoch": 1.153815421494207, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 2888 + }, + { + "epoch": 1.1542149420695167, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2889 + }, + { + "epoch": 1.1546144626448263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2890 + }, + { + "epoch": 1.1550139832201358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 2891 + }, + { + "epoch": 1.1554135037954454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2892 + }, + { + "epoch": 1.1558130243707552, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2893 + }, + { + "epoch": 1.1562125449460647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2894 + }, + { + "epoch": 1.1566120655213743, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2895 + }, + { + "epoch": 1.1570115860966839, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2896 + }, + { + "epoch": 1.1574111066719937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2897 + }, + { + "epoch": 1.1578106272473032, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 2898 + }, + { + "epoch": 1.1582101478226128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2899 + }, + { + "epoch": 1.1586096683979226, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2900 + }, + { + "epoch": 1.1590091889732321, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2901 + }, + { + "epoch": 1.1594087095485417, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2902 + }, + { + "epoch": 1.1598082301238515, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2903 + }, + { + "epoch": 1.160207750699161, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2904 + }, + { + "epoch": 1.1606072712744706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2905 + }, + { + "epoch": 1.1610067918497802, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2906 + }, + { + "epoch": 1.16140631242509, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 2907 + }, + { + "epoch": 1.1618058330003995, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2908 + }, + { + "epoch": 1.162205353575709, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2909 + }, + { + "epoch": 1.1626048741510189, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2910 + }, + { + "epoch": 1.1630043947263284, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2911 + }, + { + "epoch": 1.163403915301638, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2912 + }, + { + "epoch": 1.1638034358769476, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2913 + }, + { + "epoch": 1.1642029564522574, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2914 + }, + { + "epoch": 1.164602477027567, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2915 + }, + { + "epoch": 1.1650019976028765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2916 + }, + { + "epoch": 1.1654015181781863, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2917 + }, + { + "epoch": 1.1658010387534958, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2918 + }, + { + "epoch": 1.1662005593288054, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2919 + }, + { + "epoch": 1.1666000799041152, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2920 + }, + { + "epoch": 1.1669996004794247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2921 + }, + { + "epoch": 1.1673991210547343, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2922 + }, + { + "epoch": 1.1677986416300439, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2923 + }, + { + "epoch": 1.1681981622053537, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2924 + }, + { + "epoch": 1.1685976827806632, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2925 + }, + { + "epoch": 1.1689972033559728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2926 + }, + { + "epoch": 1.1693967239312824, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2927 + }, + { + "epoch": 1.1697962445065921, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2928 + }, + { + "epoch": 1.1701957650819017, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2929 + }, + { + "epoch": 1.1705952856572113, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2930 + }, + { + "epoch": 1.170994806232521, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2931 + }, + { + "epoch": 1.1713943268078306, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2932 + }, + { + "epoch": 1.1717938473831402, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2933 + }, + { + "epoch": 1.17219336795845, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 2934 + }, + { + "epoch": 1.1725928885337595, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2935 + }, + { + "epoch": 1.172992409109069, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2936 + }, + { + "epoch": 1.1733919296843787, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2937 + }, + { + "epoch": 1.1737914502596885, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2938 + }, + { + "epoch": 1.174190970834998, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2939 + }, + { + "epoch": 1.1745904914103076, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2940 + }, + { + "epoch": 1.1749900119856171, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2941 + }, + { + "epoch": 1.175389532560927, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2942 + }, + { + "epoch": 1.1757890531362365, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2943 + }, + { + "epoch": 1.176188573711546, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2944 + }, + { + "epoch": 1.1765880942868558, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2945 + }, + { + "epoch": 1.1769876148621654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2946 + }, + { + "epoch": 1.177387135437475, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2947 + }, + { + "epoch": 1.1777866560127848, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 2948 + }, + { + "epoch": 1.1781861765880943, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2949 + }, + { + "epoch": 1.178585697163404, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2950 + }, + { + "epoch": 1.1789852177387135, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2951 + }, + { + "epoch": 1.1793847383140232, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2952 + }, + { + "epoch": 1.1797842588893328, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2953 + }, + { + "epoch": 1.1801837794646424, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2954 + }, + { + "epoch": 1.180583300039952, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2955 + }, + { + "epoch": 1.1809828206152617, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2956 + }, + { + "epoch": 1.1813823411905713, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2957 + }, + { + "epoch": 1.1817818617658808, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2958 + }, + { + "epoch": 1.1821813823411906, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2959 + }, + { + "epoch": 1.1825809029165002, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2960 + }, + { + "epoch": 1.1829804234918098, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2961 + }, + { + "epoch": 1.1833799440671195, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 2962 + }, + { + "epoch": 1.1837794646424291, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2963 + }, + { + "epoch": 1.1841789852177387, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 2964 + }, + { + "epoch": 1.1845785057930485, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2965 + }, + { + "epoch": 1.184978026368358, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2966 + }, + { + "epoch": 1.1853775469436676, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2967 + }, + { + "epoch": 1.1857770675189772, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2968 + }, + { + "epoch": 1.186176588094287, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2969 + }, + { + "epoch": 1.1865761086695965, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 2970 + }, + { + "epoch": 1.186975629244906, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2971 + }, + { + "epoch": 1.1873751498202156, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2972 + }, + { + "epoch": 1.1877746703955254, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 2973 + }, + { + "epoch": 1.188174190970835, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2974 + }, + { + "epoch": 1.1885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2975 + }, + { + "epoch": 1.1889732321214543, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2976 + }, + { + "epoch": 1.189372752696764, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2977 + }, + { + "epoch": 1.1897722732720735, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2978 + }, + { + "epoch": 1.1901717938473833, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2979 + }, + { + "epoch": 1.1905713144226928, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2980 + }, + { + "epoch": 1.1909708349980024, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2981 + }, + { + "epoch": 1.191370355573312, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2982 + }, + { + "epoch": 1.1917698761486217, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2983 + }, + { + "epoch": 1.1921693967239313, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2984 + }, + { + "epoch": 1.1925689172992409, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 2985 + }, + { + "epoch": 1.1929684378745504, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2986 + }, + { + "epoch": 1.1933679584498602, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2987 + }, + { + "epoch": 1.1937674790251698, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2988 + }, + { + "epoch": 1.1941669996004793, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2989 + }, + { + "epoch": 1.1945665201757891, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2990 + }, + { + "epoch": 1.1949660407510987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2991 + }, + { + "epoch": 1.1953655613264083, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2992 + }, + { + "epoch": 1.195765081901718, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2993 + }, + { + "epoch": 1.1961646024770276, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2994 + }, + { + "epoch": 1.1965641230523372, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2995 + }, + { + "epoch": 1.1969636436276467, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2996 + }, + { + "epoch": 1.1973631642029565, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2997 + }, + { + "epoch": 1.197762684778266, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2998 + }, + { + "epoch": 1.1981622053535756, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2999 + }, + { + "epoch": 1.1985617259288852, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3000 + }, + { + "epoch": 1.198961246504195, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 3001 + }, + { + "epoch": 1.1993607670795046, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3002 + }, + { + "epoch": 1.1997602876548141, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3003 + }, + { + "epoch": 1.200159808230124, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 3004 + }, + { + "epoch": 1.2005593288054335, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3005 + }, + { + "epoch": 1.200958849380743, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3006 + }, + { + "epoch": 1.2013583699560528, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3007 + }, + { + "epoch": 1.2017578905313624, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3008 + }, + { + "epoch": 1.202157411106672, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3009 + }, + { + "epoch": 1.2025569316819817, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 3010 + }, + { + "epoch": 1.2029564522572913, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 3011 + }, + { + "epoch": 1.2033559728326009, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3012 + }, + { + "epoch": 1.2037554934079104, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3013 + }, + { + "epoch": 1.2041550139832202, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3014 + }, + { + "epoch": 1.2045545345585298, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3015 + }, + { + "epoch": 1.2049540551338394, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3016 + }, + { + "epoch": 1.205353575709149, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3017 + }, + { + "epoch": 1.2057530962844587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3018 + }, + { + "epoch": 1.2061526168597683, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3019 + }, + { + "epoch": 1.2065521374350778, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3020 + }, + { + "epoch": 1.2069516580103876, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3021 + }, + { + "epoch": 1.2073511785856972, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3022 + }, + { + "epoch": 1.2077506991610067, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3023 + }, + { + "epoch": 1.2081502197363165, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3024 + }, + { + "epoch": 1.208549740311626, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3025 + }, + { + "epoch": 1.2089492608869357, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3026 + }, + { + "epoch": 1.2093487814622452, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3027 + }, + { + "epoch": 1.209748302037555, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3028 + }, + { + "epoch": 1.2101478226128646, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 3029 + }, + { + "epoch": 1.2105473431881741, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3030 + }, + { + "epoch": 1.2109468637634837, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3031 + }, + { + "epoch": 1.2113463843387935, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3032 + }, + { + "epoch": 1.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3033 + }, + { + "epoch": 1.2121454254894126, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3034 + }, + { + "epoch": 1.2125449460647224, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3035 + }, + { + "epoch": 1.212944466640032, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3036 + }, + { + "epoch": 1.2133439872153415, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3037 + }, + { + "epoch": 1.2137435077906513, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3038 + }, + { + "epoch": 1.2141430283659609, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 3039 + }, + { + "epoch": 1.2145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3040 + }, + { + "epoch": 1.21494206951658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3041 + }, + { + "epoch": 1.2153415900918898, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3042 + }, + { + "epoch": 1.2157411106671994, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3043 + }, + { + "epoch": 1.216140631242509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3044 + }, + { + "epoch": 1.2165401518178185, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3045 + }, + { + "epoch": 1.2169396723931283, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3046 + }, + { + "epoch": 1.2173391929684378, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3047 + }, + { + "epoch": 1.2177387135437474, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3048 + }, + { + "epoch": 1.2181382341190572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3049 + }, + { + "epoch": 1.2185377546943668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3050 + }, + { + "epoch": 1.2189372752696763, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 3051 + }, + { + "epoch": 1.219336795844986, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3052 + }, + { + "epoch": 1.2197363164202957, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3053 + }, + { + "epoch": 1.2201358369956052, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3054 + }, + { + "epoch": 1.220535357570915, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 3055 + }, + { + "epoch": 1.2209348781462246, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3056 + }, + { + "epoch": 1.2213343987215342, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3057 + }, + { + "epoch": 1.2217339192968437, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3058 + }, + { + "epoch": 1.2221334398721535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3059 + }, + { + "epoch": 1.222532960447463, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3060 + }, + { + "epoch": 1.2229324810227726, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3061 + }, + { + "epoch": 1.2233320015980822, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3062 + }, + { + "epoch": 1.223731522173392, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3063 + }, + { + "epoch": 1.2241310427487015, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3064 + }, + { + "epoch": 1.2245305633240111, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3065 + }, + { + "epoch": 1.224930083899321, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3066 + }, + { + "epoch": 1.2253296044746305, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3067 + }, + { + "epoch": 1.22572912504994, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3068 + }, + { + "epoch": 1.2261286456252498, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3069 + }, + { + "epoch": 1.2265281662005594, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3070 + }, + { + "epoch": 1.226927686775869, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 3071 + }, + { + "epoch": 1.2273272073511785, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3072 + }, + { + "epoch": 1.2277267279264883, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 3073 + }, + { + "epoch": 1.2281262485017979, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3074 + }, + { + "epoch": 1.2285257690771074, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3075 + }, + { + "epoch": 1.228925289652417, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 3076 + }, + { + "epoch": 1.2293248102277268, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3077 + }, + { + "epoch": 1.2297243308030363, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3078 + }, + { + "epoch": 1.230123851378346, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3079 + }, + { + "epoch": 1.2305233719536557, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3080 + }, + { + "epoch": 1.2309228925289653, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3081 + }, + { + "epoch": 1.2313224131042748, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3082 + }, + { + "epoch": 1.2317219336795846, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3083 + }, + { + "epoch": 1.2321214542548942, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3084 + }, + { + "epoch": 1.2325209748302037, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3085 + }, + { + "epoch": 1.2329204954055133, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3086 + }, + { + "epoch": 1.233320015980823, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3087 + }, + { + "epoch": 1.2337195365561326, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3088 + }, + { + "epoch": 1.2341190571314422, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3089 + }, + { + "epoch": 1.2345185777067518, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3090 + }, + { + "epoch": 1.2349180982820616, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3091 + }, + { + "epoch": 1.2353176188573711, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3092 + }, + { + "epoch": 1.2357171394326807, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3093 + }, + { + "epoch": 1.2361166600079905, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3094 + }, + { + "epoch": 1.2365161805833, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3095 + }, + { + "epoch": 1.2369157011586096, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3096 + }, + { + "epoch": 1.2373152217339194, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3097 + }, + { + "epoch": 1.237714742309229, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3098 + }, + { + "epoch": 1.2381142628845385, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3099 + }, + { + "epoch": 1.2385137834598483, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 3100 + }, + { + "epoch": 1.2389133040351579, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3101 + }, + { + "epoch": 1.2393128246104674, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3102 + }, + { + "epoch": 1.239712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3103 + }, + { + "epoch": 1.2401118657610868, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3104 + }, + { + "epoch": 1.2405113863363963, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3105 + }, + { + "epoch": 1.240910906911706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3106 + }, + { + "epoch": 1.2413104274870155, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3107 + }, + { + "epoch": 1.2417099480623253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3108 + }, + { + "epoch": 1.2421094686376348, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3109 + }, + { + "epoch": 1.2425089892129444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3110 + }, + { + "epoch": 1.2429085097882542, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3111 + }, + { + "epoch": 1.2433080303635637, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3112 + }, + { + "epoch": 1.2437075509388733, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3113 + }, + { + "epoch": 1.244107071514183, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3114 + }, + { + "epoch": 1.2445065920894927, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3115 + }, + { + "epoch": 1.2449061126648022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3116 + }, + { + "epoch": 1.2453056332401118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3117 + }, + { + "epoch": 1.2457051538154216, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3118 + }, + { + "epoch": 1.2461046743907311, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3119 + }, + { + "epoch": 1.2465041949660407, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3120 + }, + { + "epoch": 1.2469037155413503, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3121 + }, + { + "epoch": 1.24730323611666, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3122 + }, + { + "epoch": 1.2477027566919696, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3123 + }, + { + "epoch": 1.2481022772672792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3124 + }, + { + "epoch": 1.248501797842589, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3125 + }, + { + "epoch": 1.2489013184178985, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3126 + }, + { + "epoch": 1.249300838993208, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3127 + }, + { + "epoch": 1.2497003595685179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3128 + }, + { + "epoch": 1.2500998801438274, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 3129 + }, + { + "epoch": 1.250499400719137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3130 + }, + { + "epoch": 1.2508989212944468, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3131 + }, + { + "epoch": 1.2512984418697564, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3132 + }, + { + "epoch": 1.251697962445066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3133 + }, + { + "epoch": 1.2520974830203755, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3134 + }, + { + "epoch": 1.252497003595685, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3135 + }, + { + "epoch": 1.2528965241709948, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3136 + }, + { + "epoch": 1.2532960447463044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3137 + }, + { + "epoch": 1.253695565321614, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3138 + }, + { + "epoch": 1.2540950858969238, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3139 + }, + { + "epoch": 1.2544946064722333, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3140 + }, + { + "epoch": 1.2548941270475429, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3141 + }, + { + "epoch": 1.2552936476228527, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3142 + }, + { + "epoch": 1.2556931681981622, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3143 + }, + { + "epoch": 1.2560926887734718, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3144 + }, + { + "epoch": 1.2564922093487816, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3145 + }, + { + "epoch": 1.2568917299240912, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3146 + }, + { + "epoch": 1.2572912504994007, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3147 + }, + { + "epoch": 1.2576907710747103, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3148 + }, + { + "epoch": 1.2580902916500198, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3149 + }, + { + "epoch": 1.2584898122253296, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3150 + }, + { + "epoch": 1.2588893328006392, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 3151 + }, + { + "epoch": 1.2592888533759488, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3152 + }, + { + "epoch": 1.2596883739512585, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3153 + }, + { + "epoch": 1.260087894526568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3154 + }, + { + "epoch": 1.2604874151018777, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3155 + }, + { + "epoch": 1.2608869356771875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3156 + }, + { + "epoch": 1.261286456252497, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3157 + }, + { + "epoch": 1.2616859768278066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3158 + }, + { + "epoch": 1.2620854974031164, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3159 + }, + { + "epoch": 1.262485017978426, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3160 + }, + { + "epoch": 1.2628845385537355, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3161 + }, + { + "epoch": 1.263284059129045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3162 + }, + { + "epoch": 1.2636835797043549, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3163 + }, + { + "epoch": 1.2640831002796644, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3164 + }, + { + "epoch": 1.264482620854974, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3165 + }, + { + "epoch": 1.2648821414302835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3166 + }, + { + "epoch": 1.2652816620055933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3167 + }, + { + "epoch": 1.265681182580903, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3168 + }, + { + "epoch": 1.2660807031562125, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3169 + }, + { + "epoch": 1.2664802237315222, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3170 + }, + { + "epoch": 1.2668797443068318, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 3171 + }, + { + "epoch": 1.2672792648821414, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3172 + }, + { + "epoch": 1.2676787854574512, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3173 + }, + { + "epoch": 1.2680783060327607, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3174 + }, + { + "epoch": 1.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3175 + }, + { + "epoch": 1.26887734718338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3176 + }, + { + "epoch": 1.2692768677586896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3177 + }, + { + "epoch": 1.2696763883339992, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3178 + }, + { + "epoch": 1.2700759089093088, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3179 + }, + { + "epoch": 1.2704754294846183, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3180 + }, + { + "epoch": 1.2708749500599281, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3181 + }, + { + "epoch": 1.2712744706352377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3182 + }, + { + "epoch": 1.2716739912105472, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3183 + }, + { + "epoch": 1.272073511785857, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3184 + }, + { + "epoch": 1.2724730323611666, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3185 + }, + { + "epoch": 1.2728725529364762, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3186 + }, + { + "epoch": 1.273272073511786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3187 + }, + { + "epoch": 1.2736715940870955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3188 + }, + { + "epoch": 1.274071114662405, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3189 + }, + { + "epoch": 1.2744706352377149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3190 + }, + { + "epoch": 1.2748701558130244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3191 + }, + { + "epoch": 1.275269676388334, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3192 + }, + { + "epoch": 1.2756691969636436, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3193 + }, + { + "epoch": 1.2760687175389531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3194 + }, + { + "epoch": 1.276468238114263, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3195 + }, + { + "epoch": 1.2768677586895725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3196 + }, + { + "epoch": 1.277267279264882, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3197 + }, + { + "epoch": 1.2776667998401918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3198 + }, + { + "epoch": 1.2780663204155014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3199 + }, + { + "epoch": 1.278465840990811, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3200 + }, + { + "epoch": 1.2788653615661207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3201 + }, + { + "epoch": 1.2792648821414303, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3202 + }, + { + "epoch": 1.2796644027167399, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3203 + }, + { + "epoch": 1.2800639232920497, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 3204 + }, + { + "epoch": 1.2804634438673592, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3205 + }, + { + "epoch": 1.2808629644426688, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 3206 + }, + { + "epoch": 1.2812624850179783, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3207 + }, + { + "epoch": 1.2816620055932881, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 3208 + }, + { + "epoch": 1.2820615261685977, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3209 + }, + { + "epoch": 1.2824610467439073, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3210 + }, + { + "epoch": 1.2828605673192168, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3211 + }, + { + "epoch": 1.2832600878945266, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3212 + }, + { + "epoch": 1.2836596084698362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3213 + }, + { + "epoch": 1.2840591290451457, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3214 + }, + { + "epoch": 1.2844586496204555, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3215 + }, + { + "epoch": 1.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3216 + }, + { + "epoch": 1.2852576907710747, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3217 + }, + { + "epoch": 1.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3218 + }, + { + "epoch": 1.286056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3219 + }, + { + "epoch": 1.2864562524970036, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 3220 + }, + { + "epoch": 1.2868557730723134, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3221 + }, + { + "epoch": 1.287255293647623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3222 + }, + { + "epoch": 1.2876548142229325, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3223 + }, + { + "epoch": 1.288054334798242, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3224 + }, + { + "epoch": 1.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3225 + }, + { + "epoch": 1.2888533759488614, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3226 + }, + { + "epoch": 1.289252896524171, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3227 + }, + { + "epoch": 1.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3228 + }, + { + "epoch": 1.2900519376747903, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3229 + }, + { + "epoch": 1.2904514582500999, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3230 + }, + { + "epoch": 1.2908509788254094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3231 + }, + { + "epoch": 1.2912504994007192, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3232 + }, + { + "epoch": 1.2916500199760288, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3233 + }, + { + "epoch": 1.2920495405513384, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3234 + }, + { + "epoch": 1.2924490611266481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3235 + }, + { + "epoch": 1.2928485817019577, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3236 + }, + { + "epoch": 1.2932481022772673, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3237 + }, + { + "epoch": 1.2936476228525768, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3238 + }, + { + "epoch": 1.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3239 + }, + { + "epoch": 1.2944466640031962, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 3240 + }, + { + "epoch": 1.2948461845785058, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3241 + }, + { + "epoch": 1.2952457051538153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3242 + }, + { + "epoch": 1.295645225729125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3243 + }, + { + "epoch": 1.2960447463044347, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3244 + }, + { + "epoch": 1.2964442668797442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 3245 + }, + { + "epoch": 1.296843787455054, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3246 + }, + { + "epoch": 1.2972433080303636, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3247 + }, + { + "epoch": 1.2976428286056731, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3248 + }, + { + "epoch": 1.298042349180983, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3249 + }, + { + "epoch": 1.2984418697562925, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3250 + }, + { + "epoch": 1.298841390331602, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3251 + }, + { + "epoch": 1.2992409109069116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3252 + }, + { + "epoch": 1.2996404314822214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3253 + }, + { + "epoch": 1.300039952057531, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3254 + }, + { + "epoch": 1.3004394726328405, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3255 + }, + { + "epoch": 1.30083899320815, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3256 + }, + { + "epoch": 1.30123851378346, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3257 + }, + { + "epoch": 1.3016380343587695, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3258 + }, + { + "epoch": 1.302037554934079, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3259 + }, + { + "epoch": 1.3024370755093888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3260 + }, + { + "epoch": 1.3028365960846984, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3261 + }, + { + "epoch": 1.303236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3262 + }, + { + "epoch": 1.3036356372353177, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3263 + }, + { + "epoch": 1.3040351578106273, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3264 + }, + { + "epoch": 1.3044346783859369, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3265 + }, + { + "epoch": 1.3048341989612466, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3266 + }, + { + "epoch": 1.3052337195365562, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3267 + }, + { + "epoch": 1.3056332401118658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3268 + }, + { + "epoch": 1.3060327606871753, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3269 + }, + { + "epoch": 1.306432281262485, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3270 + }, + { + "epoch": 1.3068318018377947, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3271 + }, + { + "epoch": 1.3072313224131042, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3272 + }, + { + "epoch": 1.3076308429884138, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3273 + }, + { + "epoch": 1.3080303635637236, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3274 + }, + { + "epoch": 1.3084298841390332, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3275 + }, + { + "epoch": 1.3088294047143427, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3276 + }, + { + "epoch": 1.3092289252896525, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3277 + }, + { + "epoch": 1.309628445864962, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3278 + }, + { + "epoch": 1.3100279664402716, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3279 + }, + { + "epoch": 1.3104274870155814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3280 + }, + { + "epoch": 1.310827007590891, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3281 + }, + { + "epoch": 1.3112265281662006, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 3282 + }, + { + "epoch": 1.3116260487415101, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3283 + }, + { + "epoch": 1.3120255693168197, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3284 + }, + { + "epoch": 1.3124250898921295, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3285 + }, + { + "epoch": 1.312824610467439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3286 + }, + { + "epoch": 1.3132241310427486, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3287 + }, + { + "epoch": 1.3136236516180584, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3288 + }, + { + "epoch": 1.314023172193368, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3289 + }, + { + "epoch": 1.3144226927686775, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3290 + }, + { + "epoch": 1.3148222133439873, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3291 + }, + { + "epoch": 1.3152217339192969, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3292 + }, + { + "epoch": 1.3156212544946064, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3293 + }, + { + "epoch": 1.3160207750699162, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3294 + }, + { + "epoch": 1.3164202956452258, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3295 + }, + { + "epoch": 1.3168198162205353, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3296 + }, + { + "epoch": 1.317219336795845, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3297 + }, + { + "epoch": 1.3176188573711547, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3298 + }, + { + "epoch": 1.3180183779464643, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3299 + }, + { + "epoch": 1.3184178985217738, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3300 + }, + { + "epoch": 1.3188174190970834, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3301 + }, + { + "epoch": 1.3192169396723932, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3302 + }, + { + "epoch": 1.3196164602477027, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3303 + }, + { + "epoch": 1.3200159808230123, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3304 + }, + { + "epoch": 1.320415501398322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3305 + }, + { + "epoch": 1.3208150219736317, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3306 + }, + { + "epoch": 1.3212145425489412, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3307 + }, + { + "epoch": 1.321614063124251, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3308 + }, + { + "epoch": 1.3220135836995606, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3309 + }, + { + "epoch": 1.3224131042748701, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3310 + }, + { + "epoch": 1.32281262485018, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3311 + }, + { + "epoch": 1.3232121454254895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3312 + }, + { + "epoch": 1.323611666000799, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3313 + }, + { + "epoch": 1.3240111865761086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3314 + }, + { + "epoch": 1.3244107071514182, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3315 + }, + { + "epoch": 1.324810227726728, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3316 + }, + { + "epoch": 1.3252097483020375, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3317 + }, + { + "epoch": 1.325609268877347, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3318 + }, + { + "epoch": 1.3260087894526569, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3319 + }, + { + "epoch": 1.3264083100279664, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3320 + }, + { + "epoch": 1.326807830603276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3321 + }, + { + "epoch": 1.3272073511785858, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3322 + }, + { + "epoch": 1.3276068717538954, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3323 + }, + { + "epoch": 1.328006392329205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3324 + }, + { + "epoch": 1.3284059129045147, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3325 + }, + { + "epoch": 1.3288054334798243, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3326 + }, + { + "epoch": 1.3292049540551338, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3327 + }, + { + "epoch": 1.3296044746304434, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3328 + }, + { + "epoch": 1.330003995205753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3329 + }, + { + "epoch": 1.3304035157810628, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3330 + }, + { + "epoch": 1.3308030363563723, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3331 + }, + { + "epoch": 1.3312025569316819, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3332 + }, + { + "epoch": 1.3316020775069917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3333 + }, + { + "epoch": 1.3320015980823012, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3334 + }, + { + "epoch": 1.3324011186576108, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3335 + }, + { + "epoch": 1.3328006392329206, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3336 + }, + { + "epoch": 1.3332001598082301, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3337 + }, + { + "epoch": 1.3335996803835397, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3338 + }, + { + "epoch": 1.3339992009588495, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3339 + }, + { + "epoch": 1.334398721534159, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3340 + }, + { + "epoch": 1.3347982421094686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3341 + }, + { + "epoch": 1.3351977626847782, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3342 + }, + { + "epoch": 1.335597283260088, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 3343 + }, + { + "epoch": 1.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3344 + }, + { + "epoch": 1.336396324410707, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3345 + }, + { + "epoch": 1.3367958449860167, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 3346 + }, + { + "epoch": 1.3371953655613265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3347 + }, + { + "epoch": 1.337594886136636, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3348 + }, + { + "epoch": 1.3379944067119456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3349 + }, + { + "epoch": 1.3383939272872554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3350 + }, + { + "epoch": 1.338793447862565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3351 + }, + { + "epoch": 1.3391929684378745, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3352 + }, + { + "epoch": 1.3395924890131843, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3353 + }, + { + "epoch": 1.3399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3354 + }, + { + "epoch": 1.3403915301638034, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 3355 + }, + { + "epoch": 1.3407910507391132, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3356 + }, + { + "epoch": 1.3411905713144228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3357 + }, + { + "epoch": 1.3415900918897323, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3358 + }, + { + "epoch": 1.341989612465042, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3359 + }, + { + "epoch": 1.3423891330403515, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3360 + }, + { + "epoch": 1.3427886536156612, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3361 + }, + { + "epoch": 1.3431881741909708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3362 + }, + { + "epoch": 1.3435876947662804, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3363 + }, + { + "epoch": 1.3439872153415902, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3364 + }, + { + "epoch": 1.3443867359168997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3365 + }, + { + "epoch": 1.3447862564922093, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3366 + }, + { + "epoch": 1.345185777067519, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3367 + }, + { + "epoch": 1.3455852976428286, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3368 + }, + { + "epoch": 1.3459848182181382, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3369 + }, + { + "epoch": 1.346384338793448, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3370 + }, + { + "epoch": 1.3467838593687576, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 3371 + }, + { + "epoch": 1.3471833799440671, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3372 + }, + { + "epoch": 1.3475829005193767, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3373 + }, + { + "epoch": 1.3479824210946862, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3374 + }, + { + "epoch": 1.348381941669996, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 3375 + }, + { + "epoch": 1.3487814622453056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3376 + }, + { + "epoch": 1.3491809828206152, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3377 + }, + { + "epoch": 1.349580503395925, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 3378 + }, + { + "epoch": 1.3499800239712345, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 3379 + }, + { + "epoch": 1.350379544546544, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3380 + }, + { + "epoch": 1.3507790651218539, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3381 + }, + { + "epoch": 1.3511785856971634, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3382 + }, + { + "epoch": 1.351578106272473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3383 + }, + { + "epoch": 1.3519776268477828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3384 + }, + { + "epoch": 1.3523771474230923, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 3385 + }, + { + "epoch": 1.352776667998402, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3386 + }, + { + "epoch": 1.3531761885737115, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3387 + }, + { + "epoch": 1.3535757091490213, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3388 + }, + { + "epoch": 1.3539752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3389 + }, + { + "epoch": 1.3543747502996404, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3390 + }, + { + "epoch": 1.35477427087495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3391 + }, + { + "epoch": 1.3551737914502597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 3392 + }, + { + "epoch": 1.3555733120255693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3393 + }, + { + "epoch": 1.3559728326008789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3394 + }, + { + "epoch": 1.3563723531761886, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3395 + }, + { + "epoch": 1.3567718737514982, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3396 + }, + { + "epoch": 1.3571713943268078, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3397 + }, + { + "epoch": 1.3575709149021176, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3398 + }, + { + "epoch": 1.3579704354774271, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 3399 + }, + { + "epoch": 1.3583699560527367, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3400 + }, + { + "epoch": 1.3587694766280465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3401 + }, + { + "epoch": 1.359168997203356, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3402 + }, + { + "epoch": 1.3595685177786656, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3403 + }, + { + "epoch": 1.3599680383539752, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3404 + }, + { + "epoch": 1.3603675589292847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3405 + }, + { + "epoch": 1.3607670795045945, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 3406 + }, + { + "epoch": 1.361166600079904, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3407 + }, + { + "epoch": 1.3615661206552137, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3408 + }, + { + "epoch": 1.3619656412305234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3409 + }, + { + "epoch": 1.362365161805833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3410 + }, + { + "epoch": 1.3627646823811426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3411 + }, + { + "epoch": 1.3631642029564524, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 3412 + }, + { + "epoch": 1.363563723531762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3413 + }, + { + "epoch": 1.3639632441070715, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3414 + }, + { + "epoch": 1.3643627646823813, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3415 + }, + { + "epoch": 1.3647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3416 + }, + { + "epoch": 1.3651618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3417 + }, + { + "epoch": 1.36556132640831, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3418 + }, + { + "epoch": 1.3659608469836195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3419 + }, + { + "epoch": 1.3663603675589293, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3420 + }, + { + "epoch": 1.3667598881342389, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3421 + }, + { + "epoch": 1.3671594087095484, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3422 + }, + { + "epoch": 1.3675589292848582, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3423 + }, + { + "epoch": 1.3679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3424 + }, + { + "epoch": 1.3683579704354774, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3425 + }, + { + "epoch": 1.3687574910107871, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3426 + }, + { + "epoch": 1.3691570115860967, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3427 + }, + { + "epoch": 1.3695565321614063, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3428 + }, + { + "epoch": 1.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3429 + }, + { + "epoch": 1.3703555733120256, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3430 + }, + { + "epoch": 1.3707550938873352, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3431 + }, + { + "epoch": 1.3711546144626447, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3432 + }, + { + "epoch": 1.3715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3433 + }, + { + "epoch": 1.371953655613264, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 3434 + }, + { + "epoch": 1.3723531761885737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3435 + }, + { + "epoch": 1.3727526967638832, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3436 + }, + { + "epoch": 1.373152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 3437 + }, + { + "epoch": 1.3735517379145026, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3438 + }, + { + "epoch": 1.3739512584898121, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3439 + }, + { + "epoch": 1.374350779065122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3440 + }, + { + "epoch": 1.3747502996404315, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3441 + }, + { + "epoch": 1.375149820215741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3442 + }, + { + "epoch": 1.3755493407910508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3443 + }, + { + "epoch": 1.3759488613663604, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3444 + }, + { + "epoch": 1.37634838194167, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3445 + }, + { + "epoch": 1.3767479025169798, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3446 + }, + { + "epoch": 1.3771474230922893, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3447 + }, + { + "epoch": 1.3775469436675989, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3448 + }, + { + "epoch": 1.3779464642429085, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3449 + }, + { + "epoch": 1.378345984818218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3450 + }, + { + "epoch": 1.3787455053935278, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3451 + }, + { + "epoch": 1.3791450259688374, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3452 + }, + { + "epoch": 1.379544546544147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3453 + }, + { + "epoch": 1.3799440671194567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3454 + }, + { + "epoch": 1.3803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3455 + }, + { + "epoch": 1.3807431082700758, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3456 + }, + { + "epoch": 1.3811426288453856, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3457 + }, + { + "epoch": 1.3815421494206952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3458 + }, + { + "epoch": 1.3819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3459 + }, + { + "epoch": 1.3823411905713145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3460 + }, + { + "epoch": 1.3827407111466241, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3461 + }, + { + "epoch": 1.3831402317219337, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3462 + }, + { + "epoch": 1.3835397522972432, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3463 + }, + { + "epoch": 1.3839392728725528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3464 + }, + { + "epoch": 1.3843387934478626, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3465 + }, + { + "epoch": 1.3847383140231722, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3466 + }, + { + "epoch": 1.3851378345984817, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3467 + }, + { + "epoch": 1.3855373551737915, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3468 + }, + { + "epoch": 1.385936875749101, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3469 + }, + { + "epoch": 1.3863363963244106, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3470 + }, + { + "epoch": 1.3867359168997204, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3471 + }, + { + "epoch": 1.38713543747503, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3472 + }, + { + "epoch": 1.3875349580503396, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3473 + }, + { + "epoch": 1.3879344786256493, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3474 + }, + { + "epoch": 1.388333999200959, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3475 + }, + { + "epoch": 1.3887335197762685, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3476 + }, + { + "epoch": 1.389133040351578, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3477 + }, + { + "epoch": 1.3895325609268878, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3478 + }, + { + "epoch": 1.3899320815021974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3479 + }, + { + "epoch": 1.390331602077507, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3480 + }, + { + "epoch": 1.3907311226528165, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3481 + }, + { + "epoch": 1.3911306432281263, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3482 + }, + { + "epoch": 1.3915301638034359, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3483 + }, + { + "epoch": 1.3919296843787454, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3484 + }, + { + "epoch": 1.3923292049540552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3485 + }, + { + "epoch": 1.3927287255293648, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3486 + }, + { + "epoch": 1.3931282461046743, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3487 + }, + { + "epoch": 1.3935277666799841, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3488 + }, + { + "epoch": 1.3939272872552937, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3489 + }, + { + "epoch": 1.3943268078306033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3490 + }, + { + "epoch": 1.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3491 + }, + { + "epoch": 1.3951258489812226, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3492 + }, + { + "epoch": 1.3955253695565322, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3493 + }, + { + "epoch": 1.3959248901318417, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3494 + }, + { + "epoch": 1.3963244107071513, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3495 + }, + { + "epoch": 1.396723931282461, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3496 + }, + { + "epoch": 1.3971234518577706, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3497 + }, + { + "epoch": 1.3975229724330802, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3498 + }, + { + "epoch": 1.39792249300839, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3499 + }, + { + "epoch": 1.3983220135836996, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3500 + }, + { + "epoch": 1.3987215341590091, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3501 + }, + { + "epoch": 1.399121054734319, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3502 + }, + { + "epoch": 1.3995205753096285, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3503 + }, + { + "epoch": 1.399920095884938, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3504 + }, + { + "epoch": 1.4003196164602478, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3505 + }, + { + "epoch": 1.4007191370355574, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3506 + }, + { + "epoch": 1.401118657610867, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3507 + }, + { + "epoch": 1.4015181781861765, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3508 + }, + { + "epoch": 1.401917698761486, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3509 + }, + { + "epoch": 1.4023172193367959, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3510 + }, + { + "epoch": 1.4027167399121054, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3511 + }, + { + "epoch": 1.403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3512 + }, + { + "epoch": 1.4035157810627248, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3513 + }, + { + "epoch": 1.4039153016380344, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3514 + }, + { + "epoch": 1.404314822213344, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3515 + }, + { + "epoch": 1.4047143427886537, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3516 + }, + { + "epoch": 1.4051138633639633, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3517 + }, + { + "epoch": 1.4055133839392728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3518 + }, + { + "epoch": 1.4059129045145826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3519 + }, + { + "epoch": 1.4063124250898922, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3520 + }, + { + "epoch": 1.4067119456652017, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3521 + }, + { + "epoch": 1.4071114662405113, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3522 + }, + { + "epoch": 1.407510986815821, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3523 + }, + { + "epoch": 1.4079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3524 + }, + { + "epoch": 1.4083100279664402, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3525 + }, + { + "epoch": 1.4087095485417498, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3526 + }, + { + "epoch": 1.4091090691170596, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3527 + }, + { + "epoch": 1.4095085896923691, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3528 + }, + { + "epoch": 1.4099081102676787, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3529 + }, + { + "epoch": 1.4103076308429885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 3530 + }, + { + "epoch": 1.410707151418298, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3531 + }, + { + "epoch": 1.4111066719936076, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 3532 + }, + { + "epoch": 1.4115061925689174, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3533 + }, + { + "epoch": 1.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3534 + }, + { + "epoch": 1.4123052337195365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3535 + }, + { + "epoch": 1.4127047542948463, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3536 + }, + { + "epoch": 1.4131042748701559, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3537 + }, + { + "epoch": 1.4135037954454654, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3538 + }, + { + "epoch": 1.413903316020775, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3539 + }, + { + "epoch": 1.4143028365960846, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3540 + }, + { + "epoch": 1.4147023571713944, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3541 + }, + { + "epoch": 1.415101877746704, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3542 + }, + { + "epoch": 1.4155013983220135, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3543 + }, + { + "epoch": 1.4159009188973233, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3544 + }, + { + "epoch": 1.4163004394726328, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3545 + }, + { + "epoch": 1.4166999600479424, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3546 + }, + { + "epoch": 1.4170994806232522, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3547 + }, + { + "epoch": 1.4174990011985618, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3548 + }, + { + "epoch": 1.4178985217738713, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3549 + }, + { + "epoch": 1.418298042349181, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 3550 + }, + { + "epoch": 1.4186975629244907, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3551 + }, + { + "epoch": 1.4190970834998002, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3552 + }, + { + "epoch": 1.4194966040751098, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3553 + }, + { + "epoch": 1.4198961246504194, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3554 + }, + { + "epoch": 1.4202956452257292, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3555 + }, + { + "epoch": 1.4206951658010387, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3556 + }, + { + "epoch": 1.4210946863763483, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3557 + }, + { + "epoch": 1.421494206951658, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3558 + }, + { + "epoch": 1.4218937275269676, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3559 + }, + { + "epoch": 1.4222932481022772, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3560 + }, + { + "epoch": 1.422692768677587, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3561 + }, + { + "epoch": 1.4230922892528965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3562 + }, + { + "epoch": 1.4234918098282061, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3563 + }, + { + "epoch": 1.423891330403516, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3564 + }, + { + "epoch": 1.4242908509788255, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3565 + }, + { + "epoch": 1.424690371554135, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3566 + }, + { + "epoch": 1.4250898921294446, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3567 + }, + { + "epoch": 1.4254894127047544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3568 + }, + { + "epoch": 1.425888933280064, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3569 + }, + { + "epoch": 1.4262884538553735, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3570 + }, + { + "epoch": 1.426687974430683, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3571 + }, + { + "epoch": 1.4270874950059929, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3572 + }, + { + "epoch": 1.4274870155813024, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3573 + }, + { + "epoch": 1.427886536156612, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3574 + }, + { + "epoch": 1.4282860567319218, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3575 + }, + { + "epoch": 1.4286855773072313, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3576 + }, + { + "epoch": 1.429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 3577 + }, + { + "epoch": 1.4294846184578507, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3578 + }, + { + "epoch": 1.4298841390331603, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3579 + }, + { + "epoch": 1.4302836596084698, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3580 + }, + { + "epoch": 1.4306831801837796, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3581 + }, + { + "epoch": 1.4310827007590892, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3582 + }, + { + "epoch": 1.4314822213343987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3583 + }, + { + "epoch": 1.4318817419097083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3584 + }, + { + "epoch": 1.4322812624850179, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3585 + }, + { + "epoch": 1.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3586 + }, + { + "epoch": 1.4330803036356372, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3587 + }, + { + "epoch": 1.4334798242109468, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3588 + }, + { + "epoch": 1.4338793447862566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3589 + }, + { + "epoch": 1.4342788653615661, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3590 + }, + { + "epoch": 1.4346783859368757, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3591 + }, + { + "epoch": 1.4350779065121855, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3592 + }, + { + "epoch": 1.435477427087495, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3593 + }, + { + "epoch": 1.4358769476628046, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3594 + }, + { + "epoch": 1.4362764682381144, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 3595 + }, + { + "epoch": 1.436675988813424, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3596 + }, + { + "epoch": 1.4370755093887335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3597 + }, + { + "epoch": 1.437475029964043, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3598 + }, + { + "epoch": 1.4378745505393526, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3599 + }, + { + "epoch": 1.4382740711146624, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3600 + }, + { + "epoch": 1.438673591689972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3601 + }, + { + "epoch": 1.4390731122652816, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3602 + }, + { + "epoch": 1.4394726328405913, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3603 + }, + { + "epoch": 1.439872153415901, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3604 + }, + { + "epoch": 1.4402716739912105, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3605 + }, + { + "epoch": 1.4406711945665203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3606 + }, + { + "epoch": 1.4410707151418298, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3607 + }, + { + "epoch": 1.4414702357171394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3608 + }, + { + "epoch": 1.4418697562924492, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3609 + }, + { + "epoch": 1.4422692768677587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3610 + }, + { + "epoch": 1.4426687974430683, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3611 + }, + { + "epoch": 1.4430683180183779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3612 + }, + { + "epoch": 1.4434678385936877, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3613 + }, + { + "epoch": 1.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3614 + }, + { + "epoch": 1.4442668797443068, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3615 + }, + { + "epoch": 1.4446664003196164, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 3616 + }, + { + "epoch": 1.4450659208949261, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3617 + }, + { + "epoch": 1.4454654414702357, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3618 + }, + { + "epoch": 1.4458649620455453, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3619 + }, + { + "epoch": 1.446264482620855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3620 + }, + { + "epoch": 1.4466640031961646, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3621 + }, + { + "epoch": 1.4470635237714742, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3622 + }, + { + "epoch": 1.447463044346784, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3623 + }, + { + "epoch": 1.4478625649220935, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3624 + }, + { + "epoch": 1.448262085497403, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3625 + }, + { + "epoch": 1.4486616060727129, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3626 + }, + { + "epoch": 1.4490611266480224, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 3627 + }, + { + "epoch": 1.449460647223332, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3628 + }, + { + "epoch": 1.4498601677986416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3629 + }, + { + "epoch": 1.4502596883739511, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3630 + }, + { + "epoch": 1.450659208949261, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3631 + }, + { + "epoch": 1.4510587295245705, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3632 + }, + { + "epoch": 1.45145825009988, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3633 + }, + { + "epoch": 1.4518577706751898, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3634 + }, + { + "epoch": 1.4522572912504994, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3635 + }, + { + "epoch": 1.452656811825809, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3636 + }, + { + "epoch": 1.4530563324011188, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3637 + }, + { + "epoch": 1.4534558529764283, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3638 + }, + { + "epoch": 1.4538553735517379, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3639 + }, + { + "epoch": 1.4542548941270477, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3640 + }, + { + "epoch": 1.4546544147023572, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3641 + }, + { + "epoch": 1.4550539352776668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3642 + }, + { + "epoch": 1.4554534558529764, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3643 + }, + { + "epoch": 1.455852976428286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3644 + }, + { + "epoch": 1.4562524970035957, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3645 + }, + { + "epoch": 1.4566520175789053, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 3646 + }, + { + "epoch": 1.4570515381542148, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3647 + }, + { + "epoch": 1.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3648 + }, + { + "epoch": 1.4578505793048342, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3649 + }, + { + "epoch": 1.4582500998801438, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3650 + }, + { + "epoch": 1.4586496204554535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3651 + }, + { + "epoch": 1.459049141030763, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3652 + }, + { + "epoch": 1.4594486616060727, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3653 + }, + { + "epoch": 1.4598481821813825, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3654 + }, + { + "epoch": 1.460247702756692, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 3655 + }, + { + "epoch": 1.4606472233320016, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3656 + }, + { + "epoch": 1.4610467439073112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3657 + }, + { + "epoch": 1.461446264482621, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3658 + }, + { + "epoch": 1.4618457850579305, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3659 + }, + { + "epoch": 1.46224530563324, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3660 + }, + { + "epoch": 1.4626448262085496, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3661 + }, + { + "epoch": 1.4630443467838594, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3662 + }, + { + "epoch": 1.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3663 + }, + { + "epoch": 1.4638433879344785, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3664 + }, + { + "epoch": 1.4642429085097883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3665 + }, + { + "epoch": 1.464642429085098, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3666 + }, + { + "epoch": 1.4650419496604075, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3667 + }, + { + "epoch": 1.4654414702357172, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 3668 + }, + { + "epoch": 1.4658409908110268, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3669 + }, + { + "epoch": 1.4662405113863364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3670 + }, + { + "epoch": 1.4666400319616462, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3671 + }, + { + "epoch": 1.4670395525369557, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3672 + }, + { + "epoch": 1.4674390731122653, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3673 + }, + { + "epoch": 1.4678385936875749, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3674 + }, + { + "epoch": 1.4682381142628844, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3675 + }, + { + "epoch": 1.4686376348381942, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 3676 + }, + { + "epoch": 1.4690371554135038, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 3677 + }, + { + "epoch": 1.4694366759888133, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3678 + }, + { + "epoch": 1.4698361965641231, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3679 + }, + { + "epoch": 1.4702357171394327, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3680 + }, + { + "epoch": 1.4706352377147422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 3681 + }, + { + "epoch": 1.471034758290052, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3682 + }, + { + "epoch": 1.4714342788653616, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3683 + }, + { + "epoch": 1.4718337994406712, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3684 + }, + { + "epoch": 1.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3685 + }, + { + "epoch": 1.4726328405912905, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3686 + }, + { + "epoch": 1.4730323611666, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3687 + }, + { + "epoch": 1.4734318817419096, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3688 + }, + { + "epoch": 1.4738314023172192, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3689 + }, + { + "epoch": 1.474230922892529, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3690 + }, + { + "epoch": 1.4746304434678386, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3691 + }, + { + "epoch": 1.4750299640431481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3692 + }, + { + "epoch": 1.475429484618458, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3693 + }, + { + "epoch": 1.4758290051937675, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3694 + }, + { + "epoch": 1.476228525769077, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3695 + }, + { + "epoch": 1.4766280463443868, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3696 + }, + { + "epoch": 1.4770275669196964, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3697 + }, + { + "epoch": 1.477427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3698 + }, + { + "epoch": 1.4778266080703157, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3699 + }, + { + "epoch": 1.4782261286456253, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3700 + }, + { + "epoch": 1.4786256492209349, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3701 + }, + { + "epoch": 1.4790251697962444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3702 + }, + { + "epoch": 1.479424690371554, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3703 + }, + { + "epoch": 1.4798242109468638, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3704 + }, + { + "epoch": 1.4802237315221733, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3705 + }, + { + "epoch": 1.480623252097483, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3706 + }, + { + "epoch": 1.4810227726727927, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3707 + }, + { + "epoch": 1.4814222932481023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3708 + }, + { + "epoch": 1.4818218138234118, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 3709 + }, + { + "epoch": 1.4822213343987216, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 3710 + }, + { + "epoch": 1.4826208549740312, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3711 + }, + { + "epoch": 1.4830203755493407, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 3712 + }, + { + "epoch": 1.4834198961246505, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3713 + }, + { + "epoch": 1.48381941669996, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3714 + }, + { + "epoch": 1.4842189372752697, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3715 + }, + { + "epoch": 1.4846184578505794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3716 + }, + { + "epoch": 1.485017978425889, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3717 + }, + { + "epoch": 1.4854174990011986, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3718 + }, + { + "epoch": 1.4858170195765081, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3719 + }, + { + "epoch": 1.4862165401518177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3720 + }, + { + "epoch": 1.4866160607271275, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3721 + }, + { + "epoch": 1.487015581302437, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3722 + }, + { + "epoch": 1.4874151018777466, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3723 + }, + { + "epoch": 1.4878146224530564, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3724 + }, + { + "epoch": 1.488214143028366, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3725 + }, + { + "epoch": 1.4886136636036755, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3726 + }, + { + "epoch": 1.4890131841789853, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3727 + }, + { + "epoch": 1.4894127047542949, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3728 + }, + { + "epoch": 1.4898122253296044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3729 + }, + { + "epoch": 1.4902117459049142, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3730 + }, + { + "epoch": 1.4906112664802238, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3731 + }, + { + "epoch": 1.4910107870555334, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3732 + }, + { + "epoch": 1.491410307630843, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3733 + }, + { + "epoch": 1.4918098282061525, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3734 + }, + { + "epoch": 1.4922093487814623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3735 + }, + { + "epoch": 1.4926088693567718, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3736 + }, + { + "epoch": 1.4930083899320814, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3737 + }, + { + "epoch": 1.4934079105073912, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3738 + }, + { + "epoch": 1.4938074310827008, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3739 + }, + { + "epoch": 1.4942069516580103, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3740 + }, + { + "epoch": 1.49460647223332, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3741 + }, + { + "epoch": 1.4950059928086297, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3742 + }, + { + "epoch": 1.4954055133839392, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3743 + }, + { + "epoch": 1.495805033959249, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3744 + }, + { + "epoch": 1.4962045545345586, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3745 + }, + { + "epoch": 1.4966040751098681, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3746 + }, + { + "epoch": 1.4970035956851777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3747 + }, + { + "epoch": 1.4974031162604873, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3748 + }, + { + "epoch": 1.497802636835797, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3749 + }, + { + "epoch": 1.4982021574111066, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3750 + }, + { + "epoch": 1.4986016779864162, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3751 + }, + { + "epoch": 1.499001198561726, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3752 + }, + { + "epoch": 1.4994007191370355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3753 + }, + { + "epoch": 1.499800239712345, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3754 + }, + { + "epoch": 1.500199760287655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3755 + }, + { + "epoch": 1.5005992808629645, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3756 + }, + { + "epoch": 1.500998801438274, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3757 + }, + { + "epoch": 1.5013983220135838, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3758 + }, + { + "epoch": 1.5017978425888934, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3759 + }, + { + "epoch": 1.502197363164203, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 3760 + }, + { + "epoch": 1.5025968837395127, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3761 + }, + { + "epoch": 1.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3762 + }, + { + "epoch": 1.5033959248901319, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3763 + }, + { + "epoch": 1.5037954454654414, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3764 + }, + { + "epoch": 1.504194966040751, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3765 + }, + { + "epoch": 1.5045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3766 + }, + { + "epoch": 1.5049940071913703, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3767 + }, + { + "epoch": 1.50539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3768 + }, + { + "epoch": 1.5057930483419897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3769 + }, + { + "epoch": 1.5061925689172992, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3770 + }, + { + "epoch": 1.5065920894926088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3771 + }, + { + "epoch": 1.5069916100679186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3772 + }, + { + "epoch": 1.5073911306432282, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3773 + }, + { + "epoch": 1.5077906512185377, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3774 + }, + { + "epoch": 1.5081901717938475, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3775 + }, + { + "epoch": 1.5085896923691569, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3776 + }, + { + "epoch": 1.5089892129444666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3777 + }, + { + "epoch": 1.5093887335197764, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3778 + }, + { + "epoch": 1.5097882540950858, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3779 + }, + { + "epoch": 1.5101877746703956, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3780 + }, + { + "epoch": 1.5105872952457051, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3781 + }, + { + "epoch": 1.5109868158210147, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3782 + }, + { + "epoch": 1.5113863363963245, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3783 + }, + { + "epoch": 1.511785856971634, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3784 + }, + { + "epoch": 1.5121853775469436, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3785 + }, + { + "epoch": 1.5125848981222534, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3786 + }, + { + "epoch": 1.512984418697563, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3787 + }, + { + "epoch": 1.5133839392728725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3788 + }, + { + "epoch": 1.5137834598481823, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3789 + }, + { + "epoch": 1.5141829804234919, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3790 + }, + { + "epoch": 1.5145825009988014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3791 + }, + { + "epoch": 1.5149820215741112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3792 + }, + { + "epoch": 1.5153815421494206, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3793 + }, + { + "epoch": 1.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3794 + }, + { + "epoch": 1.51618058330004, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3795 + }, + { + "epoch": 1.5165801038753495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3796 + }, + { + "epoch": 1.5169796244506593, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3797 + }, + { + "epoch": 1.5173791450259688, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3798 + }, + { + "epoch": 1.5177786656012784, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3799 + }, + { + "epoch": 1.5181781861765882, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3800 + }, + { + "epoch": 1.5185777067518977, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3801 + }, + { + "epoch": 1.5189772273272073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3802 + }, + { + "epoch": 1.519376747902517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3803 + }, + { + "epoch": 1.5197762684778267, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3804 + }, + { + "epoch": 1.5201757890531362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3805 + }, + { + "epoch": 1.520575309628446, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 3806 + }, + { + "epoch": 1.5209748302037553, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3807 + }, + { + "epoch": 1.5213743507790651, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3808 + }, + { + "epoch": 1.5217738713543747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3809 + }, + { + "epoch": 1.5221733919296843, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3810 + }, + { + "epoch": 1.522572912504994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3811 + }, + { + "epoch": 1.5229724330803036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3812 + }, + { + "epoch": 1.5233719536556132, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3813 + }, + { + "epoch": 1.523771474230923, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 3814 + }, + { + "epoch": 1.5241709948062325, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3815 + }, + { + "epoch": 1.524570515381542, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3816 + }, + { + "epoch": 1.5249700359568519, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3817 + }, + { + "epoch": 1.5253695565321614, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3818 + }, + { + "epoch": 1.525769077107471, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3819 + }, + { + "epoch": 1.5261685976827808, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3820 + }, + { + "epoch": 1.5265681182580901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 3821 + }, + { + "epoch": 1.5269676388334, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3822 + }, + { + "epoch": 1.5273671594087097, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3823 + }, + { + "epoch": 1.527766679984019, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 3824 + }, + { + "epoch": 1.5281662005593288, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 3825 + }, + { + "epoch": 1.5285657211346384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3826 + }, + { + "epoch": 1.528965241709948, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3827 + }, + { + "epoch": 1.5293647622852578, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3828 + }, + { + "epoch": 1.5297642828605673, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3829 + }, + { + "epoch": 1.5301638034358769, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3830 + }, + { + "epoch": 1.5305633240111867, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3831 + }, + { + "epoch": 1.5309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 3832 + }, + { + "epoch": 1.5313623651618058, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3833 + }, + { + "epoch": 1.5317618857371156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 3834 + }, + { + "epoch": 1.5321614063124251, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3835 + }, + { + "epoch": 1.5325609268877347, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3836 + }, + { + "epoch": 1.5329604474630445, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3837 + }, + { + "epoch": 1.5333599680383538, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3838 + }, + { + "epoch": 1.5337594886136636, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3839 + }, + { + "epoch": 1.5341590091889732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 3840 + }, + { + "epoch": 1.5345585297642828, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3841 + }, + { + "epoch": 1.5349580503395925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3842 + }, + { + "epoch": 1.535357570914902, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3843 + }, + { + "epoch": 1.5357570914902117, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3844 + }, + { + "epoch": 1.5361566120655215, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 3845 + }, + { + "epoch": 1.536556132640831, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3846 + }, + { + "epoch": 1.5369556532161406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3847 + }, + { + "epoch": 1.5373551737914504, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3848 + }, + { + "epoch": 1.53775469436676, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3849 + }, + { + "epoch": 1.5381542149420695, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3850 + }, + { + "epoch": 1.5385537355173793, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3851 + }, + { + "epoch": 1.5389532560926886, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3852 + }, + { + "epoch": 1.5393527766679984, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3853 + }, + { + "epoch": 1.539752297243308, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3854 + }, + { + "epoch": 1.5401518178186175, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 3855 + }, + { + "epoch": 1.5405513383939273, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3856 + }, + { + "epoch": 1.540950858969237, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3857 + }, + { + "epoch": 1.5413503795445465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3858 + }, + { + "epoch": 1.5417499001198562, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3859 + }, + { + "epoch": 1.5421494206951658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3860 + }, + { + "epoch": 1.5425489412704754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3861 + }, + { + "epoch": 1.5429484618457852, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3862 + }, + { + "epoch": 1.5433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3863 + }, + { + "epoch": 1.5437475029964043, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3864 + }, + { + "epoch": 1.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3865 + }, + { + "epoch": 1.5445465441470234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3866 + }, + { + "epoch": 1.5449460647223332, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3867 + }, + { + "epoch": 1.545345585297643, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3868 + }, + { + "epoch": 1.5457451058729523, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3869 + }, + { + "epoch": 1.5461446264482621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 3870 + }, + { + "epoch": 1.5465441470235717, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3871 + }, + { + "epoch": 1.5469436675988812, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3872 + }, + { + "epoch": 1.547343188174191, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3873 + }, + { + "epoch": 1.5477427087495006, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3874 + }, + { + "epoch": 1.5481422293248102, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3875 + }, + { + "epoch": 1.54854174990012, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 3876 + }, + { + "epoch": 1.5489412704754295, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3877 + }, + { + "epoch": 1.549340791050739, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3878 + }, + { + "epoch": 1.5497403116260489, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3879 + }, + { + "epoch": 1.5501398322013584, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3880 + }, + { + "epoch": 1.550539352776668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3881 + }, + { + "epoch": 1.5509388733519778, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3882 + }, + { + "epoch": 1.5513383939272871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3883 + }, + { + "epoch": 1.551737914502597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 3884 + }, + { + "epoch": 1.5521374350779065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3885 + }, + { + "epoch": 1.552536955653216, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 3886 + }, + { + "epoch": 1.5529364762285258, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3887 + }, + { + "epoch": 1.5533359968038354, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3888 + }, + { + "epoch": 1.553735517379145, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3889 + }, + { + "epoch": 1.5541350379544547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3890 + }, + { + "epoch": 1.5545345585297643, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3891 + }, + { + "epoch": 1.5549340791050739, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3892 + }, + { + "epoch": 1.5553335996803836, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3893 + }, + { + "epoch": 1.5557331202556932, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3894 + }, + { + "epoch": 1.5561326408310028, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3895 + }, + { + "epoch": 1.5565321614063126, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3896 + }, + { + "epoch": 1.556931681981622, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3897 + }, + { + "epoch": 1.5573312025569317, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3898 + }, + { + "epoch": 1.5577307231322413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3899 + }, + { + "epoch": 1.5581302437075508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3900 + }, + { + "epoch": 1.5585297642828606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3901 + }, + { + "epoch": 1.5589292848581702, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3902 + }, + { + "epoch": 1.5593288054334797, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3903 + }, + { + "epoch": 1.5597283260087895, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3904 + }, + { + "epoch": 1.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3905 + }, + { + "epoch": 1.5605273671594087, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3906 + }, + { + "epoch": 1.5609268877347184, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3907 + }, + { + "epoch": 1.561326408310028, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3908 + }, + { + "epoch": 1.5617259288853376, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3909 + }, + { + "epoch": 1.5621254494606474, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3910 + }, + { + "epoch": 1.5625249700359567, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3911 + }, + { + "epoch": 1.5629244906112665, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3912 + }, + { + "epoch": 1.5633240111865763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3913 + }, + { + "epoch": 1.5637235317618856, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3914 + }, + { + "epoch": 1.5641230523371954, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3915 + }, + { + "epoch": 1.564522572912505, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3916 + }, + { + "epoch": 1.5649220934878145, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3917 + }, + { + "epoch": 1.5653216140631243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3918 + }, + { + "epoch": 1.5657211346384339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 3919 + }, + { + "epoch": 1.5661206552137434, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3920 + }, + { + "epoch": 1.5665201757890532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3921 + }, + { + "epoch": 1.5669196963643628, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 3922 + }, + { + "epoch": 1.5673192169396724, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3923 + }, + { + "epoch": 1.5677187375149821, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3924 + }, + { + "epoch": 1.5681182580902915, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3925 + }, + { + "epoch": 1.5685177786656013, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3926 + }, + { + "epoch": 1.568917299240911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3927 + }, + { + "epoch": 1.5693168198162204, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3928 + }, + { + "epoch": 1.5697163403915302, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3929 + }, + { + "epoch": 1.5701158609668397, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3930 + }, + { + "epoch": 1.5705153815421493, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3931 + }, + { + "epoch": 1.570914902117459, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3932 + }, + { + "epoch": 1.5713144226927687, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3933 + }, + { + "epoch": 1.5717139432680782, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3934 + }, + { + "epoch": 1.572113463843388, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3935 + }, + { + "epoch": 1.5725129844186976, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3936 + }, + { + "epoch": 1.5729125049940071, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3937 + }, + { + "epoch": 1.573312025569317, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 3938 + }, + { + "epoch": 1.5737115461446265, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3939 + }, + { + "epoch": 1.574111066719936, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3940 + }, + { + "epoch": 1.5745105872952458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3941 + }, + { + "epoch": 1.5749101078705552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3942 + }, + { + "epoch": 1.575309628445865, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3943 + }, + { + "epoch": 1.5757091490211745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3944 + }, + { + "epoch": 1.576108669596484, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3945 + }, + { + "epoch": 1.5765081901717939, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3946 + }, + { + "epoch": 1.5769077107471035, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3947 + }, + { + "epoch": 1.577307231322413, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3948 + }, + { + "epoch": 1.5777067518977228, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3949 + }, + { + "epoch": 1.5781062724730324, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3950 + }, + { + "epoch": 1.578505793048342, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3951 + }, + { + "epoch": 1.5789053136236517, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3952 + }, + { + "epoch": 1.5793048341989613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3953 + }, + { + "epoch": 1.5797043547742708, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3954 + }, + { + "epoch": 1.5801038753495806, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3955 + }, + { + "epoch": 1.58050339592489, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3956 + }, + { + "epoch": 1.5809029165001998, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 3957 + }, + { + "epoch": 1.5813024370755095, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3958 + }, + { + "epoch": 1.581701957650819, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 3959 + }, + { + "epoch": 1.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3960 + }, + { + "epoch": 1.5825009988014382, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3961 + }, + { + "epoch": 1.5829005193767478, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3962 + }, + { + "epoch": 1.5833000399520576, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3963 + }, + { + "epoch": 1.5836995605273672, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 3964 + }, + { + "epoch": 1.5840990811026767, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 3965 + }, + { + "epoch": 1.5844986016779865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3966 + }, + { + "epoch": 1.584898122253296, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 3967 + }, + { + "epoch": 1.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3968 + }, + { + "epoch": 1.5856971634039154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3969 + }, + { + "epoch": 1.5860966839792248, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3970 + }, + { + "epoch": 1.5864962045545346, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3971 + }, + { + "epoch": 1.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3972 + }, + { + "epoch": 1.5872952457051537, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3973 + }, + { + "epoch": 1.5876947662804635, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3974 + }, + { + "epoch": 1.588094286855773, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3975 + }, + { + "epoch": 1.5884938074310826, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3976 + }, + { + "epoch": 1.5888933280063924, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3977 + }, + { + "epoch": 1.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3978 + }, + { + "epoch": 1.5896923691570115, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 3979 + }, + { + "epoch": 1.5900918897323213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3980 + }, + { + "epoch": 1.5904914103076309, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3981 + }, + { + "epoch": 1.5908909308829404, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3982 + }, + { + "epoch": 1.5912904514582502, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3983 + }, + { + "epoch": 1.5916899720335598, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3984 + }, + { + "epoch": 1.5920894926088693, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3985 + }, + { + "epoch": 1.5924890131841791, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3986 + }, + { + "epoch": 1.5928885337594885, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 3987 + }, + { + "epoch": 1.5932880543347983, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3988 + }, + { + "epoch": 1.5936875749101078, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3989 + }, + { + "epoch": 1.5940870954854174, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3990 + }, + { + "epoch": 1.5944866160607272, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3991 + }, + { + "epoch": 1.5948861366360367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3992 + }, + { + "epoch": 1.5952856572113463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3993 + }, + { + "epoch": 1.595685177786656, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3994 + }, + { + "epoch": 1.5960846983619656, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3995 + }, + { + "epoch": 1.5964842189372752, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3996 + }, + { + "epoch": 1.596883739512585, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3997 + }, + { + "epoch": 1.5972832600878946, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3998 + }, + { + "epoch": 1.5976827806632041, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3999 + }, + { + "epoch": 1.598082301238514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4000 + }, + { + "epoch": 1.5984818218138233, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4001 + }, + { + "epoch": 1.598881342389133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4002 + }, + { + "epoch": 1.5992808629644428, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4003 + }, + { + "epoch": 1.5996803835397522, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4004 + }, + { + "epoch": 1.600079904115062, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4005 + }, + { + "epoch": 1.6004794246903715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4006 + }, + { + "epoch": 1.600878945265681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4007 + }, + { + "epoch": 1.6012784658409909, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4008 + }, + { + "epoch": 1.6016779864163004, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4009 + }, + { + "epoch": 1.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4010 + }, + { + "epoch": 1.6024770275669198, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4011 + }, + { + "epoch": 1.6028765481422294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4012 + }, + { + "epoch": 1.603276068717539, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4013 + }, + { + "epoch": 1.6036755892928487, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4014 + }, + { + "epoch": 1.604075109868158, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4015 + }, + { + "epoch": 1.6044746304434678, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 4016 + }, + { + "epoch": 1.6048741510187776, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4017 + }, + { + "epoch": 1.605273671594087, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4018 + }, + { + "epoch": 1.6056731921693967, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4019 + }, + { + "epoch": 1.6060727127447063, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4020 + }, + { + "epoch": 1.6064722333200159, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4021 + }, + { + "epoch": 1.6068717538953257, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4022 + }, + { + "epoch": 1.6072712744706352, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4023 + }, + { + "epoch": 1.6076707950459448, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4024 + }, + { + "epoch": 1.6080703156212546, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4025 + }, + { + "epoch": 1.6084698361965641, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4026 + }, + { + "epoch": 1.6088693567718737, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4027 + }, + { + "epoch": 1.6092688773471835, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4028 + }, + { + "epoch": 1.609668397922493, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4029 + }, + { + "epoch": 1.6100679184978026, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4030 + }, + { + "epoch": 1.6104674390731124, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 4031 + }, + { + "epoch": 1.6108669596484217, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4032 + }, + { + "epoch": 1.6112664802237315, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4033 + }, + { + "epoch": 1.611666000799041, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 4034 + }, + { + "epoch": 1.6120655213743507, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4035 + }, + { + "epoch": 1.6124650419496604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4036 + }, + { + "epoch": 1.61286456252497, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4037 + }, + { + "epoch": 1.6132640831002796, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 4038 + }, + { + "epoch": 1.6136636036755894, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4039 + }, + { + "epoch": 1.614063124250899, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4040 + }, + { + "epoch": 1.6144626448262085, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4041 + }, + { + "epoch": 1.6148621654015183, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4042 + }, + { + "epoch": 1.6152616859768278, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4043 + }, + { + "epoch": 1.6156612065521374, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4044 + }, + { + "epoch": 1.6160607271274472, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4045 + }, + { + "epoch": 1.6164602477027565, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4046 + }, + { + "epoch": 1.6168597682780663, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4047 + }, + { + "epoch": 1.617259288853376, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4048 + }, + { + "epoch": 1.6176588094286855, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4049 + }, + { + "epoch": 1.6180583300039952, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4050 + }, + { + "epoch": 1.6184578505793048, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4051 + }, + { + "epoch": 1.6188573711546144, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4052 + }, + { + "epoch": 1.6192568917299242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4053 + }, + { + "epoch": 1.6196564123052337, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4054 + }, + { + "epoch": 1.6200559328805433, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4055 + }, + { + "epoch": 1.620455453455853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4056 + }, + { + "epoch": 1.6208549740311626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4057 + }, + { + "epoch": 1.6212544946064722, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4058 + }, + { + "epoch": 1.621654015181782, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4059 + }, + { + "epoch": 1.6220535357570913, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4060 + }, + { + "epoch": 1.6224530563324011, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4061 + }, + { + "epoch": 1.622852576907711, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4062 + }, + { + "epoch": 1.6232520974830202, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4063 + }, + { + "epoch": 1.62365161805833, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4064 + }, + { + "epoch": 1.6240511386336396, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4065 + }, + { + "epoch": 1.6244506592089492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4066 + }, + { + "epoch": 1.624850179784259, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4067 + }, + { + "epoch": 1.6252497003595685, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4068 + }, + { + "epoch": 1.625649220934878, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4069 + }, + { + "epoch": 1.6260487415101879, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4070 + }, + { + "epoch": 1.6264482620854974, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4071 + }, + { + "epoch": 1.626847782660807, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4072 + }, + { + "epoch": 1.6272473032361168, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4073 + }, + { + "epoch": 1.6276468238114263, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4074 + }, + { + "epoch": 1.628046344386736, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 4075 + }, + { + "epoch": 1.6284458649620457, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4076 + }, + { + "epoch": 1.628845385537355, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4077 + }, + { + "epoch": 1.6292449061126648, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4078 + }, + { + "epoch": 1.6296444266879744, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 4079 + }, + { + "epoch": 1.630043947263284, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4080 + }, + { + "epoch": 1.6304434678385937, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4081 + }, + { + "epoch": 1.6308429884139033, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4082 + }, + { + "epoch": 1.6312425089892129, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4083 + }, + { + "epoch": 1.6316420295645226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4084 + }, + { + "epoch": 1.6320415501398322, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4085 + }, + { + "epoch": 1.6324410707151418, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4086 + }, + { + "epoch": 1.6328405912904516, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4087 + }, + { + "epoch": 1.6332401118657611, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4088 + }, + { + "epoch": 1.6336396324410707, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4089 + }, + { + "epoch": 1.6340391530163805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4090 + }, + { + "epoch": 1.6344386735916898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4091 + }, + { + "epoch": 1.6348381941669996, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4092 + }, + { + "epoch": 1.6352377147423094, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4093 + }, + { + "epoch": 1.6356372353176187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4094 + }, + { + "epoch": 1.6360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4095 + }, + { + "epoch": 1.636436276468238, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4096 + }, + { + "epoch": 1.6368357970435476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 4097 + }, + { + "epoch": 1.6372353176188574, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4098 + }, + { + "epoch": 1.637634838194167, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4099 + }, + { + "epoch": 1.6380343587694766, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 4100 + }, + { + "epoch": 1.6384338793447863, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4101 + }, + { + "epoch": 1.638833399920096, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4102 + }, + { + "epoch": 1.6392329204954055, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4103 + }, + { + "epoch": 1.6396324410707153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4104 + }, + { + "epoch": 1.6400319616460246, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4105 + }, + { + "epoch": 1.6404314822213344, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4106 + }, + { + "epoch": 1.6408310027966442, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4107 + }, + { + "epoch": 1.6412305233719535, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4108 + }, + { + "epoch": 1.6416300439472633, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4109 + }, + { + "epoch": 1.6420295645225729, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4110 + }, + { + "epoch": 1.6424290850978824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 4111 + }, + { + "epoch": 1.6428286056731922, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4112 + }, + { + "epoch": 1.6432281262485018, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4113 + }, + { + "epoch": 1.6436276468238114, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4114 + }, + { + "epoch": 1.6440271673991211, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4115 + }, + { + "epoch": 1.6444266879744307, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4116 + }, + { + "epoch": 1.6448262085497403, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4117 + }, + { + "epoch": 1.64522572912505, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4118 + }, + { + "epoch": 1.6456252497003596, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4119 + }, + { + "epoch": 1.6460247702756692, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4120 + }, + { + "epoch": 1.646424290850979, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4121 + }, + { + "epoch": 1.6468238114262883, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4122 + }, + { + "epoch": 1.647223332001598, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4123 + }, + { + "epoch": 1.6476228525769077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4124 + }, + { + "epoch": 1.6480223731522172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4125 + }, + { + "epoch": 1.648421893727527, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4126 + }, + { + "epoch": 1.6488214143028366, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4127 + }, + { + "epoch": 1.6492209348781461, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4128 + }, + { + "epoch": 1.649620455453456, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4129 + }, + { + "epoch": 1.6500199760287655, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4130 + }, + { + "epoch": 1.650419496604075, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 4131 + }, + { + "epoch": 1.6508190171793848, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4132 + }, + { + "epoch": 1.6512185377546944, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4133 + }, + { + "epoch": 1.651618058330004, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4134 + }, + { + "epoch": 1.6520175789053138, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4135 + }, + { + "epoch": 1.652417099480623, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 4136 + }, + { + "epoch": 1.6528166200559329, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4137 + }, + { + "epoch": 1.6532161406312427, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4138 + }, + { + "epoch": 1.653615661206552, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4139 + }, + { + "epoch": 1.6540151817818618, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4140 + }, + { + "epoch": 1.6544147023571714, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4141 + }, + { + "epoch": 1.654814222932481, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4142 + }, + { + "epoch": 1.6552137435077907, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4143 + }, + { + "epoch": 1.6556132640831003, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4144 + }, + { + "epoch": 1.6560127846584098, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4145 + }, + { + "epoch": 1.6564123052337196, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4146 + }, + { + "epoch": 1.6568118258090292, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4147 + }, + { + "epoch": 1.6572113463843388, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4148 + }, + { + "epoch": 1.6576108669596485, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4149 + }, + { + "epoch": 1.6580103875349579, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4150 + }, + { + "epoch": 1.6584099081102677, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4151 + }, + { + "epoch": 1.6588094286855775, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4152 + }, + { + "epoch": 1.6592089492608868, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4153 + }, + { + "epoch": 1.6596084698361966, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4154 + }, + { + "epoch": 1.6600079904115062, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4155 + }, + { + "epoch": 1.6604075109868157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4156 + }, + { + "epoch": 1.6608070315621255, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4157 + }, + { + "epoch": 1.661206552137435, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4158 + }, + { + "epoch": 1.6616060727127446, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4159 + }, + { + "epoch": 1.6620055932880544, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4160 + }, + { + "epoch": 1.662405113863364, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4161 + }, + { + "epoch": 1.6628046344386735, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4162 + }, + { + "epoch": 1.6632041550139833, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4163 + }, + { + "epoch": 1.663603675589293, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4164 + }, + { + "epoch": 1.6640031961646025, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4165 + }, + { + "epoch": 1.6644027167399122, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4166 + }, + { + "epoch": 1.6648022373152216, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4167 + }, + { + "epoch": 1.6652017578905314, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4168 + }, + { + "epoch": 1.665601278465841, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 4169 + }, + { + "epoch": 1.6660007990411505, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4170 + }, + { + "epoch": 1.6664003196164603, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4171 + }, + { + "epoch": 1.6667998401917699, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 4172 + }, + { + "epoch": 1.6671993607670794, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4173 + }, + { + "epoch": 1.6675988813423892, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4174 + }, + { + "epoch": 1.6679984019176988, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4175 + }, + { + "epoch": 1.6683979224930083, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 4176 + }, + { + "epoch": 1.6687974430683181, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4177 + }, + { + "epoch": 1.6691969636436277, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4178 + }, + { + "epoch": 1.6695964842189372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4179 + }, + { + "epoch": 1.669996004794247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4180 + }, + { + "epoch": 1.6703955253695564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4181 + }, + { + "epoch": 1.6707950459448662, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4182 + }, + { + "epoch": 1.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4183 + }, + { + "epoch": 1.6715940870954853, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4184 + }, + { + "epoch": 1.671993607670795, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4185 + }, + { + "epoch": 1.6723931282461046, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 4186 + }, + { + "epoch": 1.6727926488214142, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4187 + }, + { + "epoch": 1.673192169396724, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4188 + }, + { + "epoch": 1.6735916899720336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4189 + }, + { + "epoch": 1.6739912105473431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4190 + }, + { + "epoch": 1.674390731122653, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4191 + }, + { + "epoch": 1.6747902516979625, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4192 + }, + { + "epoch": 1.675189772273272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4193 + }, + { + "epoch": 1.6755892928485818, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4194 + }, + { + "epoch": 1.6759888134238912, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4195 + }, + { + "epoch": 1.676388333999201, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4196 + }, + { + "epoch": 1.6767878545745107, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4197 + }, + { + "epoch": 1.67718737514982, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4198 + }, + { + "epoch": 1.6775868957251299, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4199 + }, + { + "epoch": 1.6779864163004394, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4200 + }, + { + "epoch": 1.678385936875749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 4201 + }, + { + "epoch": 1.6787854574510588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4202 + }, + { + "epoch": 1.6791849780263683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4203 + }, + { + "epoch": 1.679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 4204 + }, + { + "epoch": 1.6799840191769877, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4205 + }, + { + "epoch": 1.6803835397522973, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4206 + }, + { + "epoch": 1.6807830603276068, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4207 + }, + { + "epoch": 1.6811825809029166, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4208 + }, + { + "epoch": 1.6815821014782262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4209 + }, + { + "epoch": 1.6819816220535357, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4210 + }, + { + "epoch": 1.6823811426288455, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4211 + }, + { + "epoch": 1.6827806632041549, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4212 + }, + { + "epoch": 1.6831801837794647, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 4213 + }, + { + "epoch": 1.6835797043547742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4214 + }, + { + "epoch": 1.6839792249300838, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 4215 + }, + { + "epoch": 1.6843787455053936, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4216 + }, + { + "epoch": 1.6847782660807031, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 4217 + }, + { + "epoch": 1.6851777866560127, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4218 + }, + { + "epoch": 1.6855773072313225, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4219 + }, + { + "epoch": 1.685976827806632, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 4220 + }, + { + "epoch": 1.6863763483819416, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 4221 + }, + { + "epoch": 1.6867758689572514, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4222 + }, + { + "epoch": 1.687175389532561, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4223 + }, + { + "epoch": 1.6875749101078705, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4224 + }, + { + "epoch": 1.6879744306831803, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4225 + }, + { + "epoch": 1.6883739512584897, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4226 + }, + { + "epoch": 1.6887734718337994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4227 + }, + { + "epoch": 1.6891729924091092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4228 + }, + { + "epoch": 1.6895725129844186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4229 + }, + { + "epoch": 1.6899720335597284, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4230 + }, + { + "epoch": 1.690371554135038, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4231 + }, + { + "epoch": 1.6907710747103475, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4232 + }, + { + "epoch": 1.6911705952856573, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4233 + }, + { + "epoch": 1.6915701158609668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4234 + }, + { + "epoch": 1.6919696364362764, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4235 + }, + { + "epoch": 1.6923691570115862, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4236 + }, + { + "epoch": 1.6927686775868958, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4237 + }, + { + "epoch": 1.6931681981622053, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4238 + }, + { + "epoch": 1.693567718737515, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4239 + }, + { + "epoch": 1.6939672393128244, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4240 + }, + { + "epoch": 1.6943667598881342, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4241 + }, + { + "epoch": 1.694766280463444, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4242 + }, + { + "epoch": 1.6951658010387534, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4243 + }, + { + "epoch": 1.6955653216140631, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4244 + }, + { + "epoch": 1.6959648421893727, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4245 + }, + { + "epoch": 1.6963643627646823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4246 + }, + { + "epoch": 1.696763883339992, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4247 + }, + { + "epoch": 1.6971634039153016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4248 + }, + { + "epoch": 1.6975629244906112, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4249 + }, + { + "epoch": 1.697962445065921, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4250 + }, + { + "epoch": 1.6983619656412305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4251 + }, + { + "epoch": 1.69876148621654, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4252 + }, + { + "epoch": 1.69916100679185, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4253 + }, + { + "epoch": 1.6995605273671595, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 4254 + }, + { + "epoch": 1.699960047942469, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4255 + }, + { + "epoch": 1.7003595685177788, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4256 + }, + { + "epoch": 1.7007590890930881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4257 + }, + { + "epoch": 1.701158609668398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4258 + }, + { + "epoch": 1.7015581302437075, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4259 + }, + { + "epoch": 1.701957650819017, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4260 + }, + { + "epoch": 1.7023571713943269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4261 + }, + { + "epoch": 1.7027566919696364, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4262 + }, + { + "epoch": 1.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4263 + }, + { + "epoch": 1.7035557331202558, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4264 + }, + { + "epoch": 1.7039552536955653, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4265 + }, + { + "epoch": 1.704354774270875, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4266 + }, + { + "epoch": 1.7047542948461847, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4267 + }, + { + "epoch": 1.7051538154214942, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4268 + }, + { + "epoch": 1.7055533359968038, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4269 + }, + { + "epoch": 1.7059528565721136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4270 + }, + { + "epoch": 1.706352377147423, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4271 + }, + { + "epoch": 1.7067518977227327, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4272 + }, + { + "epoch": 1.7071514182980425, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4273 + }, + { + "epoch": 1.7075509388733519, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4274 + }, + { + "epoch": 1.7079504594486616, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4275 + }, + { + "epoch": 1.7083499800239712, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4276 + }, + { + "epoch": 1.7087495005992808, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4277 + }, + { + "epoch": 1.7091490211745906, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4278 + }, + { + "epoch": 1.7095485417499001, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4279 + }, + { + "epoch": 1.7099480623252097, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4280 + }, + { + "epoch": 1.7103475829005195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4281 + }, + { + "epoch": 1.710747103475829, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4282 + }, + { + "epoch": 1.7111466240511386, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4283 + }, + { + "epoch": 1.7115461446264484, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4284 + }, + { + "epoch": 1.7119456652017577, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4285 + }, + { + "epoch": 1.7123451857770675, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4286 + }, + { + "epoch": 1.7127447063523773, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4287 + }, + { + "epoch": 1.7131442269276866, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4288 + }, + { + "epoch": 1.7135437475029964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4289 + }, + { + "epoch": 1.713943268078306, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4290 + }, + { + "epoch": 1.7143427886536156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4291 + }, + { + "epoch": 1.7147423092289253, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4292 + }, + { + "epoch": 1.715141829804235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4293 + }, + { + "epoch": 1.7155413503795445, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4294 + }, + { + "epoch": 1.7159408709548543, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4295 + }, + { + "epoch": 1.7163403915301638, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4296 + }, + { + "epoch": 1.7167399121054734, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4297 + }, + { + "epoch": 1.7171394326807832, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4298 + }, + { + "epoch": 1.7175389532560927, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4299 + }, + { + "epoch": 1.7179384738314023, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 4300 + }, + { + "epoch": 1.718337994406712, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4301 + }, + { + "epoch": 1.7187375149820214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4302 + }, + { + "epoch": 1.7191370355573312, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4303 + }, + { + "epoch": 1.7195365561326408, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4304 + }, + { + "epoch": 1.7199360767079503, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4305 + }, + { + "epoch": 1.7203355972832601, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4306 + }, + { + "epoch": 1.7207351178585697, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4307 + }, + { + "epoch": 1.7211346384338793, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 4308 + }, + { + "epoch": 1.721534159009189, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4309 + }, + { + "epoch": 1.7219336795844986, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4310 + }, + { + "epoch": 1.7223332001598082, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4311 + }, + { + "epoch": 1.722732720735118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4312 + }, + { + "epoch": 1.7231322413104275, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4313 + }, + { + "epoch": 1.723531761885737, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4314 + }, + { + "epoch": 1.7239312824610469, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4315 + }, + { + "epoch": 1.7243308030363562, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4316 + }, + { + "epoch": 1.724730323611666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4317 + }, + { + "epoch": 1.7251298441869758, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4318 + }, + { + "epoch": 1.7255293647622851, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4319 + }, + { + "epoch": 1.725928885337595, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4320 + }, + { + "epoch": 1.7263284059129045, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4321 + }, + { + "epoch": 1.726727926488214, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4322 + }, + { + "epoch": 1.7271274470635238, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4323 + }, + { + "epoch": 1.7275269676388334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4324 + }, + { + "epoch": 1.727926488214143, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4325 + }, + { + "epoch": 1.7283260087894528, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4326 + }, + { + "epoch": 1.7287255293647623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4327 + }, + { + "epoch": 1.7291250499400719, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4328 + }, + { + "epoch": 1.7295245705153817, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4329 + }, + { + "epoch": 1.729924091090691, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4330 + }, + { + "epoch": 1.7303236116660008, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4331 + }, + { + "epoch": 1.7307231322413106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4332 + }, + { + "epoch": 1.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4333 + }, + { + "epoch": 1.7315221733919297, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4334 + }, + { + "epoch": 1.7319216939672393, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4335 + }, + { + "epoch": 1.7323212145425488, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4336 + }, + { + "epoch": 1.7327207351178586, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4337 + }, + { + "epoch": 1.7331202556931682, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4338 + }, + { + "epoch": 1.7335197762684778, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4339 + }, + { + "epoch": 1.7339192968437875, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4340 + }, + { + "epoch": 1.734318817419097, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4341 + }, + { + "epoch": 1.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4342 + }, + { + "epoch": 1.7351178585697165, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4343 + }, + { + "epoch": 1.735517379145026, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4344 + }, + { + "epoch": 1.7359168997203356, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4345 + }, + { + "epoch": 1.7363164202956454, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4346 + }, + { + "epoch": 1.7367159408709547, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4347 + }, + { + "epoch": 1.7371154614462645, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4348 + }, + { + "epoch": 1.737514982021574, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4349 + }, + { + "epoch": 1.7379145025968836, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4350 + }, + { + "epoch": 1.7383140231721934, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4351 + }, + { + "epoch": 1.738713543747503, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4352 + }, + { + "epoch": 1.7391130643228125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4353 + }, + { + "epoch": 1.7395125848981223, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4354 + }, + { + "epoch": 1.739912105473432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4355 + }, + { + "epoch": 1.7403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4356 + }, + { + "epoch": 1.7407111466240512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 4357 + }, + { + "epoch": 1.7411106671993608, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4358 + }, + { + "epoch": 1.7415101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4359 + }, + { + "epoch": 1.7419097083499802, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4360 + }, + { + "epoch": 1.7423092289252895, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4361 + }, + { + "epoch": 1.7427087495005993, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4362 + }, + { + "epoch": 1.743108270075909, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4363 + }, + { + "epoch": 1.7435077906512184, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4364 + }, + { + "epoch": 1.7439073112265282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4365 + }, + { + "epoch": 1.7443068318018378, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4366 + }, + { + "epoch": 1.7447063523771473, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4367 + }, + { + "epoch": 1.7451058729524571, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4368 + }, + { + "epoch": 1.7455053935277667, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4369 + }, + { + "epoch": 1.7459049141030762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4370 + }, + { + "epoch": 1.746304434678386, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4371 + }, + { + "epoch": 1.7467039552536956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4372 + }, + { + "epoch": 1.7471034758290052, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4373 + }, + { + "epoch": 1.747502996404315, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4374 + }, + { + "epoch": 1.7479025169796243, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4375 + }, + { + "epoch": 1.748302037554934, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4376 + }, + { + "epoch": 1.7487015581302439, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4377 + }, + { + "epoch": 1.7491010787055532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4378 + }, + { + "epoch": 1.749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4379 + }, + { + "epoch": 1.7499001198561726, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4380 + }, + { + "epoch": 1.7502996404314821, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4381 + }, + { + "epoch": 1.750699161006792, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4382 + }, + { + "epoch": 1.7510986815821015, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4383 + }, + { + "epoch": 1.751498202157411, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4384 + }, + { + "epoch": 1.7518977227327208, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4385 + }, + { + "epoch": 1.7522972433080304, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4386 + }, + { + "epoch": 1.75269676388334, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4387 + }, + { + "epoch": 1.7530962844586497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4388 + }, + { + "epoch": 1.7534958050339593, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4389 + }, + { + "epoch": 1.7538953256092689, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4390 + }, + { + "epoch": 1.7542948461845786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4391 + }, + { + "epoch": 1.754694366759888, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4392 + }, + { + "epoch": 1.7550938873351978, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 4393 + }, + { + "epoch": 1.7554934079105073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4394 + }, + { + "epoch": 1.755892928485817, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4395 + }, + { + "epoch": 1.7562924490611267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4396 + }, + { + "epoch": 1.7566919696364363, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4397 + }, + { + "epoch": 1.7570914902117458, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4398 + }, + { + "epoch": 1.7574910107870556, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4399 + }, + { + "epoch": 1.7578905313623652, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4400 + }, + { + "epoch": 1.7582900519376747, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4401 + }, + { + "epoch": 1.7586895725129845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4402 + }, + { + "epoch": 1.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4403 + }, + { + "epoch": 1.7594886136636037, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4404 + }, + { + "epoch": 1.7598881342389134, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4405 + }, + { + "epoch": 1.7602876548142228, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 4406 + }, + { + "epoch": 1.7606871753895326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 4407 + }, + { + "epoch": 1.7610866959648424, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4408 + }, + { + "epoch": 1.7614862165401517, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 4409 + }, + { + "epoch": 1.7618857371154615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4410 + }, + { + "epoch": 1.762285257690771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4411 + }, + { + "epoch": 1.7626847782660806, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4412 + }, + { + "epoch": 1.7630842988413904, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4413 + }, + { + "epoch": 1.7634838194167, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4414 + }, + { + "epoch": 1.7638833399920095, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4415 + }, + { + "epoch": 1.7642828605673193, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4416 + }, + { + "epoch": 1.7646823811426289, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4417 + }, + { + "epoch": 1.7650819017179384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4418 + }, + { + "epoch": 1.7654814222932482, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4419 + }, + { + "epoch": 1.7658809428685576, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4420 + }, + { + "epoch": 1.7662804634438674, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4421 + }, + { + "epoch": 1.7666799840191771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4422 + }, + { + "epoch": 1.7670795045944865, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4423 + }, + { + "epoch": 1.7674790251697963, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4424 + }, + { + "epoch": 1.7678785457451058, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4425 + }, + { + "epoch": 1.7682780663204154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4426 + }, + { + "epoch": 1.7686775868957252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4427 + }, + { + "epoch": 1.7690771074710347, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4428 + }, + { + "epoch": 1.7694766280463443, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4429 + }, + { + "epoch": 1.769876148621654, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4430 + }, + { + "epoch": 1.7702756691969637, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4431 + }, + { + "epoch": 1.7706751897722732, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4432 + }, + { + "epoch": 1.771074710347583, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4433 + }, + { + "epoch": 1.7714742309228926, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4434 + }, + { + "epoch": 1.7718737514982021, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4435 + }, + { + "epoch": 1.772273272073512, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4436 + }, + { + "epoch": 1.7726727926488213, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4437 + }, + { + "epoch": 1.773072313224131, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4438 + }, + { + "epoch": 1.7734718337994406, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4439 + }, + { + "epoch": 1.7738713543747502, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4440 + }, + { + "epoch": 1.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4441 + }, + { + "epoch": 1.7746703955253695, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4442 + }, + { + "epoch": 1.775069916100679, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 4443 + }, + { + "epoch": 1.7754694366759889, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 4444 + }, + { + "epoch": 1.7758689572512985, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4445 + }, + { + "epoch": 1.776268477826608, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 4446 + }, + { + "epoch": 1.7766679984019178, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 4447 + }, + { + "epoch": 1.7770675189772274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4448 + }, + { + "epoch": 1.777467039552537, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 4449 + }, + { + "epoch": 1.7778665601278467, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 4450 + }, + { + "epoch": 1.778266080703156, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4451 + }, + { + "epoch": 1.7786656012784658, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4452 + }, + { + "epoch": 1.7790651218537756, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4453 + }, + { + "epoch": 1.779464642429085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4454 + }, + { + "epoch": 1.7798641630043948, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4455 + }, + { + "epoch": 1.7802636835797043, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4456 + }, + { + "epoch": 1.780663204155014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4457 + }, + { + "epoch": 1.7810627247303237, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4458 + }, + { + "epoch": 1.7814622453056332, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4459 + }, + { + "epoch": 1.7818617658809428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4460 + }, + { + "epoch": 1.7822612864562526, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4461 + }, + { + "epoch": 1.7826608070315622, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4462 + }, + { + "epoch": 1.7830603276068717, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4463 + }, + { + "epoch": 1.7834598481821815, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4464 + }, + { + "epoch": 1.7838593687574908, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4465 + }, + { + "epoch": 1.7842588893328006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4466 + }, + { + "epoch": 1.7846584099081104, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4467 + }, + { + "epoch": 1.7850579304834198, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4468 + }, + { + "epoch": 1.7854574510587295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4469 + }, + { + "epoch": 1.7858569716340391, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4470 + }, + { + "epoch": 1.7862564922093487, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4471 + }, + { + "epoch": 1.7866560127846585, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4472 + }, + { + "epoch": 1.787055533359968, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4473 + }, + { + "epoch": 1.7874550539352776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4474 + }, + { + "epoch": 1.7878545745105874, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4475 + }, + { + "epoch": 1.788254095085897, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4476 + }, + { + "epoch": 1.7886536156612065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4477 + }, + { + "epoch": 1.7890531362365163, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4478 + }, + { + "epoch": 1.7894526568118259, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4479 + }, + { + "epoch": 1.7898521773871354, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4480 + }, + { + "epoch": 1.7902516979624452, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4481 + }, + { + "epoch": 1.7906512185377546, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 4482 + }, + { + "epoch": 1.7910507391130643, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4483 + }, + { + "epoch": 1.791450259688374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4484 + }, + { + "epoch": 1.7918497802636835, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4485 + }, + { + "epoch": 1.7922493008389933, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4486 + }, + { + "epoch": 1.7926488214143028, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4487 + }, + { + "epoch": 1.7930483419896124, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4488 + }, + { + "epoch": 1.7934478625649222, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4489 + }, + { + "epoch": 1.7938473831402317, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4490 + }, + { + "epoch": 1.7942469037155413, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4491 + }, + { + "epoch": 1.794646424290851, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4492 + }, + { + "epoch": 1.7950459448661606, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4493 + }, + { + "epoch": 1.7954454654414702, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4494 + }, + { + "epoch": 1.79584498601678, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4495 + }, + { + "epoch": 1.7962445065920893, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4496 + }, + { + "epoch": 1.7966440271673991, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4497 + }, + { + "epoch": 1.797043547742709, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4498 + }, + { + "epoch": 1.7974430683180183, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4499 + }, + { + "epoch": 1.797842588893328, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4500 + }, + { + "epoch": 1.7982421094686376, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4501 + }, + { + "epoch": 1.7986416300439472, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4502 + }, + { + "epoch": 1.799041150619257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4503 + }, + { + "epoch": 1.7994406711945665, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4504 + }, + { + "epoch": 1.799840191769876, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4505 + }, + { + "epoch": 1.8002397123451859, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4506 + }, + { + "epoch": 1.8006392329204954, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4507 + }, + { + "epoch": 1.801038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4508 + }, + { + "epoch": 1.8014382740711148, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4509 + }, + { + "epoch": 1.8018377946464241, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4510 + }, + { + "epoch": 1.802237315221734, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4511 + }, + { + "epoch": 1.8026368357970437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4512 + }, + { + "epoch": 1.803036356372353, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4513 + }, + { + "epoch": 1.8034358769476628, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4514 + }, + { + "epoch": 1.8038353975229724, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4515 + }, + { + "epoch": 1.804234918098282, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 4516 + }, + { + "epoch": 1.8046344386735917, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4517 + }, + { + "epoch": 1.8050339592489013, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4518 + }, + { + "epoch": 1.8054334798242109, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4519 + }, + { + "epoch": 1.8058330003995207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4520 + }, + { + "epoch": 1.8062325209748302, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4521 + }, + { + "epoch": 1.8066320415501398, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 4522 + }, + { + "epoch": 1.8070315621254496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4523 + }, + { + "epoch": 1.8074310827007591, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4524 + }, + { + "epoch": 1.8078306032760687, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4525 + }, + { + "epoch": 1.8082301238513785, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4526 + }, + { + "epoch": 1.8086296444266878, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4527 + }, + { + "epoch": 1.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4528 + }, + { + "epoch": 1.8094286855773072, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4529 + }, + { + "epoch": 1.8098282061526167, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4530 + }, + { + "epoch": 1.8102277267279265, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4531 + }, + { + "epoch": 1.810627247303236, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4532 + }, + { + "epoch": 1.8110267678785457, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4533 + }, + { + "epoch": 1.8114262884538554, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4534 + }, + { + "epoch": 1.811825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4535 + }, + { + "epoch": 1.8122253296044746, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4536 + }, + { + "epoch": 1.8126248501797844, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4537 + }, + { + "epoch": 1.813024370755094, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 4538 + }, + { + "epoch": 1.8134238913304035, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4539 + }, + { + "epoch": 1.8138234119057133, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 4540 + }, + { + "epoch": 1.8142229324810226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 4541 + }, + { + "epoch": 1.8146224530563324, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4542 + }, + { + "epoch": 1.8150219736316422, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4543 + }, + { + "epoch": 1.8154214942069515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4544 + }, + { + "epoch": 1.8158210147822613, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4545 + }, + { + "epoch": 1.8162205353575709, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4546 + }, + { + "epoch": 1.8166200559328805, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4547 + }, + { + "epoch": 1.8170195765081902, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4548 + }, + { + "epoch": 1.8174190970834998, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4549 + }, + { + "epoch": 1.8178186176588094, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4550 + }, + { + "epoch": 1.8182181382341192, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4551 + }, + { + "epoch": 1.8186176588094287, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4552 + }, + { + "epoch": 1.8190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4553 + }, + { + "epoch": 1.819416699960048, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4554 + }, + { + "epoch": 1.8198162205353574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4555 + }, + { + "epoch": 1.8202157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4556 + }, + { + "epoch": 1.820615261685977, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4557 + }, + { + "epoch": 1.8210147822612863, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4558 + }, + { + "epoch": 1.821414302836596, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4559 + }, + { + "epoch": 1.8218138234119057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4560 + }, + { + "epoch": 1.8222133439872152, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4561 + }, + { + "epoch": 1.822612864562525, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4562 + }, + { + "epoch": 1.8230123851378346, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4563 + }, + { + "epoch": 1.8234119057131442, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4564 + }, + { + "epoch": 1.823811426288454, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4565 + }, + { + "epoch": 1.8242109468637635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4566 + }, + { + "epoch": 1.824610467439073, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4567 + }, + { + "epoch": 1.8250099880143829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4568 + }, + { + "epoch": 1.8254095085896924, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4569 + }, + { + "epoch": 1.825809029165002, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4570 + }, + { + "epoch": 1.8262085497403118, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4571 + }, + { + "epoch": 1.8266080703156211, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4572 + }, + { + "epoch": 1.827007590890931, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4573 + }, + { + "epoch": 1.8274071114662405, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4574 + }, + { + "epoch": 1.82780663204155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4575 + }, + { + "epoch": 1.8282061526168598, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4576 + }, + { + "epoch": 1.8286056731921694, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4577 + }, + { + "epoch": 1.829005193767479, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4578 + }, + { + "epoch": 1.8294047143427887, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4579 + }, + { + "epoch": 1.8298042349180983, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4580 + }, + { + "epoch": 1.8302037554934079, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4581 + }, + { + "epoch": 1.8306032760687176, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4582 + }, + { + "epoch": 1.8310027966440272, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4583 + }, + { + "epoch": 1.8314023172193368, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4584 + }, + { + "epoch": 1.8318018377946466, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4585 + }, + { + "epoch": 1.832201358369956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4586 + }, + { + "epoch": 1.8326008789452657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4587 + }, + { + "epoch": 1.8330003995205755, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4588 + }, + { + "epoch": 1.8333999200958848, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4589 + }, + { + "epoch": 1.8337994406711946, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4590 + }, + { + "epoch": 1.8341989612465042, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4591 + }, + { + "epoch": 1.8345984818218137, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4592 + }, + { + "epoch": 1.8349980023971235, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4593 + }, + { + "epoch": 1.835397522972433, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4594 + }, + { + "epoch": 1.8357970435477426, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4595 + }, + { + "epoch": 1.8361965641230524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4596 + }, + { + "epoch": 1.836596084698362, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4597 + }, + { + "epoch": 1.8369956052736716, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4598 + }, + { + "epoch": 1.8373951258489813, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4599 + }, + { + "epoch": 1.8377946464242907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4600 + }, + { + "epoch": 1.8381941669996005, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4601 + }, + { + "epoch": 1.8385936875749103, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4602 + }, + { + "epoch": 1.8389932081502196, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4603 + }, + { + "epoch": 1.8393927287255294, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4604 + }, + { + "epoch": 1.839792249300839, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4605 + }, + { + "epoch": 1.8401917698761485, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4606 + }, + { + "epoch": 1.8405912904514583, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4607 + }, + { + "epoch": 1.8409908110267679, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 4608 + }, + { + "epoch": 1.8413903316020774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4609 + }, + { + "epoch": 1.8417898521773872, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4610 + }, + { + "epoch": 1.8421893727526968, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4611 + }, + { + "epoch": 1.8425888933280063, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4612 + }, + { + "epoch": 1.8429884139033161, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4613 + }, + { + "epoch": 1.8433879344786257, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4614 + }, + { + "epoch": 1.8437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4615 + }, + { + "epoch": 1.844186975629245, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4616 + }, + { + "epoch": 1.8445864962045544, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4617 + }, + { + "epoch": 1.8449860167798642, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4618 + }, + { + "epoch": 1.8453855373551737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4619 + }, + { + "epoch": 1.8457850579304833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4620 + }, + { + "epoch": 1.846184578505793, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4621 + }, + { + "epoch": 1.8465840990811027, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4622 + }, + { + "epoch": 1.8469836196564122, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4623 + }, + { + "epoch": 1.847383140231722, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4624 + }, + { + "epoch": 1.8477826608070316, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4625 + }, + { + "epoch": 1.8481821813823411, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4626 + }, + { + "epoch": 1.848581701957651, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4627 + }, + { + "epoch": 1.8489812225329605, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4628 + }, + { + "epoch": 1.84938074310827, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4629 + }, + { + "epoch": 1.8497802636835798, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4630 + }, + { + "epoch": 1.8501797842588892, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4631 + }, + { + "epoch": 1.850579304834199, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4632 + }, + { + "epoch": 1.8509788254095088, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 4633 + }, + { + "epoch": 1.851378345984818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4634 + }, + { + "epoch": 1.8517778665601279, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4635 + }, + { + "epoch": 1.8521773871354374, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4636 + }, + { + "epoch": 1.852576907710747, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4637 + }, + { + "epoch": 1.8529764282860568, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4638 + }, + { + "epoch": 1.8533759488613664, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4639 + }, + { + "epoch": 1.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4640 + }, + { + "epoch": 1.8541749900119857, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4641 + }, + { + "epoch": 1.8545745105872953, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 4642 + }, + { + "epoch": 1.8549740311626048, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4643 + }, + { + "epoch": 1.8553735517379146, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4644 + }, + { + "epoch": 1.855773072313224, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4645 + }, + { + "epoch": 1.8561725928885338, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4646 + }, + { + "epoch": 1.8565721134638435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4647 + }, + { + "epoch": 1.8569716340391529, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4648 + }, + { + "epoch": 1.8573711546144627, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4649 + }, + { + "epoch": 1.8577706751897722, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4650 + }, + { + "epoch": 1.8581701957650818, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4651 + }, + { + "epoch": 1.8585697163403916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4652 + }, + { + "epoch": 1.8589692369157012, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4653 + }, + { + "epoch": 1.8593687574910107, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4654 + }, + { + "epoch": 1.8597682780663205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4655 + }, + { + "epoch": 1.86016779864163, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4656 + }, + { + "epoch": 1.8605673192169396, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4657 + }, + { + "epoch": 1.8609668397922494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4658 + }, + { + "epoch": 1.861366360367559, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4659 + }, + { + "epoch": 1.8617658809428685, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4660 + }, + { + "epoch": 1.8621654015181783, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4661 + }, + { + "epoch": 1.8625649220934877, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4662 + }, + { + "epoch": 1.8629644426687975, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4663 + }, + { + "epoch": 1.863363963244107, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4664 + }, + { + "epoch": 1.8637634838194166, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4665 + }, + { + "epoch": 1.8641630043947264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4666 + }, + { + "epoch": 1.864562524970036, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4667 + }, + { + "epoch": 1.8649620455453455, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4668 + }, + { + "epoch": 1.8653615661206553, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4669 + }, + { + "epoch": 1.8657610866959649, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4670 + }, + { + "epoch": 1.8661606072712744, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4671 + }, + { + "epoch": 1.8665601278465842, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4672 + }, + { + "epoch": 1.8669596484218938, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4673 + }, + { + "epoch": 1.8673591689972033, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4674 + }, + { + "epoch": 1.8677586895725131, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4675 + }, + { + "epoch": 1.8681582101478225, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4676 + }, + { + "epoch": 1.8685577307231322, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4677 + }, + { + "epoch": 1.868957251298442, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4678 + }, + { + "epoch": 1.8693567718737514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4679 + }, + { + "epoch": 1.8697562924490612, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4680 + }, + { + "epoch": 1.8701558130243707, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4681 + }, + { + "epoch": 1.8705553335996803, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4682 + }, + { + "epoch": 1.87095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4683 + }, + { + "epoch": 1.8713543747502996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4684 + }, + { + "epoch": 1.8717538953256092, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4685 + }, + { + "epoch": 1.872153415900919, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4686 + }, + { + "epoch": 1.8725529364762286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4687 + }, + { + "epoch": 1.8729524570515381, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4688 + }, + { + "epoch": 1.873351977626848, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 4689 + }, + { + "epoch": 1.8737514982021573, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4690 + }, + { + "epoch": 1.874151018777467, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4691 + }, + { + "epoch": 1.8745505393527768, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4692 + }, + { + "epoch": 1.8749500599280862, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4693 + }, + { + "epoch": 1.875349580503396, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4694 + }, + { + "epoch": 1.8757491010787055, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4695 + }, + { + "epoch": 1.876148621654015, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4696 + }, + { + "epoch": 1.8765481422293249, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4697 + }, + { + "epoch": 1.8769476628046344, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4698 + }, + { + "epoch": 1.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4699 + }, + { + "epoch": 1.8777467039552538, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4700 + }, + { + "epoch": 1.8781462245305633, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4701 + }, + { + "epoch": 1.878545745105873, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4702 + }, + { + "epoch": 1.8789452656811827, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4703 + }, + { + "epoch": 1.8793447862564923, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4704 + }, + { + "epoch": 1.8797443068318018, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4705 + }, + { + "epoch": 1.8801438274071116, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4706 + }, + { + "epoch": 1.880543347982421, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4707 + }, + { + "epoch": 1.8809428685577307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4708 + }, + { + "epoch": 1.8813423891330403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4709 + }, + { + "epoch": 1.8817419097083499, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4710 + }, + { + "epoch": 1.8821414302836597, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4711 + }, + { + "epoch": 1.8825409508589692, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4712 + }, + { + "epoch": 1.8829404714342788, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4713 + }, + { + "epoch": 1.8833399920095886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4714 + }, + { + "epoch": 1.8837395125848981, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4715 + }, + { + "epoch": 1.8841390331602077, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4716 + }, + { + "epoch": 1.8845385537355175, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4717 + }, + { + "epoch": 1.884938074310827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4718 + }, + { + "epoch": 1.8853375948861366, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4719 + }, + { + "epoch": 1.8857371154614464, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4720 + }, + { + "epoch": 1.8861366360367557, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4721 + }, + { + "epoch": 1.8865361566120655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4722 + }, + { + "epoch": 1.8869356771873753, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4723 + }, + { + "epoch": 1.8873351977626847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4724 + }, + { + "epoch": 1.8877347183379944, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4725 + }, + { + "epoch": 1.888134238913304, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4726 + }, + { + "epoch": 1.8885337594886136, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4727 + }, + { + "epoch": 1.8889332800639234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4728 + }, + { + "epoch": 1.889332800639233, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4729 + }, + { + "epoch": 1.8897323212145425, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4730 + }, + { + "epoch": 1.8901318417898523, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4731 + }, + { + "epoch": 1.8905313623651618, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4732 + }, + { + "epoch": 1.8909308829404714, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 4733 + }, + { + "epoch": 1.8913304035157812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4734 + }, + { + "epoch": 1.8917299240910905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4735 + }, + { + "epoch": 1.8921294446664003, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 4736 + }, + { + "epoch": 1.89252896524171, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4737 + }, + { + "epoch": 1.8929284858170194, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4738 + }, + { + "epoch": 1.8933280063923292, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4739 + }, + { + "epoch": 1.8937275269676388, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4740 + }, + { + "epoch": 1.8941270475429484, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4741 + }, + { + "epoch": 1.8945265681182581, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4742 + }, + { + "epoch": 1.8949260886935677, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4743 + }, + { + "epoch": 1.8953256092688773, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4744 + }, + { + "epoch": 1.895725129844187, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4745 + }, + { + "epoch": 1.8961246504194966, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4746 + }, + { + "epoch": 1.8965241709948062, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4747 + }, + { + "epoch": 1.896923691570116, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4748 + }, + { + "epoch": 1.8973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4749 + }, + { + "epoch": 1.897722732720735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4750 + }, + { + "epoch": 1.898122253296045, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4751 + }, + { + "epoch": 1.8985217738713542, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 4752 + }, + { + "epoch": 1.898921294446664, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4753 + }, + { + "epoch": 1.8993208150219736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4754 + }, + { + "epoch": 1.8997203355972831, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4755 + }, + { + "epoch": 1.900119856172593, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4756 + }, + { + "epoch": 1.9005193767479025, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4757 + }, + { + "epoch": 1.900918897323212, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4758 + }, + { + "epoch": 1.9013184178985219, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4759 + }, + { + "epoch": 1.9017179384738314, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4760 + }, + { + "epoch": 1.902117459049141, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4761 + }, + { + "epoch": 1.9025169796244508, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4762 + }, + { + "epoch": 1.9029165001997603, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4763 + }, + { + "epoch": 1.90331602077507, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4764 + }, + { + "epoch": 1.9037155413503797, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4765 + }, + { + "epoch": 1.904115061925689, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4766 + }, + { + "epoch": 1.9045145825009988, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4767 + }, + { + "epoch": 1.9049141030763086, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4768 + }, + { + "epoch": 1.905313623651618, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4769 + }, + { + "epoch": 1.9057131442269277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4770 + }, + { + "epoch": 1.9061126648022373, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4771 + }, + { + "epoch": 1.9065121853775469, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4772 + }, + { + "epoch": 1.9069117059528566, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4773 + }, + { + "epoch": 1.9073112265281662, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4774 + }, + { + "epoch": 1.9077107471034758, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4775 + }, + { + "epoch": 1.9081102676787856, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4776 + }, + { + "epoch": 1.9085097882540951, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4777 + }, + { + "epoch": 1.9089093088294047, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4778 + }, + { + "epoch": 1.9093088294047145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4779 + }, + { + "epoch": 1.9097083499800238, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4780 + }, + { + "epoch": 1.9101078705553336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4781 + }, + { + "epoch": 1.9105073911306434, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4782 + }, + { + "epoch": 1.9109069117059527, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4783 + }, + { + "epoch": 1.9113064322812625, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4784 + }, + { + "epoch": 1.911705952856572, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4785 + }, + { + "epoch": 1.9121054734318816, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4786 + }, + { + "epoch": 1.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4787 + }, + { + "epoch": 1.912904514582501, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4788 + }, + { + "epoch": 1.9133040351578106, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4789 + }, + { + "epoch": 1.9137035557331203, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 4790 + }, + { + "epoch": 1.91410307630843, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4791 + }, + { + "epoch": 1.9145025968837395, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4792 + }, + { + "epoch": 1.9149021174590493, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4793 + }, + { + "epoch": 1.9153016380343588, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4794 + }, + { + "epoch": 1.9157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4795 + }, + { + "epoch": 1.9161006791849782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4796 + }, + { + "epoch": 1.9165001997602875, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4797 + }, + { + "epoch": 1.9168997203355973, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4798 + }, + { + "epoch": 1.9172992409109069, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4799 + }, + { + "epoch": 1.9176987614862164, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 4800 + }, + { + "epoch": 1.9180982820615262, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4801 + }, + { + "epoch": 1.9184978026368358, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4802 + }, + { + "epoch": 1.9188973232121453, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 4803 + }, + { + "epoch": 1.9192968437874551, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 4804 + }, + { + "epoch": 1.9196963643627647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4805 + }, + { + "epoch": 1.9200958849380743, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4806 + }, + { + "epoch": 1.920495405513384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4807 + }, + { + "epoch": 1.9208949260886936, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4808 + }, + { + "epoch": 1.9212944466640032, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4809 + }, + { + "epoch": 1.921693967239313, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4810 + }, + { + "epoch": 1.9220934878146223, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4811 + }, + { + "epoch": 1.922493008389932, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4812 + }, + { + "epoch": 1.9228925289652419, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4813 + }, + { + "epoch": 1.9232920495405512, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4814 + }, + { + "epoch": 1.923691570115861, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 4815 + }, + { + "epoch": 1.9240910906911706, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4816 + }, + { + "epoch": 1.9244906112664801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4817 + }, + { + "epoch": 1.92489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4818 + }, + { + "epoch": 1.9252896524170995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4819 + }, + { + "epoch": 1.925689172992409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4820 + }, + { + "epoch": 1.9260886935677188, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4821 + }, + { + "epoch": 1.9264882141430284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4822 + }, + { + "epoch": 1.926887734718338, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4823 + }, + { + "epoch": 1.9272872552936477, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4824 + }, + { + "epoch": 1.927686775868957, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4825 + }, + { + "epoch": 1.9280862964442669, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4826 + }, + { + "epoch": 1.9284858170195767, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4827 + }, + { + "epoch": 1.928885337594886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4828 + }, + { + "epoch": 1.9292848581701958, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4829 + }, + { + "epoch": 1.9296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4830 + }, + { + "epoch": 1.930083899320815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4831 + }, + { + "epoch": 1.9304834198961247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4832 + }, + { + "epoch": 1.9308829404714343, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4833 + }, + { + "epoch": 1.9312824610467438, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4834 + }, + { + "epoch": 1.9316819816220536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4835 + }, + { + "epoch": 1.9320815021973632, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4836 + }, + { + "epoch": 1.9324810227726728, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4837 + }, + { + "epoch": 1.9328805433479825, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4838 + }, + { + "epoch": 1.933280063923292, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4839 + }, + { + "epoch": 1.9336795844986017, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4840 + }, + { + "epoch": 1.9340791050739115, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4841 + }, + { + "epoch": 1.9344786256492208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4842 + }, + { + "epoch": 1.9348781462245306, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4843 + }, + { + "epoch": 1.9352776667998401, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4844 + }, + { + "epoch": 1.9356771873751497, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4845 + }, + { + "epoch": 1.9360767079504595, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4846 + }, + { + "epoch": 1.936476228525769, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4847 + }, + { + "epoch": 1.9368757491010786, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4848 + }, + { + "epoch": 1.9372752696763884, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4849 + }, + { + "epoch": 1.937674790251698, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4850 + }, + { + "epoch": 1.9380743108270075, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 4851 + }, + { + "epoch": 1.9384738314023173, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4852 + }, + { + "epoch": 1.938873351977627, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4853 + }, + { + "epoch": 1.9392728725529365, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4854 + }, + { + "epoch": 1.9396723931282462, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4855 + }, + { + "epoch": 1.9400719137035556, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4856 + }, + { + "epoch": 1.9404714342788654, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4857 + }, + { + "epoch": 1.940870954854175, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4858 + }, + { + "epoch": 1.9412704754294845, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4859 + }, + { + "epoch": 1.9416699960047943, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4860 + }, + { + "epoch": 1.9420695165801038, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4861 + }, + { + "epoch": 1.9424690371554134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4862 + }, + { + "epoch": 1.9428685577307232, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4863 + }, + { + "epoch": 1.9432680783060328, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4864 + }, + { + "epoch": 1.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4865 + }, + { + "epoch": 1.9440671194566521, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4866 + }, + { + "epoch": 1.9444666400319617, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4867 + }, + { + "epoch": 1.9448661606072712, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4868 + }, + { + "epoch": 1.945265681182581, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4869 + }, + { + "epoch": 1.9456652017578904, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4870 + }, + { + "epoch": 1.9460647223332002, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4871 + }, + { + "epoch": 1.94646424290851, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4872 + }, + { + "epoch": 1.9468637634838193, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4873 + }, + { + "epoch": 1.947263284059129, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4874 + }, + { + "epoch": 1.9476628046344386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4875 + }, + { + "epoch": 1.9480623252097482, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4876 + }, + { + "epoch": 1.948461845785058, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4877 + }, + { + "epoch": 1.9488613663603676, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4878 + }, + { + "epoch": 1.9492608869356771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4879 + }, + { + "epoch": 1.949660407510987, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4880 + }, + { + "epoch": 1.9500599280862965, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4881 + }, + { + "epoch": 1.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4882 + }, + { + "epoch": 1.9508589692369158, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4883 + }, + { + "epoch": 1.9512584898122254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4884 + }, + { + "epoch": 1.951658010387535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4885 + }, + { + "epoch": 1.9520575309628447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4886 + }, + { + "epoch": 1.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4887 + }, + { + "epoch": 1.9528565721134639, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4888 + }, + { + "epoch": 1.9532560926887734, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4889 + }, + { + "epoch": 1.953655613264083, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 4890 + }, + { + "epoch": 1.9540551338393928, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4891 + }, + { + "epoch": 1.9544546544147023, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4892 + }, + { + "epoch": 1.954854174990012, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4893 + }, + { + "epoch": 1.9552536955653217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4894 + }, + { + "epoch": 1.9556532161406313, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4895 + }, + { + "epoch": 1.9560527367159408, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4896 + }, + { + "epoch": 1.9564522572912506, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4897 + }, + { + "epoch": 1.9568517778665602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4898 + }, + { + "epoch": 1.9572512984418697, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4899 + }, + { + "epoch": 1.9576508190171795, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4900 + }, + { + "epoch": 1.9580503395924889, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4901 + }, + { + "epoch": 1.9584498601677987, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4902 + }, + { + "epoch": 1.9588493807431082, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4903 + }, + { + "epoch": 1.9592489013184178, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4904 + }, + { + "epoch": 1.9596484218937276, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4905 + }, + { + "epoch": 1.9600479424690371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4906 + }, + { + "epoch": 1.9604474630443467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4907 + }, + { + "epoch": 1.9608469836196565, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4908 + }, + { + "epoch": 1.961246504194966, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4909 + }, + { + "epoch": 1.9616460247702756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4910 + }, + { + "epoch": 1.9620455453455854, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4911 + }, + { + "epoch": 1.962445065920895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4912 + }, + { + "epoch": 1.9628445864962045, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 4913 + }, + { + "epoch": 1.9632441070715143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4914 + }, + { + "epoch": 1.9636436276468237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 4915 + }, + { + "epoch": 1.9640431482221334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4916 + }, + { + "epoch": 1.9644426687974432, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 4917 + }, + { + "epoch": 1.9648421893727526, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4918 + }, + { + "epoch": 1.9652417099480624, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4919 + }, + { + "epoch": 1.965641230523372, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4920 + }, + { + "epoch": 1.9660407510986815, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4921 + }, + { + "epoch": 1.9664402716739913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4922 + }, + { + "epoch": 1.9668397922493008, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4923 + }, + { + "epoch": 1.9672393128246104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4924 + }, + { + "epoch": 1.9676388333999202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 4925 + }, + { + "epoch": 1.9680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4926 + }, + { + "epoch": 1.9684378745505393, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4927 + }, + { + "epoch": 1.968837395125849, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4928 + }, + { + "epoch": 1.9692369157011587, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4929 + }, + { + "epoch": 1.9696364362764682, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4930 + }, + { + "epoch": 1.970035956851778, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4931 + }, + { + "epoch": 1.9704354774270874, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4932 + }, + { + "epoch": 1.9708349980023971, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4933 + }, + { + "epoch": 1.9712345185777067, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 4934 + }, + { + "epoch": 1.9716340391530163, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4935 + }, + { + "epoch": 1.972033559728326, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4936 + }, + { + "epoch": 1.9724330803036356, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4937 + }, + { + "epoch": 1.9728326008789452, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4938 + }, + { + "epoch": 1.973232121454255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4939 + }, + { + "epoch": 1.9736316420295645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4940 + }, + { + "epoch": 1.974031162604874, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4941 + }, + { + "epoch": 1.9744306831801839, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4942 + }, + { + "epoch": 1.9748302037554935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4943 + }, + { + "epoch": 1.975229724330803, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4944 + }, + { + "epoch": 1.9756292449061128, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4945 + }, + { + "epoch": 1.9760287654814221, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4946 + }, + { + "epoch": 1.976428286056732, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4947 + }, + { + "epoch": 1.9768278066320415, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4948 + }, + { + "epoch": 1.977227327207351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4949 + }, + { + "epoch": 1.9776268477826608, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4950 + }, + { + "epoch": 1.9780263683579704, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4951 + }, + { + "epoch": 1.97842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 4952 + }, + { + "epoch": 1.9788254095085898, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4953 + }, + { + "epoch": 1.9792249300838993, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4954 + }, + { + "epoch": 1.979624450659209, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4955 + }, + { + "epoch": 1.9800239712345187, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4956 + }, + { + "epoch": 1.9804234918098282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4957 + }, + { + "epoch": 1.9808230123851378, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4958 + }, + { + "epoch": 1.9812225329604476, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4959 + }, + { + "epoch": 1.981622053535757, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4960 + }, + { + "epoch": 1.9820215741110667, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4961 + }, + { + "epoch": 1.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4962 + }, + { + "epoch": 1.9828206152616858, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 4963 + }, + { + "epoch": 1.9832201358369956, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4964 + }, + { + "epoch": 1.9836196564123052, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4965 + }, + { + "epoch": 1.9840191769876148, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4966 + }, + { + "epoch": 1.9844186975629245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4967 + }, + { + "epoch": 1.9848182181382341, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4968 + }, + { + "epoch": 1.9852177387135437, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4969 + }, + { + "epoch": 1.9856172592888535, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4970 + }, + { + "epoch": 1.986016779864163, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4971 + }, + { + "epoch": 1.9864163004394726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4972 + }, + { + "epoch": 1.9868158210147824, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4973 + }, + { + "epoch": 1.987215341590092, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4974 + }, + { + "epoch": 1.9876148621654015, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4975 + }, + { + "epoch": 1.9880143827407113, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4976 + }, + { + "epoch": 1.9884139033160206, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4977 + }, + { + "epoch": 1.9888134238913304, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4978 + }, + { + "epoch": 1.98921294446664, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4979 + }, + { + "epoch": 1.9896124650419496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4980 + }, + { + "epoch": 1.9900119856172593, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4981 + }, + { + "epoch": 1.990411506192569, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4982 + }, + { + "epoch": 1.9908110267678785, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4983 + }, + { + "epoch": 1.9912105473431883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4984 + }, + { + "epoch": 1.9916100679184978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4985 + }, + { + "epoch": 1.9920095884938074, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4986 + }, + { + "epoch": 1.9924091090691172, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4987 + }, + { + "epoch": 1.9928086296444267, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4988 + }, + { + "epoch": 1.9932081502197363, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4989 + }, + { + "epoch": 1.993607670795046, + "grad_norm": 0.04541015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4990 + }, + { + "epoch": 1.9940071913703554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4991 + }, + { + "epoch": 1.9944067119456652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4992 + }, + { + "epoch": 1.9948062325209748, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 4993 + }, + { + "epoch": 1.9952057530962843, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4994 + }, + { + "epoch": 1.9956052736715941, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4995 + }, + { + "epoch": 1.9960047942469037, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4996 + }, + { + "epoch": 1.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4997 + }, + { + "epoch": 1.996803835397523, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4998 + }, + { + "epoch": 1.9972033559728326, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4999 + }, + { + "epoch": 1.9976028765481422, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5000 + }, + { + "epoch": 1.998002397123452, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5001 + }, + { + "epoch": 1.9984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5002 + }, + { + "epoch": 1.998801438274071, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5003 + }, + { + "epoch": 1.9992009588493809, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5004 + }, + { + "epoch": 1.9996004794246902, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5005 + }, + { + "epoch": 2.0, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5006 + }, + { + "epoch": 2.00039952057531, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5007 + }, + { + "epoch": 2.000799041150619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5008 + }, + { + "epoch": 2.001198561725929, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5009 + }, + { + "epoch": 2.0015980823012387, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5010 + }, + { + "epoch": 2.001997602876548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5011 + }, + { + "epoch": 2.002397123451858, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5012 + }, + { + "epoch": 2.0027966440271676, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5013 + }, + { + "epoch": 2.003196164602477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 5014 + }, + { + "epoch": 2.0035956851777867, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5015 + }, + { + "epoch": 2.003995205753096, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5016 + }, + { + "epoch": 2.004394726328406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5017 + }, + { + "epoch": 2.0047942469037157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5018 + }, + { + "epoch": 2.005193767479025, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5019 + }, + { + "epoch": 2.005593288054335, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5020 + }, + { + "epoch": 2.0059928086296446, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5021 + }, + { + "epoch": 2.006392329204954, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5022 + }, + { + "epoch": 2.0067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5023 + }, + { + "epoch": 2.0071913703555735, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 5024 + }, + { + "epoch": 2.007590890930883, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5025 + }, + { + "epoch": 2.0079904115061926, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5026 + }, + { + "epoch": 2.0083899320815024, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5027 + }, + { + "epoch": 2.0087894526568117, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 5028 + }, + { + "epoch": 2.0091889732321215, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 5029 + }, + { + "epoch": 2.009588493807431, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5030 + }, + { + "epoch": 2.0099880143827407, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5031 + }, + { + "epoch": 2.0103875349580504, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5032 + }, + { + "epoch": 2.01078705553336, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5033 + }, + { + "epoch": 2.0111865761086696, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5034 + }, + { + "epoch": 2.0115860966839794, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5035 + }, + { + "epoch": 2.0119856172592887, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5036 + }, + { + "epoch": 2.0123851378345985, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5037 + }, + { + "epoch": 2.0127846584099083, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5038 + }, + { + "epoch": 2.0131841789852176, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5039 + }, + { + "epoch": 2.0135836995605274, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5040 + }, + { + "epoch": 2.013983220135837, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 5041 + }, + { + "epoch": 2.0143827407111465, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 5042 + }, + { + "epoch": 2.0147822612864563, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 5043 + }, + { + "epoch": 2.0151817818617657, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5044 + }, + { + "epoch": 2.0155813024370755, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5045 + }, + { + "epoch": 2.0159808230123852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5046 + }, + { + "epoch": 2.0163803435876946, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5047 + }, + { + "epoch": 2.0167798641630044, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5048 + }, + { + "epoch": 2.017179384738314, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5049 + }, + { + "epoch": 2.0175789053136235, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5050 + }, + { + "epoch": 2.0179784258889333, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5051 + }, + { + "epoch": 2.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5052 + }, + { + "epoch": 2.0187774670395524, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5053 + }, + { + "epoch": 2.019176987614862, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5054 + }, + { + "epoch": 2.019576508190172, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5055 + }, + { + "epoch": 2.0199760287654813, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5056 + }, + { + "epoch": 2.020375549340791, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5057 + }, + { + "epoch": 2.020775069916101, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5058 + }, + { + "epoch": 2.0211745904914102, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5059 + }, + { + "epoch": 2.02157411106672, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5060 + }, + { + "epoch": 2.0219736316420294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5061 + }, + { + "epoch": 2.022373152217339, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5062 + }, + { + "epoch": 2.022772672792649, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5063 + }, + { + "epoch": 2.0231721933679583, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5064 + }, + { + "epoch": 2.023571713943268, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5065 + }, + { + "epoch": 2.023971234518578, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5066 + }, + { + "epoch": 2.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5067 + }, + { + "epoch": 2.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5068 + }, + { + "epoch": 2.0251697962445068, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5069 + }, + { + "epoch": 2.025569316819816, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5070 + }, + { + "epoch": 2.025968837395126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5071 + }, + { + "epoch": 2.0263683579704357, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5072 + }, + { + "epoch": 2.026767878545745, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5073 + }, + { + "epoch": 2.027167399121055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5074 + }, + { + "epoch": 2.027566919696364, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5075 + }, + { + "epoch": 2.027966440271674, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5076 + }, + { + "epoch": 2.0283659608469837, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5077 + }, + { + "epoch": 2.028765481422293, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5078 + }, + { + "epoch": 2.029165001997603, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5079 + }, + { + "epoch": 2.0295645225729126, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5080 + }, + { + "epoch": 2.029964043148222, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5081 + }, + { + "epoch": 2.0303635637235318, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5082 + }, + { + "epoch": 2.0307630842988416, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5083 + }, + { + "epoch": 2.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5084 + }, + { + "epoch": 2.0315621254494607, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 5085 + }, + { + "epoch": 2.0319616460247705, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5086 + }, + { + "epoch": 2.03236116660008, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5087 + }, + { + "epoch": 2.0327606871753896, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5088 + }, + { + "epoch": 2.033160207750699, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5089 + }, + { + "epoch": 2.0335597283260087, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5090 + }, + { + "epoch": 2.0339592489013185, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5091 + }, + { + "epoch": 2.034358769476628, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5092 + }, + { + "epoch": 2.0347582900519376, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5093 + }, + { + "epoch": 2.0351578106272474, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5094 + }, + { + "epoch": 2.0355573312025568, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5095 + }, + { + "epoch": 2.0359568517778666, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5096 + }, + { + "epoch": 2.0363563723531763, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5097 + }, + { + "epoch": 2.0367558929284857, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5098 + }, + { + "epoch": 2.0371554135037955, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5099 + }, + { + "epoch": 2.0375549340791053, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 5100 + }, + { + "epoch": 2.0379544546544146, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5101 + }, + { + "epoch": 2.0383539752297244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5102 + }, + { + "epoch": 2.038753495805034, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5103 + }, + { + "epoch": 2.0391530163803435, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5104 + }, + { + "epoch": 2.0395525369556533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5105 + }, + { + "epoch": 2.0399520575309626, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5106 + }, + { + "epoch": 2.0403515781062724, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5107 + }, + { + "epoch": 2.040751098681582, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 5108 + }, + { + "epoch": 2.0411506192568916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 5109 + }, + { + "epoch": 2.0415501398322013, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5110 + }, + { + "epoch": 2.041949660407511, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5111 + }, + { + "epoch": 2.0423491809828205, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5112 + }, + { + "epoch": 2.0427487015581303, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5113 + }, + { + "epoch": 2.04314822213344, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5114 + }, + { + "epoch": 2.0435477427087494, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5115 + }, + { + "epoch": 2.043947263284059, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 5116 + }, + { + "epoch": 2.044346783859369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5117 + }, + { + "epoch": 2.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5118 + }, + { + "epoch": 2.045145825009988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5119 + }, + { + "epoch": 2.0455453455852974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 5120 + }, + { + "epoch": 2.0459448661606072, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5121 + }, + { + "epoch": 2.046344386735917, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5122 + }, + { + "epoch": 2.0467439073112264, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5123 + }, + { + "epoch": 2.047143427886536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5124 + }, + { + "epoch": 2.047542948461846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5125 + }, + { + "epoch": 2.0479424690371553, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5126 + }, + { + "epoch": 2.048341989612465, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5127 + }, + { + "epoch": 2.048741510187775, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5128 + }, + { + "epoch": 2.049141030763084, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5129 + }, + { + "epoch": 2.049540551338394, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5130 + }, + { + "epoch": 2.0499400719137038, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5131 + }, + { + "epoch": 2.050339592489013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5132 + }, + { + "epoch": 2.050739113064323, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5133 + }, + { + "epoch": 2.0511386336396322, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5134 + }, + { + "epoch": 2.051538154214942, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5135 + }, + { + "epoch": 2.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 5136 + }, + { + "epoch": 2.052337195365561, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5137 + }, + { + "epoch": 2.052736715940871, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 5138 + }, + { + "epoch": 2.0531362365161807, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5139 + }, + { + "epoch": 2.05353575709149, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5140 + }, + { + "epoch": 2.0539352776668, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5141 + }, + { + "epoch": 2.0543347982421096, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5142 + }, + { + "epoch": 2.054734318817419, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5143 + }, + { + "epoch": 2.0551338393927288, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5144 + }, + { + "epoch": 2.0555333599680385, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 5145 + }, + { + "epoch": 2.055932880543348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5146 + }, + { + "epoch": 2.0563324011186577, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5147 + }, + { + "epoch": 2.0567319216939675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5148 + }, + { + "epoch": 2.057131442269277, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5149 + }, + { + "epoch": 2.0575309628445866, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5150 + }, + { + "epoch": 2.057930483419896, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5151 + }, + { + "epoch": 2.0583300039952057, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5152 + }, + { + "epoch": 2.0587295245705155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5153 + }, + { + "epoch": 2.059129045145825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5154 + }, + { + "epoch": 2.0595285657211346, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5155 + }, + { + "epoch": 2.0599280862964444, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5156 + }, + { + "epoch": 2.0603276068717538, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5157 + }, + { + "epoch": 2.0607271274470635, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5158 + }, + { + "epoch": 2.0611266480223733, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5159 + }, + { + "epoch": 2.0615261685976827, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5160 + }, + { + "epoch": 2.0619256891729925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5161 + }, + { + "epoch": 2.0623252097483022, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5162 + }, + { + "epoch": 2.0627247303236116, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5163 + }, + { + "epoch": 2.0631242508989214, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5164 + }, + { + "epoch": 2.0635237714742307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5165 + }, + { + "epoch": 2.0639232920495405, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5166 + }, + { + "epoch": 2.0643228126248503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5167 + }, + { + "epoch": 2.0647223332001596, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5168 + }, + { + "epoch": 2.0651218537754694, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5169 + }, + { + "epoch": 2.065521374350779, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5170 + }, + { + "epoch": 2.0659208949260885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5171 + }, + { + "epoch": 2.0663204155013983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5172 + }, + { + "epoch": 2.066719936076708, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5173 + }, + { + "epoch": 2.0671194566520175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 5174 + }, + { + "epoch": 2.0675189772273272, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5175 + }, + { + "epoch": 2.067918497802637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5176 + }, + { + "epoch": 2.0683180183779464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 5177 + }, + { + "epoch": 2.068717538953256, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5178 + }, + { + "epoch": 2.0691170595285655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5179 + }, + { + "epoch": 2.0695165801038753, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5180 + }, + { + "epoch": 2.069916100679185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5181 + }, + { + "epoch": 2.0703156212544944, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5182 + }, + { + "epoch": 2.070715141829804, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5183 + }, + { + "epoch": 2.071114662405114, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5184 + }, + { + "epoch": 2.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5185 + }, + { + "epoch": 2.071913703555733, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5186 + }, + { + "epoch": 2.072313224131043, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5187 + }, + { + "epoch": 2.0727127447063523, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5188 + }, + { + "epoch": 2.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5189 + }, + { + "epoch": 2.073511785856972, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 5190 + }, + { + "epoch": 2.073911306432281, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5191 + }, + { + "epoch": 2.074310827007591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5192 + }, + { + "epoch": 2.0747103475829007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5193 + }, + { + "epoch": 2.07510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5194 + }, + { + "epoch": 2.07550938873352, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5195 + }, + { + "epoch": 2.075908909308829, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5196 + }, + { + "epoch": 2.076308429884139, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5197 + }, + { + "epoch": 2.076707950459449, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5198 + }, + { + "epoch": 2.077107471034758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5199 + }, + { + "epoch": 2.077506991610068, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5200 + }, + { + "epoch": 2.0779065121853777, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5201 + }, + { + "epoch": 2.078306032760687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5202 + }, + { + "epoch": 2.078705553335997, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5203 + }, + { + "epoch": 2.0791050739113066, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5204 + }, + { + "epoch": 2.079504594486616, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 5205 + }, + { + "epoch": 2.0799041150619257, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5206 + }, + { + "epoch": 2.0803036356372355, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5207 + }, + { + "epoch": 2.080703156212545, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5208 + }, + { + "epoch": 2.0811026767878547, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5209 + }, + { + "epoch": 2.081502197363164, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5210 + }, + { + "epoch": 2.081901717938474, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5211 + }, + { + "epoch": 2.0823012385137836, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5212 + }, + { + "epoch": 2.082700759089093, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5213 + }, + { + "epoch": 2.0831002796644027, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 5214 + }, + { + "epoch": 2.0834998002397125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5215 + }, + { + "epoch": 2.083899320815022, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5216 + }, + { + "epoch": 2.0842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5217 + }, + { + "epoch": 2.0846983619656414, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5218 + }, + { + "epoch": 2.0850978825409507, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5219 + }, + { + "epoch": 2.0854974031162605, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5220 + }, + { + "epoch": 2.0858969236915703, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5221 + }, + { + "epoch": 2.0862964442668797, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 5222 + }, + { + "epoch": 2.0866959648421894, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5223 + }, + { + "epoch": 2.087095485417499, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5224 + }, + { + "epoch": 2.0874950059928086, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5225 + }, + { + "epoch": 2.0878945265681184, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5226 + }, + { + "epoch": 2.0882940471434277, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5227 + }, + { + "epoch": 2.0886935677187375, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5228 + }, + { + "epoch": 2.0890930882940473, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5229 + }, + { + "epoch": 2.0894926088693566, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5230 + }, + { + "epoch": 2.0898921294446664, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5231 + }, + { + "epoch": 2.090291650019976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5232 + }, + { + "epoch": 2.0906911705952855, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5233 + }, + { + "epoch": 2.0910906911705953, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5234 + }, + { + "epoch": 2.091490211745905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5235 + }, + { + "epoch": 2.0918897323212144, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5236 + }, + { + "epoch": 2.0922892528965242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5237 + }, + { + "epoch": 2.092688773471834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5238 + }, + { + "epoch": 2.0930882940471434, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5239 + }, + { + "epoch": 2.093487814622453, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5240 + }, + { + "epoch": 2.0938873351977625, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5241 + }, + { + "epoch": 2.0942868557730723, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5242 + }, + { + "epoch": 2.094686376348382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5243 + }, + { + "epoch": 2.0950858969236914, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5244 + }, + { + "epoch": 2.095485417499001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5245 + }, + { + "epoch": 2.095884938074311, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5246 + }, + { + "epoch": 2.0962844586496203, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5247 + }, + { + "epoch": 2.09668397922493, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5248 + }, + { + "epoch": 2.09708349980024, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5249 + }, + { + "epoch": 2.0974830203755492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5250 + }, + { + "epoch": 2.097882540950859, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5251 + }, + { + "epoch": 2.098282061526169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 5252 + }, + { + "epoch": 2.098681582101478, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5253 + }, + { + "epoch": 2.099081102676788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5254 + }, + { + "epoch": 2.0994806232520973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5255 + }, + { + "epoch": 2.099880143827407, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5256 + }, + { + "epoch": 2.100279664402717, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5257 + }, + { + "epoch": 2.100679184978026, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5258 + }, + { + "epoch": 2.101078705553336, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5259 + }, + { + "epoch": 2.1014782261286458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5260 + }, + { + "epoch": 2.101877746703955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5261 + }, + { + "epoch": 2.102277267279265, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5262 + }, + { + "epoch": 2.1026767878545747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5263 + }, + { + "epoch": 2.103076308429884, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5264 + }, + { + "epoch": 2.103475829005194, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5265 + }, + { + "epoch": 2.1038753495805036, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5266 + }, + { + "epoch": 2.104274870155813, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5267 + }, + { + "epoch": 2.1046743907311227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5268 + }, + { + "epoch": 2.105073911306432, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5269 + }, + { + "epoch": 2.105473431881742, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5270 + }, + { + "epoch": 2.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5271 + }, + { + "epoch": 2.106272473032361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5272 + }, + { + "epoch": 2.1066719936076708, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5273 + }, + { + "epoch": 2.1070715141829806, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5274 + }, + { + "epoch": 2.10747103475829, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5275 + }, + { + "epoch": 2.1078705553335997, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5276 + }, + { + "epoch": 2.1082700759089095, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5277 + }, + { + "epoch": 2.108669596484219, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5278 + }, + { + "epoch": 2.1090691170595286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5279 + }, + { + "epoch": 2.1094686376348384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5280 + }, + { + "epoch": 2.1098681582101477, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5281 + }, + { + "epoch": 2.1102676787854575, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5282 + }, + { + "epoch": 2.1106671993607673, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5283 + }, + { + "epoch": 2.1110667199360766, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5284 + }, + { + "epoch": 2.1114662405113864, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5285 + }, + { + "epoch": 2.1118657610866958, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5286 + }, + { + "epoch": 2.1122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5287 + }, + { + "epoch": 2.1126648022373153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5288 + }, + { + "epoch": 2.1130643228126247, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5289 + }, + { + "epoch": 2.1134638433879345, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5290 + }, + { + "epoch": 2.1138633639632443, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5291 + }, + { + "epoch": 2.1142628845385536, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5292 + }, + { + "epoch": 2.1146624051138634, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5293 + }, + { + "epoch": 2.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5294 + }, + { + "epoch": 2.1154614462644825, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5295 + }, + { + "epoch": 2.1158609668397923, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5296 + }, + { + "epoch": 2.116260487415102, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 5297 + }, + { + "epoch": 2.1166600079904114, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5298 + }, + { + "epoch": 2.117059528565721, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5299 + }, + { + "epoch": 2.1174590491410306, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5300 + }, + { + "epoch": 2.1178585697163403, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5301 + }, + { + "epoch": 2.11825809029165, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5302 + }, + { + "epoch": 2.1186576108669595, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5303 + }, + { + "epoch": 2.1190571314422693, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5304 + }, + { + "epoch": 2.119456652017579, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5305 + }, + { + "epoch": 2.1198561725928884, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5306 + }, + { + "epoch": 2.120255693168198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5307 + }, + { + "epoch": 2.120655213743508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5308 + }, + { + "epoch": 2.1210547343188173, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5309 + }, + { + "epoch": 2.121454254894127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5310 + }, + { + "epoch": 2.121853775469437, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5311 + }, + { + "epoch": 2.122253296044746, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5312 + }, + { + "epoch": 2.122652816620056, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5313 + }, + { + "epoch": 2.1230523371953653, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5314 + }, + { + "epoch": 2.123451857770675, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5315 + }, + { + "epoch": 2.123851378345985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5316 + }, + { + "epoch": 2.1242508989212943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5317 + }, + { + "epoch": 2.124650419496604, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5318 + }, + { + "epoch": 2.125049940071914, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5319 + }, + { + "epoch": 2.125449460647223, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5320 + }, + { + "epoch": 2.125848981222533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5321 + }, + { + "epoch": 2.1262485017978427, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5322 + }, + { + "epoch": 2.126648022373152, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5323 + }, + { + "epoch": 2.127047542948462, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5324 + }, + { + "epoch": 2.1274470635237717, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5325 + }, + { + "epoch": 2.127846584099081, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5326 + }, + { + "epoch": 2.128246104674391, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5327 + }, + { + "epoch": 2.1286456252497006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5328 + }, + { + "epoch": 2.12904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5329 + }, + { + "epoch": 2.1294446664003197, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5330 + }, + { + "epoch": 2.129844186975629, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5331 + }, + { + "epoch": 2.130243707550939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5332 + }, + { + "epoch": 2.1306432281262486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5333 + }, + { + "epoch": 2.131042748701558, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5334 + }, + { + "epoch": 2.1314422692768678, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5335 + }, + { + "epoch": 2.1318417898521775, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5336 + }, + { + "epoch": 2.132241310427487, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5337 + }, + { + "epoch": 2.1326408310027967, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5338 + }, + { + "epoch": 2.1330403515781065, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5339 + }, + { + "epoch": 2.133439872153416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5340 + }, + { + "epoch": 2.1338393927287256, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5341 + }, + { + "epoch": 2.134238913304035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5342 + }, + { + "epoch": 2.1346384338793447, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5343 + }, + { + "epoch": 2.1350379544546545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5344 + }, + { + "epoch": 2.135437475029964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5345 + }, + { + "epoch": 2.1358369956052736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5346 + }, + { + "epoch": 2.1362365161805834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5347 + }, + { + "epoch": 2.1366360367558928, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5348 + }, + { + "epoch": 2.1370355573312025, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5349 + }, + { + "epoch": 2.1374350779065123, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5350 + }, + { + "epoch": 2.1378345984818217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 5351 + }, + { + "epoch": 2.1382341190571315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5352 + }, + { + "epoch": 2.1386336396324412, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5353 + }, + { + "epoch": 2.1390331602077506, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5354 + }, + { + "epoch": 2.1394326807830604, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5355 + }, + { + "epoch": 2.13983220135837, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5356 + }, + { + "epoch": 2.1402317219336795, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 5357 + }, + { + "epoch": 2.1406312425089893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5358 + }, + { + "epoch": 2.1410307630842986, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5359 + }, + { + "epoch": 2.1414302836596084, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5360 + }, + { + "epoch": 2.141829804234918, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5361 + }, + { + "epoch": 2.1422293248102275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5362 + }, + { + "epoch": 2.1426288453855373, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5363 + }, + { + "epoch": 2.143028365960847, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5364 + }, + { + "epoch": 2.1434278865361565, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5365 + }, + { + "epoch": 2.1438274071114662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5366 + }, + { + "epoch": 2.144226927686776, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5367 + }, + { + "epoch": 2.1446264482620854, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5368 + }, + { + "epoch": 2.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5369 + }, + { + "epoch": 2.145425489412705, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5370 + }, + { + "epoch": 2.1458250099880143, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5371 + }, + { + "epoch": 2.146224530563324, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5372 + }, + { + "epoch": 2.146624051138634, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5373 + }, + { + "epoch": 2.147023571713943, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5374 + }, + { + "epoch": 2.147423092289253, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5375 + }, + { + "epoch": 2.1478226128645623, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5376 + }, + { + "epoch": 2.148222133439872, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5377 + }, + { + "epoch": 2.148621654015182, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5378 + }, + { + "epoch": 2.1490211745904912, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5379 + }, + { + "epoch": 2.149420695165801, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5380 + }, + { + "epoch": 2.149820215741111, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5381 + }, + { + "epoch": 2.15021973631642, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5382 + }, + { + "epoch": 2.15061925689173, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5383 + }, + { + "epoch": 2.1510187774670397, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5384 + }, + { + "epoch": 2.151418298042349, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5385 + }, + { + "epoch": 2.151817818617659, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5386 + }, + { + "epoch": 2.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5387 + }, + { + "epoch": 2.152616859768278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5388 + }, + { + "epoch": 2.1530163803435878, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5389 + }, + { + "epoch": 2.153415900918897, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5390 + }, + { + "epoch": 2.153815421494207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5391 + }, + { + "epoch": 2.1542149420695167, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5392 + }, + { + "epoch": 2.154614462644826, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5393 + }, + { + "epoch": 2.155013983220136, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5394 + }, + { + "epoch": 2.1554135037954456, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5395 + }, + { + "epoch": 2.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5396 + }, + { + "epoch": 2.1562125449460647, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5397 + }, + { + "epoch": 2.1566120655213745, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5398 + }, + { + "epoch": 2.157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5399 + }, + { + "epoch": 2.1574111066719937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5400 + }, + { + "epoch": 2.1578106272473034, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5401 + }, + { + "epoch": 2.158210147822613, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5402 + }, + { + "epoch": 2.1586096683979226, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5403 + }, + { + "epoch": 2.159009188973232, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5404 + }, + { + "epoch": 2.1594087095485417, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5405 + }, + { + "epoch": 2.1598082301238515, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5406 + }, + { + "epoch": 2.160207750699161, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5407 + }, + { + "epoch": 2.1606072712744706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5408 + }, + { + "epoch": 2.1610067918497804, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5409 + }, + { + "epoch": 2.1614063124250897, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5410 + }, + { + "epoch": 2.1618058330003995, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5411 + }, + { + "epoch": 2.1622053535757093, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5412 + }, + { + "epoch": 2.1626048741510187, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5413 + }, + { + "epoch": 2.1630043947263284, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5414 + }, + { + "epoch": 2.1634039153016382, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5415 + }, + { + "epoch": 2.1638034358769476, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5416 + }, + { + "epoch": 2.1642029564522574, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5417 + }, + { + "epoch": 2.164602477027567, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5418 + }, + { + "epoch": 2.1650019976028765, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5419 + }, + { + "epoch": 2.1654015181781863, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5420 + }, + { + "epoch": 2.1658010387534956, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5421 + }, + { + "epoch": 2.1662005593288054, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5422 + }, + { + "epoch": 2.166600079904115, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5423 + }, + { + "epoch": 2.1669996004794245, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5424 + }, + { + "epoch": 2.1673991210547343, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5425 + }, + { + "epoch": 2.167798641630044, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5426 + }, + { + "epoch": 2.1681981622053534, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5427 + }, + { + "epoch": 2.1685976827806632, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5428 + }, + { + "epoch": 2.168997203355973, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5429 + }, + { + "epoch": 2.1693967239312824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5430 + }, + { + "epoch": 2.169796244506592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5431 + }, + { + "epoch": 2.1701957650819015, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5432 + }, + { + "epoch": 2.1705952856572113, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5433 + }, + { + "epoch": 2.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5434 + }, + { + "epoch": 2.1713943268078304, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5435 + }, + { + "epoch": 2.17179384738314, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5436 + }, + { + "epoch": 2.17219336795845, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 5437 + }, + { + "epoch": 2.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5438 + }, + { + "epoch": 2.172992409109069, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5439 + }, + { + "epoch": 2.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5440 + }, + { + "epoch": 2.1737914502596882, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5441 + }, + { + "epoch": 2.174190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5442 + }, + { + "epoch": 2.174590491410308, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5443 + }, + { + "epoch": 2.174990011985617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5444 + }, + { + "epoch": 2.175389532560927, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5445 + }, + { + "epoch": 2.1757890531362367, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5446 + }, + { + "epoch": 2.176188573711546, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5447 + }, + { + "epoch": 2.176588094286856, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5448 + }, + { + "epoch": 2.176987614862165, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5449 + }, + { + "epoch": 2.177387135437475, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5450 + }, + { + "epoch": 2.1777866560127848, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5451 + }, + { + "epoch": 2.178186176588094, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 5452 + }, + { + "epoch": 2.178585697163404, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5453 + }, + { + "epoch": 2.1789852177387137, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5454 + }, + { + "epoch": 2.179384738314023, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5455 + }, + { + "epoch": 2.179784258889333, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5456 + }, + { + "epoch": 2.1801837794646426, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5457 + }, + { + "epoch": 2.180583300039952, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5458 + }, + { + "epoch": 2.1809828206152617, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5459 + }, + { + "epoch": 2.1813823411905715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5460 + }, + { + "epoch": 2.181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5461 + }, + { + "epoch": 2.1821813823411906, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5462 + }, + { + "epoch": 2.1825809029165004, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5463 + }, + { + "epoch": 2.1829804234918098, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5464 + }, + { + "epoch": 2.1833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5465 + }, + { + "epoch": 2.183779464642429, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 5466 + }, + { + "epoch": 2.1841789852177387, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 5467 + }, + { + "epoch": 2.1845785057930485, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5468 + }, + { + "epoch": 2.184978026368358, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5469 + }, + { + "epoch": 2.1853775469436676, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5470 + }, + { + "epoch": 2.1857770675189774, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5471 + }, + { + "epoch": 2.1861765880942867, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5472 + }, + { + "epoch": 2.1865761086695965, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 5473 + }, + { + "epoch": 2.1869756292449063, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5474 + }, + { + "epoch": 2.1873751498202156, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5475 + }, + { + "epoch": 2.1877746703955254, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5476 + }, + { + "epoch": 2.1881741909708348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5477 + }, + { + "epoch": 2.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5478 + }, + { + "epoch": 2.1889732321214543, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5479 + }, + { + "epoch": 2.1893727526967637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5480 + }, + { + "epoch": 2.1897722732720735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 5481 + }, + { + "epoch": 2.1901717938473833, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5482 + }, + { + "epoch": 2.1905713144226926, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5483 + }, + { + "epoch": 2.1909708349980024, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5484 + }, + { + "epoch": 2.191370355573312, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5485 + }, + { + "epoch": 2.1917698761486215, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5486 + }, + { + "epoch": 2.1921693967239313, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 5487 + }, + { + "epoch": 2.192568917299241, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5488 + }, + { + "epoch": 2.1929684378745504, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 5489 + }, + { + "epoch": 2.19336795844986, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5490 + }, + { + "epoch": 2.19376747902517, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5491 + }, + { + "epoch": 2.1941669996004793, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5492 + }, + { + "epoch": 2.194566520175789, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5493 + }, + { + "epoch": 2.1949660407510985, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5494 + }, + { + "epoch": 2.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5495 + }, + { + "epoch": 2.195765081901718, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5496 + }, + { + "epoch": 2.1961646024770274, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5497 + }, + { + "epoch": 2.196564123052337, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5498 + }, + { + "epoch": 2.196963643627647, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5499 + }, + { + "epoch": 2.1973631642029563, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5500 + }, + { + "epoch": 2.197762684778266, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5501 + }, + { + "epoch": 2.198162205353576, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5502 + }, + { + "epoch": 2.198561725928885, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5503 + }, + { + "epoch": 2.198961246504195, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5504 + }, + { + "epoch": 2.199360767079505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5505 + }, + { + "epoch": 2.199760287654814, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5506 + }, + { + "epoch": 2.200159808230124, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5507 + }, + { + "epoch": 2.2005593288054337, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5508 + }, + { + "epoch": 2.200958849380743, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5509 + }, + { + "epoch": 2.201358369956053, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5510 + }, + { + "epoch": 2.201757890531362, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5511 + }, + { + "epoch": 2.202157411106672, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5512 + }, + { + "epoch": 2.2025569316819817, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5513 + }, + { + "epoch": 2.202956452257291, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5514 + }, + { + "epoch": 2.203355972832601, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5515 + }, + { + "epoch": 2.2037554934079107, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5516 + }, + { + "epoch": 2.20415501398322, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5517 + }, + { + "epoch": 2.20455453455853, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5518 + }, + { + "epoch": 2.2049540551338396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5519 + }, + { + "epoch": 2.205353575709149, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5520 + }, + { + "epoch": 2.2057530962844587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5521 + }, + { + "epoch": 2.206152616859768, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5522 + }, + { + "epoch": 2.206552137435078, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5523 + }, + { + "epoch": 2.2069516580103876, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5524 + }, + { + "epoch": 2.207351178585697, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5525 + }, + { + "epoch": 2.2077506991610067, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5526 + }, + { + "epoch": 2.2081502197363165, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5527 + }, + { + "epoch": 2.208549740311626, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5528 + }, + { + "epoch": 2.2089492608869357, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5529 + }, + { + "epoch": 2.2093487814622454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5530 + }, + { + "epoch": 2.209748302037555, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5531 + }, + { + "epoch": 2.2101478226128646, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5532 + }, + { + "epoch": 2.2105473431881744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5533 + }, + { + "epoch": 2.2109468637634837, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5534 + }, + { + "epoch": 2.2113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5535 + }, + { + "epoch": 2.2117459049141033, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5536 + }, + { + "epoch": 2.2121454254894126, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5537 + }, + { + "epoch": 2.2125449460647224, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5538 + }, + { + "epoch": 2.2129444666400317, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5539 + }, + { + "epoch": 2.2133439872153415, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5540 + }, + { + "epoch": 2.2137435077906513, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5541 + }, + { + "epoch": 2.2141430283659607, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5542 + }, + { + "epoch": 2.2145425489412705, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5543 + }, + { + "epoch": 2.2149420695165802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5544 + }, + { + "epoch": 2.2153415900918896, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5545 + }, + { + "epoch": 2.2157411106671994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5546 + }, + { + "epoch": 2.216140631242509, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5547 + }, + { + "epoch": 2.2165401518178185, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5548 + }, + { + "epoch": 2.2169396723931283, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5549 + }, + { + "epoch": 2.217339192968438, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5550 + }, + { + "epoch": 2.2177387135437474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5551 + }, + { + "epoch": 2.218138234119057, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5552 + }, + { + "epoch": 2.218537754694367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5553 + }, + { + "epoch": 2.2189372752696763, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 5554 + }, + { + "epoch": 2.219336795844986, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5555 + }, + { + "epoch": 2.2197363164202955, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5556 + }, + { + "epoch": 2.2201358369956052, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5557 + }, + { + "epoch": 2.220535357570915, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5558 + }, + { + "epoch": 2.2209348781462244, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5559 + }, + { + "epoch": 2.221334398721534, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5560 + }, + { + "epoch": 2.221733919296844, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5561 + }, + { + "epoch": 2.2221334398721533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5562 + }, + { + "epoch": 2.222532960447463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5563 + }, + { + "epoch": 2.222932481022773, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5564 + }, + { + "epoch": 2.223332001598082, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5565 + }, + { + "epoch": 2.223731522173392, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5566 + }, + { + "epoch": 2.2241310427487013, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5567 + }, + { + "epoch": 2.224530563324011, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5568 + }, + { + "epoch": 2.224930083899321, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5569 + }, + { + "epoch": 2.2253296044746302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5570 + }, + { + "epoch": 2.22572912504994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5571 + }, + { + "epoch": 2.22612864562525, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5572 + }, + { + "epoch": 2.226528166200559, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5573 + }, + { + "epoch": 2.226927686775869, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5574 + }, + { + "epoch": 2.2273272073511787, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5575 + }, + { + "epoch": 2.227726727926488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5576 + }, + { + "epoch": 2.228126248501798, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5577 + }, + { + "epoch": 2.2285257690771076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5578 + }, + { + "epoch": 2.228925289652417, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5579 + }, + { + "epoch": 2.2293248102277268, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5580 + }, + { + "epoch": 2.2297243308030366, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5581 + }, + { + "epoch": 2.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5582 + }, + { + "epoch": 2.2305233719536557, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5583 + }, + { + "epoch": 2.230922892528965, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5584 + }, + { + "epoch": 2.231322413104275, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5585 + }, + { + "epoch": 2.2317219336795846, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5586 + }, + { + "epoch": 2.232121454254894, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 5587 + }, + { + "epoch": 2.2325209748302037, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5588 + }, + { + "epoch": 2.2329204954055135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 5589 + }, + { + "epoch": 2.233320015980823, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5590 + }, + { + "epoch": 2.2337195365561326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5591 + }, + { + "epoch": 2.2341190571314424, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5592 + }, + { + "epoch": 2.2345185777067518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 5593 + }, + { + "epoch": 2.2349180982820616, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5594 + }, + { + "epoch": 2.2353176188573713, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5595 + }, + { + "epoch": 2.2357171394326807, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5596 + }, + { + "epoch": 2.2361166600079905, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5597 + }, + { + "epoch": 2.2365161805833003, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5598 + }, + { + "epoch": 2.2369157011586096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5599 + }, + { + "epoch": 2.2373152217339194, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5600 + }, + { + "epoch": 2.2377147423092287, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5601 + }, + { + "epoch": 2.2381142628845385, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5602 + }, + { + "epoch": 2.2385137834598483, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5603 + }, + { + "epoch": 2.2389133040351576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5604 + }, + { + "epoch": 2.2393128246104674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 5605 + }, + { + "epoch": 2.239712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5606 + }, + { + "epoch": 2.2401118657610866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5607 + }, + { + "epoch": 2.2405113863363963, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5608 + }, + { + "epoch": 2.240910906911706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5609 + }, + { + "epoch": 2.2413104274870155, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5610 + }, + { + "epoch": 2.2417099480623253, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5611 + }, + { + "epoch": 2.2421094686376346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5612 + }, + { + "epoch": 2.2425089892129444, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5613 + }, + { + "epoch": 2.242908509788254, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5614 + }, + { + "epoch": 2.2433080303635635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5615 + }, + { + "epoch": 2.2437075509388733, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5616 + }, + { + "epoch": 2.244107071514183, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5617 + }, + { + "epoch": 2.2445065920894924, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 5618 + }, + { + "epoch": 2.2449061126648022, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5619 + }, + { + "epoch": 2.245305633240112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5620 + }, + { + "epoch": 2.2457051538154214, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5621 + }, + { + "epoch": 2.246104674390731, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5622 + }, + { + "epoch": 2.246504194966041, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5623 + }, + { + "epoch": 2.2469037155413503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5624 + }, + { + "epoch": 2.24730323611666, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5625 + }, + { + "epoch": 2.24770275669197, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5626 + }, + { + "epoch": 2.248102277267279, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5627 + }, + { + "epoch": 2.248501797842589, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5628 + }, + { + "epoch": 2.2489013184178983, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5629 + }, + { + "epoch": 2.249300838993208, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5630 + }, + { + "epoch": 2.249700359568518, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5631 + }, + { + "epoch": 2.2500998801438272, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 5632 + }, + { + "epoch": 2.250499400719137, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5633 + }, + { + "epoch": 2.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5634 + }, + { + "epoch": 2.251298441869756, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5635 + }, + { + "epoch": 2.251697962445066, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5636 + }, + { + "epoch": 2.2520974830203757, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5637 + }, + { + "epoch": 2.252497003595685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 5638 + }, + { + "epoch": 2.252896524170995, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5639 + }, + { + "epoch": 2.2532960447463046, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 5640 + }, + { + "epoch": 2.253695565321614, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5641 + }, + { + "epoch": 2.2540950858969238, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5642 + }, + { + "epoch": 2.2544946064722335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5643 + }, + { + "epoch": 2.254894127047543, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5644 + }, + { + "epoch": 2.2552936476228527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5645 + }, + { + "epoch": 2.255693168198162, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5646 + }, + { + "epoch": 2.256092688773472, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5647 + }, + { + "epoch": 2.2564922093487816, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5648 + }, + { + "epoch": 2.256891729924091, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5649 + }, + { + "epoch": 2.2572912504994007, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5650 + }, + { + "epoch": 2.2576907710747105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 5651 + }, + { + "epoch": 2.25809029165002, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5652 + }, + { + "epoch": 2.2584898122253296, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5653 + }, + { + "epoch": 2.2588893328006394, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5654 + }, + { + "epoch": 2.2592888533759488, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5655 + }, + { + "epoch": 2.2596883739512585, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5656 + }, + { + "epoch": 2.260087894526568, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5657 + }, + { + "epoch": 2.2604874151018777, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5658 + }, + { + "epoch": 2.2608869356771875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5659 + }, + { + "epoch": 2.2612864562524972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5660 + }, + { + "epoch": 2.2616859768278066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5661 + }, + { + "epoch": 2.2620854974031164, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5662 + }, + { + "epoch": 2.2624850179784257, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5663 + }, + { + "epoch": 2.2628845385537355, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5664 + }, + { + "epoch": 2.2632840591290453, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5665 + }, + { + "epoch": 2.2636835797043546, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5666 + }, + { + "epoch": 2.2640831002796644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 5667 + }, + { + "epoch": 2.264482620854974, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5668 + }, + { + "epoch": 2.2648821414302835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5669 + }, + { + "epoch": 2.2652816620055933, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5670 + }, + { + "epoch": 2.265681182580903, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5671 + }, + { + "epoch": 2.2660807031562125, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5672 + }, + { + "epoch": 2.2664802237315222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5673 + }, + { + "epoch": 2.2668797443068316, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5674 + }, + { + "epoch": 2.2672792648821414, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5675 + }, + { + "epoch": 2.267678785457451, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5676 + }, + { + "epoch": 2.2680783060327605, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5677 + }, + { + "epoch": 2.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5678 + }, + { + "epoch": 2.26887734718338, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 5679 + }, + { + "epoch": 2.2692768677586894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5680 + }, + { + "epoch": 2.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5681 + }, + { + "epoch": 2.270075908909309, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5682 + }, + { + "epoch": 2.2704754294846183, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5683 + }, + { + "epoch": 2.270874950059928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5684 + }, + { + "epoch": 2.271274470635238, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5685 + }, + { + "epoch": 2.2716739912105472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5686 + }, + { + "epoch": 2.272073511785857, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5687 + }, + { + "epoch": 2.272473032361167, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5688 + }, + { + "epoch": 2.272872552936476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5689 + }, + { + "epoch": 2.273272073511786, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5690 + }, + { + "epoch": 2.2736715940870953, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5691 + }, + { + "epoch": 2.274071114662405, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5692 + }, + { + "epoch": 2.274470635237715, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5693 + }, + { + "epoch": 2.274870155813024, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5694 + }, + { + "epoch": 2.275269676388334, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5695 + }, + { + "epoch": 2.275669196963644, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 5696 + }, + { + "epoch": 2.276068717538953, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5697 + }, + { + "epoch": 2.276468238114263, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5698 + }, + { + "epoch": 2.2768677586895727, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5699 + }, + { + "epoch": 2.277267279264882, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 5700 + }, + { + "epoch": 2.277666799840192, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5701 + }, + { + "epoch": 2.278066320415501, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5702 + }, + { + "epoch": 2.278465840990811, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5703 + }, + { + "epoch": 2.2788653615661207, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5704 + }, + { + "epoch": 2.2792648821414305, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5705 + }, + { + "epoch": 2.27966440271674, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5706 + }, + { + "epoch": 2.2800639232920497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5707 + }, + { + "epoch": 2.280463443867359, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5708 + }, + { + "epoch": 2.280862964442669, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5709 + }, + { + "epoch": 2.2812624850179786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5710 + }, + { + "epoch": 2.281662005593288, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5711 + }, + { + "epoch": 2.2820615261685977, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5712 + }, + { + "epoch": 2.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5713 + }, + { + "epoch": 2.282860567319217, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5714 + }, + { + "epoch": 2.2832600878945266, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5715 + }, + { + "epoch": 2.2836596084698364, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5716 + }, + { + "epoch": 2.2840591290451457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5717 + }, + { + "epoch": 2.2844586496204555, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5718 + }, + { + "epoch": 2.284858170195765, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5719 + }, + { + "epoch": 2.2852576907710747, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5720 + }, + { + "epoch": 2.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5721 + }, + { + "epoch": 2.286056731921694, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5722 + }, + { + "epoch": 2.2864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5723 + }, + { + "epoch": 2.2868557730723134, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5724 + }, + { + "epoch": 2.2872552936476227, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5725 + }, + { + "epoch": 2.2876548142229325, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5726 + }, + { + "epoch": 2.2880543347982423, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5727 + }, + { + "epoch": 2.2884538553735516, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5728 + }, + { + "epoch": 2.2888533759488614, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5729 + }, + { + "epoch": 2.289252896524171, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5730 + }, + { + "epoch": 2.2896524170994805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5731 + }, + { + "epoch": 2.2900519376747903, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5732 + }, + { + "epoch": 2.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5733 + }, + { + "epoch": 2.2908509788254094, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5734 + }, + { + "epoch": 2.2912504994007192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 5735 + }, + { + "epoch": 2.2916500199760286, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 5736 + }, + { + "epoch": 2.2920495405513384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5737 + }, + { + "epoch": 2.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5738 + }, + { + "epoch": 2.2928485817019575, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5739 + }, + { + "epoch": 2.2932481022772673, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5740 + }, + { + "epoch": 2.293647622852577, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5741 + }, + { + "epoch": 2.2940471434278864, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5742 + }, + { + "epoch": 2.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5743 + }, + { + "epoch": 2.294846184578506, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5744 + }, + { + "epoch": 2.2952457051538153, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5745 + }, + { + "epoch": 2.295645225729125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5746 + }, + { + "epoch": 2.2960447463044344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5747 + }, + { + "epoch": 2.2964442668797442, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5748 + }, + { + "epoch": 2.296843787455054, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5749 + }, + { + "epoch": 2.297243308030364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5750 + }, + { + "epoch": 2.297642828605673, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5751 + }, + { + "epoch": 2.298042349180983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5752 + }, + { + "epoch": 2.2984418697562923, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5753 + }, + { + "epoch": 2.298841390331602, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5754 + }, + { + "epoch": 2.299240910906912, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5755 + }, + { + "epoch": 2.299640431482221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5756 + }, + { + "epoch": 2.300039952057531, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5757 + }, + { + "epoch": 2.3004394726328408, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5758 + }, + { + "epoch": 2.30083899320815, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 5759 + }, + { + "epoch": 2.30123851378346, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5760 + }, + { + "epoch": 2.3016380343587697, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5761 + }, + { + "epoch": 2.302037554934079, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5762 + }, + { + "epoch": 2.302437075509389, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5763 + }, + { + "epoch": 2.302836596084698, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5764 + }, + { + "epoch": 2.303236116660008, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5765 + }, + { + "epoch": 2.3036356372353177, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5766 + }, + { + "epoch": 2.304035157810627, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5767 + }, + { + "epoch": 2.304434678385937, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5768 + }, + { + "epoch": 2.3048341989612466, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5769 + }, + { + "epoch": 2.305233719536556, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5770 + }, + { + "epoch": 2.3056332401118658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5771 + }, + { + "epoch": 2.3060327606871756, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5772 + }, + { + "epoch": 2.306432281262485, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5773 + }, + { + "epoch": 2.3068318018377947, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5774 + }, + { + "epoch": 2.3072313224131045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5775 + }, + { + "epoch": 2.307630842988414, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5776 + }, + { + "epoch": 2.3080303635637236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5777 + }, + { + "epoch": 2.3084298841390334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5778 + }, + { + "epoch": 2.3088294047143427, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5779 + }, + { + "epoch": 2.3092289252896525, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5780 + }, + { + "epoch": 2.309628445864962, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5781 + }, + { + "epoch": 2.3100279664402716, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5782 + }, + { + "epoch": 2.3104274870155814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5783 + }, + { + "epoch": 2.3108270075908908, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5784 + }, + { + "epoch": 2.3112265281662006, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5785 + }, + { + "epoch": 2.3116260487415103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5786 + }, + { + "epoch": 2.3120255693168197, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5787 + }, + { + "epoch": 2.3124250898921295, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5788 + }, + { + "epoch": 2.3128246104674393, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5789 + }, + { + "epoch": 2.3132241310427486, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5790 + }, + { + "epoch": 2.3136236516180584, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5791 + }, + { + "epoch": 2.3140231721933677, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5792 + }, + { + "epoch": 2.3144226927686775, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5793 + }, + { + "epoch": 2.3148222133439873, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 5794 + }, + { + "epoch": 2.315221733919297, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5795 + }, + { + "epoch": 2.3156212544946064, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5796 + }, + { + "epoch": 2.316020775069916, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5797 + }, + { + "epoch": 2.3164202956452256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5798 + }, + { + "epoch": 2.3168198162205353, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5799 + }, + { + "epoch": 2.317219336795845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5800 + }, + { + "epoch": 2.3176188573711545, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5801 + }, + { + "epoch": 2.3180183779464643, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5802 + }, + { + "epoch": 2.318417898521774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5803 + }, + { + "epoch": 2.3188174190970834, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5804 + }, + { + "epoch": 2.319216939672393, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5805 + }, + { + "epoch": 2.319616460247703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5806 + }, + { + "epoch": 2.3200159808230123, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5807 + }, + { + "epoch": 2.320415501398322, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5808 + }, + { + "epoch": 2.3208150219736314, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5809 + }, + { + "epoch": 2.321214542548941, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5810 + }, + { + "epoch": 2.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5811 + }, + { + "epoch": 2.3220135836995603, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5812 + }, + { + "epoch": 2.32241310427487, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5813 + }, + { + "epoch": 2.32281262485018, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5814 + }, + { + "epoch": 2.3232121454254893, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5815 + }, + { + "epoch": 2.323611666000799, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5816 + }, + { + "epoch": 2.324011186576109, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5817 + }, + { + "epoch": 2.324410707151418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5818 + }, + { + "epoch": 2.324810227726728, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5819 + }, + { + "epoch": 2.3252097483020377, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5820 + }, + { + "epoch": 2.325609268877347, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5821 + }, + { + "epoch": 2.326008789452657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5822 + }, + { + "epoch": 2.3264083100279667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5823 + }, + { + "epoch": 2.326807830603276, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5824 + }, + { + "epoch": 2.327207351178586, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5825 + }, + { + "epoch": 2.327606871753895, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5826 + }, + { + "epoch": 2.328006392329205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5827 + }, + { + "epoch": 2.3284059129045147, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5828 + }, + { + "epoch": 2.328805433479824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5829 + }, + { + "epoch": 2.329204954055134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5830 + }, + { + "epoch": 2.3296044746304436, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5831 + }, + { + "epoch": 2.330003995205753, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5832 + }, + { + "epoch": 2.3304035157810628, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 5833 + }, + { + "epoch": 2.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5834 + }, + { + "epoch": 2.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5835 + }, + { + "epoch": 2.3316020775069917, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5836 + }, + { + "epoch": 2.332001598082301, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5837 + }, + { + "epoch": 2.332401118657611, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 5838 + }, + { + "epoch": 2.3328006392329206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5839 + }, + { + "epoch": 2.3332001598082304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5840 + }, + { + "epoch": 2.3335996803835397, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5841 + }, + { + "epoch": 2.3339992009588495, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5842 + }, + { + "epoch": 2.334398721534159, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5843 + }, + { + "epoch": 2.3347982421094686, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5844 + }, + { + "epoch": 2.3351977626847784, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5845 + }, + { + "epoch": 2.3355972832600878, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5846 + }, + { + "epoch": 2.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5847 + }, + { + "epoch": 2.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5848 + }, + { + "epoch": 2.3367958449860167, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5849 + }, + { + "epoch": 2.3371953655613265, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5850 + }, + { + "epoch": 2.3375948861366362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5851 + }, + { + "epoch": 2.3379944067119456, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5852 + }, + { + "epoch": 2.3383939272872554, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5853 + }, + { + "epoch": 2.3387934478625647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5854 + }, + { + "epoch": 2.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5855 + }, + { + "epoch": 2.3395924890131843, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5856 + }, + { + "epoch": 2.3399920095884936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5857 + }, + { + "epoch": 2.3403915301638034, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5858 + }, + { + "epoch": 2.340791050739113, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5859 + }, + { + "epoch": 2.3411905713144225, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5860 + }, + { + "epoch": 2.3415900918897323, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5861 + }, + { + "epoch": 2.341989612465042, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5862 + }, + { + "epoch": 2.3423891330403515, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5863 + }, + { + "epoch": 2.3427886536156612, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5864 + }, + { + "epoch": 2.343188174190971, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5865 + }, + { + "epoch": 2.3435876947662804, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5866 + }, + { + "epoch": 2.34398721534159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5867 + }, + { + "epoch": 2.3443867359169, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5868 + }, + { + "epoch": 2.3447862564922093, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5869 + }, + { + "epoch": 2.345185777067519, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5870 + }, + { + "epoch": 2.3455852976428284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5871 + }, + { + "epoch": 2.345984818218138, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5872 + }, + { + "epoch": 2.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5873 + }, + { + "epoch": 2.3467838593687573, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5874 + }, + { + "epoch": 2.347183379944067, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5875 + }, + { + "epoch": 2.347582900519377, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5876 + }, + { + "epoch": 2.3479824210946862, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 5877 + }, + { + "epoch": 2.348381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5878 + }, + { + "epoch": 2.348781462245306, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5879 + }, + { + "epoch": 2.349180982820615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5880 + }, + { + "epoch": 2.349580503395925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 5881 + }, + { + "epoch": 2.3499800239712343, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5882 + }, + { + "epoch": 2.350379544546544, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5883 + }, + { + "epoch": 2.350779065121854, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5884 + }, + { + "epoch": 2.3511785856971636, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5885 + }, + { + "epoch": 2.351578106272473, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5886 + }, + { + "epoch": 2.3519776268477828, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5887 + }, + { + "epoch": 2.352377147423092, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5888 + }, + { + "epoch": 2.352776667998402, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5889 + }, + { + "epoch": 2.3531761885737117, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5890 + }, + { + "epoch": 2.353575709149021, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5891 + }, + { + "epoch": 2.353975229724331, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5892 + }, + { + "epoch": 2.3543747502996406, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5893 + }, + { + "epoch": 2.35477427087495, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5894 + }, + { + "epoch": 2.3551737914502597, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5895 + }, + { + "epoch": 2.3555733120255695, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5896 + }, + { + "epoch": 2.355972832600879, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 5897 + }, + { + "epoch": 2.3563723531761886, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5898 + }, + { + "epoch": 2.356771873751498, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5899 + }, + { + "epoch": 2.357171394326808, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5900 + }, + { + "epoch": 2.3575709149021176, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5901 + }, + { + "epoch": 2.357970435477427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5902 + }, + { + "epoch": 2.3583699560527367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5903 + }, + { + "epoch": 2.3587694766280465, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5904 + }, + { + "epoch": 2.359168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5905 + }, + { + "epoch": 2.3595685177786656, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 5906 + }, + { + "epoch": 2.3599680383539754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5907 + }, + { + "epoch": 2.3603675589292847, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5908 + }, + { + "epoch": 2.3607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5909 + }, + { + "epoch": 2.361166600079904, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5910 + }, + { + "epoch": 2.3615661206552137, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5911 + }, + { + "epoch": 2.3619656412305234, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5912 + }, + { + "epoch": 2.3623651618058332, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5913 + }, + { + "epoch": 2.3627646823811426, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5914 + }, + { + "epoch": 2.3631642029564524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5915 + }, + { + "epoch": 2.3635637235317617, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5916 + }, + { + "epoch": 2.3639632441070715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5917 + }, + { + "epoch": 2.3643627646823813, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5918 + }, + { + "epoch": 2.3647622852576906, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5919 + }, + { + "epoch": 2.3651618058330004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 5920 + }, + { + "epoch": 2.36556132640831, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5921 + }, + { + "epoch": 2.3659608469836195, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5922 + }, + { + "epoch": 2.3663603675589293, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5923 + }, + { + "epoch": 2.366759888134239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5924 + }, + { + "epoch": 2.3671594087095484, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5925 + }, + { + "epoch": 2.3675589292848582, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5926 + }, + { + "epoch": 2.3679584498601676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5927 + }, + { + "epoch": 2.3683579704354774, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5928 + }, + { + "epoch": 2.368757491010787, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5929 + }, + { + "epoch": 2.369157011586097, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5930 + }, + { + "epoch": 2.3695565321614063, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5931 + }, + { + "epoch": 2.369956052736716, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5932 + }, + { + "epoch": 2.3703555733120254, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5933 + }, + { + "epoch": 2.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5934 + }, + { + "epoch": 2.371154614462645, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5935 + }, + { + "epoch": 2.3715541350379543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5936 + }, + { + "epoch": 2.371953655613264, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5937 + }, + { + "epoch": 2.372353176188574, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5938 + }, + { + "epoch": 2.3727526967638832, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5939 + }, + { + "epoch": 2.373152217339193, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5940 + }, + { + "epoch": 2.373551737914503, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5941 + }, + { + "epoch": 2.373951258489812, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5942 + }, + { + "epoch": 2.374350779065122, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 5943 + }, + { + "epoch": 2.3747502996404313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5944 + }, + { + "epoch": 2.375149820215741, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5945 + }, + { + "epoch": 2.375549340791051, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5946 + }, + { + "epoch": 2.37594886136636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5947 + }, + { + "epoch": 2.37634838194167, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5948 + }, + { + "epoch": 2.3767479025169798, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5949 + }, + { + "epoch": 2.377147423092289, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 5950 + }, + { + "epoch": 2.377546943667599, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5951 + }, + { + "epoch": 2.3779464642429087, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5952 + }, + { + "epoch": 2.378345984818218, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 5953 + }, + { + "epoch": 2.378745505393528, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5954 + }, + { + "epoch": 2.379145025968837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5955 + }, + { + "epoch": 2.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5956 + }, + { + "epoch": 2.3799440671194567, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5957 + }, + { + "epoch": 2.3803435876947665, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5958 + }, + { + "epoch": 2.380743108270076, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5959 + }, + { + "epoch": 2.3811426288453856, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5960 + }, + { + "epoch": 2.381542149420695, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5961 + }, + { + "epoch": 2.3819416699960048, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5962 + }, + { + "epoch": 2.3823411905713145, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5963 + }, + { + "epoch": 2.382740711146624, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5964 + }, + { + "epoch": 2.3831402317219337, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5965 + }, + { + "epoch": 2.3835397522972435, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5966 + }, + { + "epoch": 2.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5967 + }, + { + "epoch": 2.3843387934478626, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5968 + }, + { + "epoch": 2.3847383140231724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5969 + }, + { + "epoch": 2.3851378345984817, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5970 + }, + { + "epoch": 2.3855373551737915, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 5971 + }, + { + "epoch": 2.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5972 + }, + { + "epoch": 2.3863363963244106, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 5973 + }, + { + "epoch": 2.3867359168997204, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5974 + }, + { + "epoch": 2.38713543747503, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5975 + }, + { + "epoch": 2.3875349580503396, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5976 + }, + { + "epoch": 2.3879344786256493, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5977 + }, + { + "epoch": 2.3883339992009587, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5978 + }, + { + "epoch": 2.3887335197762685, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5979 + }, + { + "epoch": 2.3891330403515783, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5980 + }, + { + "epoch": 2.3895325609268876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5981 + }, + { + "epoch": 2.3899320815021974, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5982 + }, + { + "epoch": 2.390331602077507, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5983 + }, + { + "epoch": 2.3907311226528165, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5984 + }, + { + "epoch": 2.3911306432281263, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5985 + }, + { + "epoch": 2.391530163803436, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 5986 + }, + { + "epoch": 2.3919296843787454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5987 + }, + { + "epoch": 2.392329204954055, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5988 + }, + { + "epoch": 2.3927287255293646, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5989 + }, + { + "epoch": 2.3931282461046743, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5990 + }, + { + "epoch": 2.393527766679984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5991 + }, + { + "epoch": 2.3939272872552935, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5992 + }, + { + "epoch": 2.3943268078306033, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5993 + }, + { + "epoch": 2.394726328405913, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5994 + }, + { + "epoch": 2.3951258489812224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5995 + }, + { + "epoch": 2.395525369556532, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5996 + }, + { + "epoch": 2.395924890131842, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5997 + }, + { + "epoch": 2.3963244107071513, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5998 + }, + { + "epoch": 2.396723931282461, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5999 + }, + { + "epoch": 2.3971234518577704, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6000 + }, + { + "epoch": 2.39752297243308, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6001 + }, + { + "epoch": 2.39792249300839, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6002 + }, + { + "epoch": 2.3983220135837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6003 + }, + { + "epoch": 2.398721534159009, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6004 + }, + { + "epoch": 2.399121054734319, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6005 + }, + { + "epoch": 2.3995205753096283, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6006 + }, + { + "epoch": 2.399920095884938, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6007 + }, + { + "epoch": 2.400319616460248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6008 + }, + { + "epoch": 2.400719137035557, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6009 + }, + { + "epoch": 2.401118657610867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6010 + }, + { + "epoch": 2.4015181781861767, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6011 + }, + { + "epoch": 2.401917698761486, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6012 + }, + { + "epoch": 2.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6013 + }, + { + "epoch": 2.4027167399121057, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6014 + }, + { + "epoch": 2.403116260487415, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6015 + }, + { + "epoch": 2.403515781062725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6016 + }, + { + "epoch": 2.403915301638034, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6017 + }, + { + "epoch": 2.404314822213344, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6018 + }, + { + "epoch": 2.4047143427886537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6019 + }, + { + "epoch": 2.4051138633639635, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6020 + }, + { + "epoch": 2.405513383939273, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6021 + }, + { + "epoch": 2.4059129045145826, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6022 + }, + { + "epoch": 2.406312425089892, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6023 + }, + { + "epoch": 2.4067119456652017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6024 + }, + { + "epoch": 2.4071114662405115, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6025 + }, + { + "epoch": 2.407510986815821, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6026 + }, + { + "epoch": 2.4079105073911307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6027 + }, + { + "epoch": 2.4083100279664404, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6028 + }, + { + "epoch": 2.40870954854175, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 6029 + }, + { + "epoch": 2.4091090691170596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6030 + }, + { + "epoch": 2.4095085896923694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6031 + }, + { + "epoch": 2.4099081102676787, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6032 + }, + { + "epoch": 2.4103076308429885, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 6033 + }, + { + "epoch": 2.410707151418298, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 6034 + }, + { + "epoch": 2.4111066719936076, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6035 + }, + { + "epoch": 2.4115061925689174, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6036 + }, + { + "epoch": 2.4119057131442267, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6037 + }, + { + "epoch": 2.4123052337195365, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6038 + }, + { + "epoch": 2.4127047542948463, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6039 + }, + { + "epoch": 2.4131042748701557, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6040 + }, + { + "epoch": 2.4135037954454654, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6041 + }, + { + "epoch": 2.4139033160207752, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6042 + }, + { + "epoch": 2.4143028365960846, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 6043 + }, + { + "epoch": 2.4147023571713944, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6044 + }, + { + "epoch": 2.4151018777467037, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6045 + }, + { + "epoch": 2.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6046 + }, + { + "epoch": 2.4159009188973233, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6047 + }, + { + "epoch": 2.416300439472633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 6048 + }, + { + "epoch": 2.4166999600479424, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6049 + }, + { + "epoch": 2.417099480623252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6050 + }, + { + "epoch": 2.4174990011985615, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6051 + }, + { + "epoch": 2.4178985217738713, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6052 + }, + { + "epoch": 2.418298042349181, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6053 + }, + { + "epoch": 2.4186975629244905, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6054 + }, + { + "epoch": 2.4190970834998002, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6055 + }, + { + "epoch": 2.41949660407511, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6056 + }, + { + "epoch": 2.4198961246504194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6057 + }, + { + "epoch": 2.420295645225729, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6058 + }, + { + "epoch": 2.420695165801039, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6059 + }, + { + "epoch": 2.4210946863763483, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6060 + }, + { + "epoch": 2.421494206951658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6061 + }, + { + "epoch": 2.4218937275269674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 6062 + }, + { + "epoch": 2.422293248102277, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6063 + }, + { + "epoch": 2.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6064 + }, + { + "epoch": 2.4230922892528968, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6065 + }, + { + "epoch": 2.423491809828206, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6066 + }, + { + "epoch": 2.423891330403516, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6067 + }, + { + "epoch": 2.4242908509788252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6068 + }, + { + "epoch": 2.424690371554135, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6069 + }, + { + "epoch": 2.425089892129445, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6070 + }, + { + "epoch": 2.425489412704754, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6071 + }, + { + "epoch": 2.425888933280064, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6072 + }, + { + "epoch": 2.4262884538553737, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6073 + }, + { + "epoch": 2.426687974430683, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6074 + }, + { + "epoch": 2.427087495005993, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6075 + }, + { + "epoch": 2.4274870155813026, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6076 + }, + { + "epoch": 2.427886536156612, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6077 + }, + { + "epoch": 2.4282860567319218, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6078 + }, + { + "epoch": 2.428685577307231, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6079 + }, + { + "epoch": 2.429085097882541, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6080 + }, + { + "epoch": 2.4294846184578507, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6081 + }, + { + "epoch": 2.42988413903316, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6082 + }, + { + "epoch": 2.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6083 + }, + { + "epoch": 2.4306831801837796, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6084 + }, + { + "epoch": 2.431082700759089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6085 + }, + { + "epoch": 2.4314822213343987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6086 + }, + { + "epoch": 2.4318817419097085, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6087 + }, + { + "epoch": 2.432281262485018, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6088 + }, + { + "epoch": 2.4326807830603276, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6089 + }, + { + "epoch": 2.433080303635637, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6090 + }, + { + "epoch": 2.4334798242109468, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6091 + }, + { + "epoch": 2.4338793447862566, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 6092 + }, + { + "epoch": 2.4342788653615663, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6093 + }, + { + "epoch": 2.4346783859368757, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6094 + }, + { + "epoch": 2.4350779065121855, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 6095 + }, + { + "epoch": 2.435477427087495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 6096 + }, + { + "epoch": 2.4358769476628046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 6097 + }, + { + "epoch": 2.4362764682381144, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6098 + }, + { + "epoch": 2.4366759888134237, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6099 + }, + { + "epoch": 2.4370755093887335, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6100 + }, + { + "epoch": 2.4374750299640433, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6101 + }, + { + "epoch": 2.4378745505393526, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6102 + }, + { + "epoch": 2.4382740711146624, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6103 + }, + { + "epoch": 2.438673591689972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6104 + }, + { + "epoch": 2.4390731122652816, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6105 + }, + { + "epoch": 2.4394726328405913, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6106 + }, + { + "epoch": 2.4398721534159007, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6107 + }, + { + "epoch": 2.4402716739912105, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6108 + }, + { + "epoch": 2.4406711945665203, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6109 + }, + { + "epoch": 2.44107071514183, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6110 + }, + { + "epoch": 2.4414702357171394, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6111 + }, + { + "epoch": 2.441869756292449, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6112 + }, + { + "epoch": 2.4422692768677585, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6113 + }, + { + "epoch": 2.4426687974430683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6114 + }, + { + "epoch": 2.443068318018378, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6115 + }, + { + "epoch": 2.4434678385936874, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6116 + }, + { + "epoch": 2.443867359168997, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6117 + }, + { + "epoch": 2.444266879744307, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 6118 + }, + { + "epoch": 2.4446664003196164, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6119 + }, + { + "epoch": 2.445065920894926, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6120 + }, + { + "epoch": 2.445465441470236, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6121 + }, + { + "epoch": 2.4458649620455453, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6122 + }, + { + "epoch": 2.446264482620855, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6123 + }, + { + "epoch": 2.4466640031961644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6124 + }, + { + "epoch": 2.447063523771474, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6125 + }, + { + "epoch": 2.447463044346784, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6126 + }, + { + "epoch": 2.4478625649220933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6127 + }, + { + "epoch": 2.448262085497403, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6128 + }, + { + "epoch": 2.448661606072713, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6129 + }, + { + "epoch": 2.4490611266480222, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6130 + }, + { + "epoch": 2.449460647223332, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6131 + }, + { + "epoch": 2.449860167798642, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6132 + }, + { + "epoch": 2.450259688373951, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6133 + }, + { + "epoch": 2.450659208949261, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6134 + }, + { + "epoch": 2.4510587295245703, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 6135 + }, + { + "epoch": 2.45145825009988, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6136 + }, + { + "epoch": 2.45185777067519, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6137 + }, + { + "epoch": 2.4522572912504996, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6138 + }, + { + "epoch": 2.452656811825809, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6139 + }, + { + "epoch": 2.4530563324011188, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6140 + }, + { + "epoch": 2.453455852976428, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6141 + }, + { + "epoch": 2.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6142 + }, + { + "epoch": 2.4542548941270477, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6143 + }, + { + "epoch": 2.454654414702357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6144 + }, + { + "epoch": 2.455053935277667, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6145 + }, + { + "epoch": 2.4554534558529766, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6146 + }, + { + "epoch": 2.455852976428286, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6147 + }, + { + "epoch": 2.4562524970035957, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6148 + }, + { + "epoch": 2.4566520175789055, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6149 + }, + { + "epoch": 2.457051538154215, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6150 + }, + { + "epoch": 2.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6151 + }, + { + "epoch": 2.457850579304834, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6152 + }, + { + "epoch": 2.4582500998801438, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6153 + }, + { + "epoch": 2.4586496204554535, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6154 + }, + { + "epoch": 2.4590491410307633, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6155 + }, + { + "epoch": 2.4594486616060727, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6156 + }, + { + "epoch": 2.4598481821813825, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6157 + }, + { + "epoch": 2.460247702756692, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6158 + }, + { + "epoch": 2.4606472233320016, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6159 + }, + { + "epoch": 2.4610467439073114, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6160 + }, + { + "epoch": 2.4614462644826207, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6161 + }, + { + "epoch": 2.4618457850579305, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6162 + }, + { + "epoch": 2.4622453056332403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6163 + }, + { + "epoch": 2.4626448262085496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6164 + }, + { + "epoch": 2.4630443467838594, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6165 + }, + { + "epoch": 2.463443867359169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6166 + }, + { + "epoch": 2.4638433879344785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6167 + }, + { + "epoch": 2.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6168 + }, + { + "epoch": 2.4646424290850977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 6169 + }, + { + "epoch": 2.4650419496604075, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6170 + }, + { + "epoch": 2.4654414702357172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6171 + }, + { + "epoch": 2.4658409908110266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6172 + }, + { + "epoch": 2.4662405113863364, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6173 + }, + { + "epoch": 2.466640031961646, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6174 + }, + { + "epoch": 2.4670395525369555, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6175 + }, + { + "epoch": 2.4674390731122653, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6176 + }, + { + "epoch": 2.467838593687575, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6177 + }, + { + "epoch": 2.4682381142628844, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6178 + }, + { + "epoch": 2.468637634838194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6179 + }, + { + "epoch": 2.4690371554135035, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6180 + }, + { + "epoch": 2.4694366759888133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6181 + }, + { + "epoch": 2.469836196564123, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6182 + }, + { + "epoch": 2.470235717139433, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6183 + }, + { + "epoch": 2.4706352377147422, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6184 + }, + { + "epoch": 2.471034758290052, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6185 + }, + { + "epoch": 2.4714342788653614, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 6186 + }, + { + "epoch": 2.471833799440671, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6187 + }, + { + "epoch": 2.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6188 + }, + { + "epoch": 2.4726328405912903, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6189 + }, + { + "epoch": 2.4730323611666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6190 + }, + { + "epoch": 2.47343188174191, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6191 + }, + { + "epoch": 2.473831402317219, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6192 + }, + { + "epoch": 2.474230922892529, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6193 + }, + { + "epoch": 2.474630443467839, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6194 + }, + { + "epoch": 2.475029964043148, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6195 + }, + { + "epoch": 2.475429484618458, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6196 + }, + { + "epoch": 2.4758290051937673, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6197 + }, + { + "epoch": 2.476228525769077, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6198 + }, + { + "epoch": 2.476628046344387, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6199 + }, + { + "epoch": 2.4770275669196966, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6200 + }, + { + "epoch": 2.477427087495006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6201 + }, + { + "epoch": 2.4778266080703157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6202 + }, + { + "epoch": 2.478226128645625, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6203 + }, + { + "epoch": 2.478625649220935, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6204 + }, + { + "epoch": 2.4790251697962447, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6205 + }, + { + "epoch": 2.479424690371554, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6206 + }, + { + "epoch": 2.479824210946864, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6207 + }, + { + "epoch": 2.4802237315221736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6208 + }, + { + "epoch": 2.480623252097483, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6209 + }, + { + "epoch": 2.4810227726727927, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6210 + }, + { + "epoch": 2.4814222932481025, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6211 + }, + { + "epoch": 2.481821813823412, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6212 + }, + { + "epoch": 2.4822213343987216, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6213 + }, + { + "epoch": 2.482620854974031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 6214 + }, + { + "epoch": 2.4830203755493407, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6215 + }, + { + "epoch": 2.4834198961246505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6216 + }, + { + "epoch": 2.48381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 6217 + }, + { + "epoch": 2.4842189372752697, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6218 + }, + { + "epoch": 2.4846184578505794, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6219 + }, + { + "epoch": 2.485017978425889, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6220 + }, + { + "epoch": 2.4854174990011986, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6221 + }, + { + "epoch": 2.4858170195765084, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6222 + }, + { + "epoch": 2.4862165401518177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6223 + }, + { + "epoch": 2.4866160607271275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6224 + }, + { + "epoch": 2.487015581302437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6225 + }, + { + "epoch": 2.4874151018777466, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6226 + }, + { + "epoch": 2.4878146224530564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6227 + }, + { + "epoch": 2.488214143028366, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6228 + }, + { + "epoch": 2.4886136636036755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6229 + }, + { + "epoch": 2.4890131841789853, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6230 + }, + { + "epoch": 2.4894127047542947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6231 + }, + { + "epoch": 2.4898122253296044, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6232 + }, + { + "epoch": 2.4902117459049142, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6233 + }, + { + "epoch": 2.4906112664802236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6234 + }, + { + "epoch": 2.4910107870555334, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6235 + }, + { + "epoch": 2.491410307630843, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6236 + }, + { + "epoch": 2.4918098282061525, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6237 + }, + { + "epoch": 2.4922093487814623, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6238 + }, + { + "epoch": 2.492608869356772, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6239 + }, + { + "epoch": 2.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6240 + }, + { + "epoch": 2.493407910507391, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6241 + }, + { + "epoch": 2.4938074310827005, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6242 + }, + { + "epoch": 2.4942069516580103, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6243 + }, + { + "epoch": 2.49460647223332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6244 + }, + { + "epoch": 2.49500599280863, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6245 + }, + { + "epoch": 2.4954055133839392, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 6246 + }, + { + "epoch": 2.495805033959249, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6247 + }, + { + "epoch": 2.4962045545345584, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6248 + }, + { + "epoch": 2.496604075109868, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6249 + }, + { + "epoch": 2.497003595685178, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6250 + }, + { + "epoch": 2.4974031162604873, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6251 + }, + { + "epoch": 2.497802636835797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6252 + }, + { + "epoch": 2.498202157411107, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6253 + }, + { + "epoch": 2.498601677986416, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6254 + }, + { + "epoch": 2.499001198561726, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 6255 + }, + { + "epoch": 2.4994007191370358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6256 + }, + { + "epoch": 2.499800239712345, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6257 + }, + { + "epoch": 2.500199760287655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 6258 + }, + { + "epoch": 2.5005992808629642, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6259 + }, + { + "epoch": 2.500998801438274, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6260 + }, + { + "epoch": 2.501398322013584, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6261 + }, + { + "epoch": 2.5017978425888936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6262 + }, + { + "epoch": 2.502197363164203, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6263 + }, + { + "epoch": 2.5025968837395127, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6264 + }, + { + "epoch": 2.502996404314822, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6265 + }, + { + "epoch": 2.503395924890132, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6266 + }, + { + "epoch": 2.5037954454654416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6267 + }, + { + "epoch": 2.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6268 + }, + { + "epoch": 2.5045944866160608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6269 + }, + { + "epoch": 2.50499400719137, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6270 + }, + { + "epoch": 2.50539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6271 + }, + { + "epoch": 2.5057930483419897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 6272 + }, + { + "epoch": 2.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6273 + }, + { + "epoch": 2.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6274 + }, + { + "epoch": 2.5069916100679186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6275 + }, + { + "epoch": 2.507391130643228, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 6276 + }, + { + "epoch": 2.5077906512185377, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6277 + }, + { + "epoch": 2.5081901717938475, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6278 + }, + { + "epoch": 2.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6279 + }, + { + "epoch": 2.5089892129444666, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6280 + }, + { + "epoch": 2.5093887335197764, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6281 + }, + { + "epoch": 2.5097882540950858, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6282 + }, + { + "epoch": 2.5101877746703956, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6283 + }, + { + "epoch": 2.5105872952457053, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6284 + }, + { + "epoch": 2.5109868158210147, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6285 + }, + { + "epoch": 2.5113863363963245, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6286 + }, + { + "epoch": 2.511785856971634, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6287 + }, + { + "epoch": 2.5121853775469436, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6288 + }, + { + "epoch": 2.5125848981222534, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6289 + }, + { + "epoch": 2.512984418697563, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6290 + }, + { + "epoch": 2.5133839392728725, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6291 + }, + { + "epoch": 2.5137834598481823, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6292 + }, + { + "epoch": 2.5141829804234916, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6293 + }, + { + "epoch": 2.5145825009988014, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6294 + }, + { + "epoch": 2.514982021574111, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6295 + }, + { + "epoch": 2.5153815421494206, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6296 + }, + { + "epoch": 2.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6297 + }, + { + "epoch": 2.5161805833000397, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6298 + }, + { + "epoch": 2.5165801038753495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6299 + }, + { + "epoch": 2.5169796244506593, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6300 + }, + { + "epoch": 2.517379145025969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6301 + }, + { + "epoch": 2.5177786656012784, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6302 + }, + { + "epoch": 2.518178186176588, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6303 + }, + { + "epoch": 2.5185777067518975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6304 + }, + { + "epoch": 2.5189772273272073, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6305 + }, + { + "epoch": 2.519376747902517, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6306 + }, + { + "epoch": 2.519776268477827, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6307 + }, + { + "epoch": 2.520175789053136, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6308 + }, + { + "epoch": 2.520575309628446, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6309 + }, + { + "epoch": 2.5209748302037553, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6310 + }, + { + "epoch": 2.521374350779065, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 6311 + }, + { + "epoch": 2.521773871354375, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6312 + }, + { + "epoch": 2.5221733919296843, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 6313 + }, + { + "epoch": 2.522572912504994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6314 + }, + { + "epoch": 2.5229724330803034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6315 + }, + { + "epoch": 2.523371953655613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6316 + }, + { + "epoch": 2.523771474230923, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6317 + }, + { + "epoch": 2.5241709948062327, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6318 + }, + { + "epoch": 2.524570515381542, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6319 + }, + { + "epoch": 2.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6320 + }, + { + "epoch": 2.525369556532161, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6321 + }, + { + "epoch": 2.525769077107471, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6322 + }, + { + "epoch": 2.526168597682781, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 6323 + }, + { + "epoch": 2.52656811825809, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6324 + }, + { + "epoch": 2.5269676388334, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6325 + }, + { + "epoch": 2.5273671594087097, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6326 + }, + { + "epoch": 2.527766679984019, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 6327 + }, + { + "epoch": 2.528166200559329, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6328 + }, + { + "epoch": 2.5285657211346386, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6329 + }, + { + "epoch": 2.528965241709948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6330 + }, + { + "epoch": 2.5293647622852578, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6331 + }, + { + "epoch": 2.529764282860567, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6332 + }, + { + "epoch": 2.530163803435877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6333 + }, + { + "epoch": 2.5305633240111867, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 6334 + }, + { + "epoch": 2.5309628445864965, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6335 + }, + { + "epoch": 2.531362365161806, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6336 + }, + { + "epoch": 2.5317618857371156, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6337 + }, + { + "epoch": 2.532161406312425, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6338 + }, + { + "epoch": 2.5325609268877347, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6339 + }, + { + "epoch": 2.5329604474630445, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6340 + }, + { + "epoch": 2.533359968038354, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6341 + }, + { + "epoch": 2.5337594886136636, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6342 + }, + { + "epoch": 2.534159009188973, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6343 + }, + { + "epoch": 2.5345585297642828, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6344 + }, + { + "epoch": 2.5349580503395925, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6345 + }, + { + "epoch": 2.5353575709149023, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6346 + }, + { + "epoch": 2.5357570914902117, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 6347 + }, + { + "epoch": 2.5361566120655215, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6348 + }, + { + "epoch": 2.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6349 + }, + { + "epoch": 2.5369556532161406, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 6350 + }, + { + "epoch": 2.5373551737914504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6351 + }, + { + "epoch": 2.53775469436676, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6352 + }, + { + "epoch": 2.5381542149420695, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6353 + }, + { + "epoch": 2.5385537355173793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6354 + }, + { + "epoch": 2.5389532560926886, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6355 + }, + { + "epoch": 2.5393527766679984, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6356 + }, + { + "epoch": 2.539752297243308, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6357 + }, + { + "epoch": 2.5401518178186175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6358 + }, + { + "epoch": 2.5405513383939273, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 6359 + }, + { + "epoch": 2.5409508589692367, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6360 + }, + { + "epoch": 2.5413503795445465, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6361 + }, + { + "epoch": 2.5417499001198562, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 6362 + }, + { + "epoch": 2.542149420695166, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6363 + }, + { + "epoch": 2.5425489412704754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6364 + }, + { + "epoch": 2.542948461845785, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6365 + }, + { + "epoch": 2.5433479824210945, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 6366 + }, + { + "epoch": 2.5437475029964043, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6367 + }, + { + "epoch": 2.544147023571714, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6368 + }, + { + "epoch": 2.5445465441470234, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6369 + }, + { + "epoch": 2.544946064722333, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6370 + }, + { + "epoch": 2.545345585297643, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6371 + }, + { + "epoch": 2.5457451058729523, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6372 + }, + { + "epoch": 2.546144626448262, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6373 + }, + { + "epoch": 2.546544147023572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6374 + }, + { + "epoch": 2.5469436675988812, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6375 + }, + { + "epoch": 2.547343188174191, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6376 + }, + { + "epoch": 2.5477427087495004, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6377 + }, + { + "epoch": 2.54814222932481, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6378 + }, + { + "epoch": 2.54854174990012, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6379 + }, + { + "epoch": 2.5489412704754297, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6380 + }, + { + "epoch": 2.549340791050739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6381 + }, + { + "epoch": 2.549740311626049, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6382 + }, + { + "epoch": 2.550139832201358, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6383 + }, + { + "epoch": 2.550539352776668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6384 + }, + { + "epoch": 2.5509388733519778, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6385 + }, + { + "epoch": 2.551338393927287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 6386 + }, + { + "epoch": 2.551737914502597, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6387 + }, + { + "epoch": 2.5521374350779062, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 6388 + }, + { + "epoch": 2.552536955653216, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6389 + }, + { + "epoch": 2.552936476228526, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6390 + }, + { + "epoch": 2.5533359968038356, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6391 + }, + { + "epoch": 2.553735517379145, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6392 + }, + { + "epoch": 2.5541350379544547, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6393 + }, + { + "epoch": 2.554534558529764, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6394 + }, + { + "epoch": 2.554934079105074, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 6395 + }, + { + "epoch": 2.5553335996803836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6396 + }, + { + "epoch": 2.5557331202556934, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6397 + }, + { + "epoch": 2.556132640831003, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6398 + }, + { + "epoch": 2.5565321614063126, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6399 + }, + { + "epoch": 2.556931681981622, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6400 + }, + { + "epoch": 2.5573312025569317, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6401 + }, + { + "epoch": 2.5577307231322415, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6402 + }, + { + "epoch": 2.558130243707551, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6403 + }, + { + "epoch": 2.5585297642828606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6404 + }, + { + "epoch": 2.55892928485817, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 6405 + }, + { + "epoch": 2.5593288054334797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 6406 + }, + { + "epoch": 2.5597283260087895, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 6407 + }, + { + "epoch": 2.5601278465840993, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6408 + }, + { + "epoch": 2.5605273671594087, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6409 + }, + { + "epoch": 2.5609268877347184, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6410 + }, + { + "epoch": 2.561326408310028, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6411 + }, + { + "epoch": 2.5617259288853376, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 6412 + }, + { + "epoch": 2.5621254494606474, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6413 + }, + { + "epoch": 2.5625249700359567, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6414 + }, + { + "epoch": 2.5629244906112665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6415 + }, + { + "epoch": 2.5633240111865763, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6416 + }, + { + "epoch": 2.5637235317618856, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 6417 + }, + { + "epoch": 2.5641230523371954, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6418 + }, + { + "epoch": 2.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 6419 + }, + { + "epoch": 2.5649220934878145, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6420 + }, + { + "epoch": 2.5653216140631243, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6421 + }, + { + "epoch": 2.5657211346384337, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6422 + }, + { + "epoch": 2.5661206552137434, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6423 + }, + { + "epoch": 2.5665201757890532, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6424 + }, + { + "epoch": 2.566919696364363, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6425 + }, + { + "epoch": 2.5673192169396724, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6426 + }, + { + "epoch": 2.567718737514982, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6427 + }, + { + "epoch": 2.5681182580902915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6428 + }, + { + "epoch": 2.5685177786656013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6429 + }, + { + "epoch": 2.568917299240911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6430 + }, + { + "epoch": 2.5693168198162204, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6431 + }, + { + "epoch": 2.56971634039153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6432 + }, + { + "epoch": 2.5701158609668395, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6433 + }, + { + "epoch": 2.5705153815421493, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 6434 + }, + { + "epoch": 2.570914902117459, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6435 + }, + { + "epoch": 2.571314422692769, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6436 + }, + { + "epoch": 2.5717139432680782, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6437 + }, + { + "epoch": 2.572113463843388, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6438 + }, + { + "epoch": 2.5725129844186974, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6439 + }, + { + "epoch": 2.572912504994007, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6440 + }, + { + "epoch": 2.573312025569317, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6441 + }, + { + "epoch": 2.5737115461446267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6442 + }, + { + "epoch": 2.574111066719936, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6443 + }, + { + "epoch": 2.574510587295246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 6444 + }, + { + "epoch": 2.574910107870555, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6445 + }, + { + "epoch": 2.575309628445865, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6446 + }, + { + "epoch": 2.5757091490211748, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6447 + }, + { + "epoch": 2.576108669596484, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6448 + }, + { + "epoch": 2.576508190171794, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6449 + }, + { + "epoch": 2.5769077107471032, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6450 + }, + { + "epoch": 2.577307231322413, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6451 + }, + { + "epoch": 2.577706751897723, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6452 + }, + { + "epoch": 2.5781062724730326, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6453 + }, + { + "epoch": 2.578505793048342, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6454 + }, + { + "epoch": 2.5789053136236517, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6455 + }, + { + "epoch": 2.579304834198961, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6456 + }, + { + "epoch": 2.579704354774271, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 6457 + }, + { + "epoch": 2.5801038753495806, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6458 + }, + { + "epoch": 2.58050339592489, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6459 + }, + { + "epoch": 2.5809029165001998, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6460 + }, + { + "epoch": 2.5813024370755095, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6461 + }, + { + "epoch": 2.581701957650819, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6462 + }, + { + "epoch": 2.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6463 + }, + { + "epoch": 2.5825009988014385, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6464 + }, + { + "epoch": 2.582900519376748, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6465 + }, + { + "epoch": 2.5833000399520576, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6466 + }, + { + "epoch": 2.583699560527367, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6467 + }, + { + "epoch": 2.5840990811026767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6468 + }, + { + "epoch": 2.5844986016779865, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6469 + }, + { + "epoch": 2.5848981222532963, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6470 + }, + { + "epoch": 2.5852976428286056, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6471 + }, + { + "epoch": 2.5856971634039154, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6472 + }, + { + "epoch": 2.5860966839792248, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6473 + }, + { + "epoch": 2.5864962045545346, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6474 + }, + { + "epoch": 2.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 6475 + }, + { + "epoch": 2.5872952457051537, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6476 + }, + { + "epoch": 2.5876947662804635, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6477 + }, + { + "epoch": 2.588094286855773, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6478 + }, + { + "epoch": 2.5884938074310826, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6479 + }, + { + "epoch": 2.5888933280063924, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6480 + }, + { + "epoch": 2.589292848581702, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6481 + }, + { + "epoch": 2.5896923691570115, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6482 + }, + { + "epoch": 2.5900918897323213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6483 + }, + { + "epoch": 2.5904914103076306, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 6484 + }, + { + "epoch": 2.5908909308829404, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6485 + }, + { + "epoch": 2.59129045145825, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6486 + }, + { + "epoch": 2.59168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6487 + }, + { + "epoch": 2.5920894926088693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6488 + }, + { + "epoch": 2.592489013184179, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6489 + }, + { + "epoch": 2.5928885337594885, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6490 + }, + { + "epoch": 2.5932880543347983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6491 + }, + { + "epoch": 2.593687574910108, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6492 + }, + { + "epoch": 2.5940870954854174, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 6493 + }, + { + "epoch": 2.594486616060727, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6494 + }, + { + "epoch": 2.5948861366360365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6495 + }, + { + "epoch": 2.5952856572113463, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6496 + }, + { + "epoch": 2.595685177786656, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6497 + }, + { + "epoch": 2.596084698361966, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6498 + }, + { + "epoch": 2.596484218937275, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6499 + }, + { + "epoch": 2.596883739512585, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6500 + }, + { + "epoch": 2.5972832600878943, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6501 + }, + { + "epoch": 2.597682780663204, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6502 + }, + { + "epoch": 2.598082301238514, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 6503 + }, + { + "epoch": 2.5984818218138233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6504 + }, + { + "epoch": 2.598881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6505 + }, + { + "epoch": 2.599280862964443, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6506 + }, + { + "epoch": 2.599680383539752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6507 + }, + { + "epoch": 2.600079904115062, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 6508 + }, + { + "epoch": 2.6004794246903717, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6509 + }, + { + "epoch": 2.600878945265681, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6510 + }, + { + "epoch": 2.601278465840991, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6511 + }, + { + "epoch": 2.6016779864163, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6512 + }, + { + "epoch": 2.60207750699161, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6513 + }, + { + "epoch": 2.60247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6514 + }, + { + "epoch": 2.6028765481422296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6515 + }, + { + "epoch": 2.603276068717539, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6516 + }, + { + "epoch": 2.6036755892928487, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6517 + }, + { + "epoch": 2.604075109868158, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6518 + }, + { + "epoch": 2.604474630443468, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6519 + }, + { + "epoch": 2.6048741510187776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6520 + }, + { + "epoch": 2.605273671594087, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6521 + }, + { + "epoch": 2.6056731921693967, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6522 + }, + { + "epoch": 2.606072712744706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 6523 + }, + { + "epoch": 2.606472233320016, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6524 + }, + { + "epoch": 2.6068717538953257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6525 + }, + { + "epoch": 2.6072712744706354, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6526 + }, + { + "epoch": 2.607670795045945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6527 + }, + { + "epoch": 2.6080703156212546, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6528 + }, + { + "epoch": 2.608469836196564, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6529 + }, + { + "epoch": 2.6088693567718737, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 6530 + }, + { + "epoch": 2.6092688773471835, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6531 + }, + { + "epoch": 2.6096683979224933, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6532 + }, + { + "epoch": 2.6100679184978026, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6533 + }, + { + "epoch": 2.6104674390731124, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6534 + }, + { + "epoch": 2.6108669596484217, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6535 + }, + { + "epoch": 2.6112664802237315, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6536 + }, + { + "epoch": 2.6116660007990413, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6537 + }, + { + "epoch": 2.6120655213743507, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6538 + }, + { + "epoch": 2.6124650419496604, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6539 + }, + { + "epoch": 2.61286456252497, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6540 + }, + { + "epoch": 2.6132640831002796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 6541 + }, + { + "epoch": 2.6136636036755894, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6542 + }, + { + "epoch": 2.614063124250899, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6543 + }, + { + "epoch": 2.6144626448262085, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6544 + }, + { + "epoch": 2.6148621654015183, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6545 + }, + { + "epoch": 2.6152616859768276, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6546 + }, + { + "epoch": 2.6156612065521374, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6547 + }, + { + "epoch": 2.616060727127447, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 6548 + }, + { + "epoch": 2.6164602477027565, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6549 + }, + { + "epoch": 2.6168597682780663, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6550 + }, + { + "epoch": 2.617259288853376, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6551 + }, + { + "epoch": 2.6176588094286855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6552 + }, + { + "epoch": 2.6180583300039952, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6553 + }, + { + "epoch": 2.618457850579305, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6554 + }, + { + "epoch": 2.6188573711546144, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6555 + }, + { + "epoch": 2.619256891729924, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6556 + }, + { + "epoch": 2.6196564123052335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6557 + }, + { + "epoch": 2.6200559328805433, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6558 + }, + { + "epoch": 2.620455453455853, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 6559 + }, + { + "epoch": 2.620854974031163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6560 + }, + { + "epoch": 2.621254494606472, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6561 + }, + { + "epoch": 2.621654015181782, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6562 + }, + { + "epoch": 2.6220535357570913, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6563 + }, + { + "epoch": 2.622453056332401, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 6564 + }, + { + "epoch": 2.622852576907711, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6565 + }, + { + "epoch": 2.6232520974830202, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6566 + }, + { + "epoch": 2.62365161805833, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6567 + }, + { + "epoch": 2.6240511386336394, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6568 + }, + { + "epoch": 2.624450659208949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6569 + }, + { + "epoch": 2.624850179784259, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6570 + }, + { + "epoch": 2.6252497003595687, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6571 + }, + { + "epoch": 2.625649220934878, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6572 + }, + { + "epoch": 2.626048741510188, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6573 + }, + { + "epoch": 2.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6574 + }, + { + "epoch": 2.626847782660807, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6575 + }, + { + "epoch": 2.6272473032361168, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6576 + }, + { + "epoch": 2.6276468238114266, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 6577 + }, + { + "epoch": 2.628046344386736, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6578 + }, + { + "epoch": 2.6284458649620457, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6579 + }, + { + "epoch": 2.628845385537355, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6580 + }, + { + "epoch": 2.629244906112665, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6581 + }, + { + "epoch": 2.6296444266879746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6582 + }, + { + "epoch": 2.630043947263284, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6583 + }, + { + "epoch": 2.6304434678385937, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6584 + }, + { + "epoch": 2.630842988413903, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 6585 + }, + { + "epoch": 2.631242508989213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6586 + }, + { + "epoch": 2.6316420295645226, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6587 + }, + { + "epoch": 2.6320415501398324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6588 + }, + { + "epoch": 2.6324410707151418, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6589 + }, + { + "epoch": 2.6328405912904516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6590 + }, + { + "epoch": 2.633240111865761, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6591 + }, + { + "epoch": 2.6336396324410707, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6592 + }, + { + "epoch": 2.6340391530163805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6593 + }, + { + "epoch": 2.63443867359169, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6594 + }, + { + "epoch": 2.6348381941669996, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6595 + }, + { + "epoch": 2.6352377147423094, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6596 + }, + { + "epoch": 2.6356372353176187, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6597 + }, + { + "epoch": 2.6360367558929285, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6598 + }, + { + "epoch": 2.6364362764682383, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6599 + }, + { + "epoch": 2.6368357970435476, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6600 + }, + { + "epoch": 2.6372353176188574, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6601 + }, + { + "epoch": 2.6376348381941668, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6602 + }, + { + "epoch": 2.6380343587694766, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6603 + }, + { + "epoch": 2.6384338793447863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6604 + }, + { + "epoch": 2.638833399920096, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6605 + }, + { + "epoch": 2.6392329204954055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6606 + }, + { + "epoch": 2.6396324410707153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6607 + }, + { + "epoch": 2.6400319616460246, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6608 + }, + { + "epoch": 2.6404314822213344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6609 + }, + { + "epoch": 2.640831002796644, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6610 + }, + { + "epoch": 2.6412305233719535, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6611 + }, + { + "epoch": 2.6416300439472633, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6612 + }, + { + "epoch": 2.6420295645225726, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6613 + }, + { + "epoch": 2.6424290850978824, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6614 + }, + { + "epoch": 2.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6615 + }, + { + "epoch": 2.643228126248502, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6616 + }, + { + "epoch": 2.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6617 + }, + { + "epoch": 2.644027167399121, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6618 + }, + { + "epoch": 2.6444266879744305, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6619 + }, + { + "epoch": 2.6448262085497403, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6620 + }, + { + "epoch": 2.64522572912505, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6621 + }, + { + "epoch": 2.64562524970036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6622 + }, + { + "epoch": 2.646024770275669, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6623 + }, + { + "epoch": 2.646424290850979, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6624 + }, + { + "epoch": 2.6468238114262883, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6625 + }, + { + "epoch": 2.647223332001598, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6626 + }, + { + "epoch": 2.647622852576908, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6627 + }, + { + "epoch": 2.6480223731522172, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6628 + }, + { + "epoch": 2.648421893727527, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6629 + }, + { + "epoch": 2.6488214143028364, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6630 + }, + { + "epoch": 2.649220934878146, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6631 + }, + { + "epoch": 2.649620455453456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 6632 + }, + { + "epoch": 2.6500199760287657, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6633 + }, + { + "epoch": 2.650419496604075, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6634 + }, + { + "epoch": 2.650819017179385, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6635 + }, + { + "epoch": 2.651218537754694, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6636 + }, + { + "epoch": 2.651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6637 + }, + { + "epoch": 2.6520175789053138, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6638 + }, + { + "epoch": 2.652417099480623, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 6639 + }, + { + "epoch": 2.652816620055933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6640 + }, + { + "epoch": 2.6532161406312427, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6641 + }, + { + "epoch": 2.653615661206552, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6642 + }, + { + "epoch": 2.654015181781862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6643 + }, + { + "epoch": 2.6544147023571716, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6644 + }, + { + "epoch": 2.654814222932481, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6645 + }, + { + "epoch": 2.6552137435077907, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6646 + }, + { + "epoch": 2.6556132640831, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 6647 + }, + { + "epoch": 2.65601278465841, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6648 + }, + { + "epoch": 2.6564123052337196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6649 + }, + { + "epoch": 2.6568118258090294, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6650 + }, + { + "epoch": 2.6572113463843388, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6651 + }, + { + "epoch": 2.6576108669596485, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6652 + }, + { + "epoch": 2.658010387534958, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6653 + }, + { + "epoch": 2.6584099081102677, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6654 + }, + { + "epoch": 2.6588094286855775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6655 + }, + { + "epoch": 2.659208949260887, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6656 + }, + { + "epoch": 2.6596084698361966, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6657 + }, + { + "epoch": 2.660007990411506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6658 + }, + { + "epoch": 2.6604075109868157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6659 + }, + { + "epoch": 2.6608070315621255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6660 + }, + { + "epoch": 2.6612065521374353, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6661 + }, + { + "epoch": 2.6616060727127446, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6662 + }, + { + "epoch": 2.6620055932880544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6663 + }, + { + "epoch": 2.6624051138633638, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6664 + }, + { + "epoch": 2.6628046344386735, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6665 + }, + { + "epoch": 2.6632041550139833, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 6666 + }, + { + "epoch": 2.663603675589293, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6667 + }, + { + "epoch": 2.6640031961646025, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6668 + }, + { + "epoch": 2.6644027167399122, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6669 + }, + { + "epoch": 2.6648022373152216, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6670 + }, + { + "epoch": 2.6652017578905314, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6671 + }, + { + "epoch": 2.665601278465841, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6672 + }, + { + "epoch": 2.6660007990411505, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6673 + }, + { + "epoch": 2.6664003196164603, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6674 + }, + { + "epoch": 2.6667998401917696, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6675 + }, + { + "epoch": 2.6671993607670794, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6676 + }, + { + "epoch": 2.667598881342389, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6677 + }, + { + "epoch": 2.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6678 + }, + { + "epoch": 2.6683979224930083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6679 + }, + { + "epoch": 2.668797443068318, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6680 + }, + { + "epoch": 2.6691969636436275, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6681 + }, + { + "epoch": 2.6695964842189372, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6682 + }, + { + "epoch": 2.669996004794247, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6683 + }, + { + "epoch": 2.6703955253695564, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6684 + }, + { + "epoch": 2.670795045944866, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6685 + }, + { + "epoch": 2.671194566520176, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6686 + }, + { + "epoch": 2.6715940870954853, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 6687 + }, + { + "epoch": 2.671993607670795, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6688 + }, + { + "epoch": 2.672393128246105, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6689 + }, + { + "epoch": 2.672792648821414, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6690 + }, + { + "epoch": 2.673192169396724, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6691 + }, + { + "epoch": 2.6735916899720333, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6692 + }, + { + "epoch": 2.673991210547343, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6693 + }, + { + "epoch": 2.674390731122653, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6694 + }, + { + "epoch": 2.6747902516979627, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6695 + }, + { + "epoch": 2.675189772273272, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6696 + }, + { + "epoch": 2.675589292848582, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6697 + }, + { + "epoch": 2.675988813423891, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6698 + }, + { + "epoch": 2.676388333999201, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6699 + }, + { + "epoch": 2.6767878545745107, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6700 + }, + { + "epoch": 2.67718737514982, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6701 + }, + { + "epoch": 2.67758689572513, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6702 + }, + { + "epoch": 2.677986416300439, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 6703 + }, + { + "epoch": 2.678385936875749, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6704 + }, + { + "epoch": 2.678785457451059, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6705 + }, + { + "epoch": 2.6791849780263686, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6706 + }, + { + "epoch": 2.679584498601678, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6707 + }, + { + "epoch": 2.6799840191769877, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6708 + }, + { + "epoch": 2.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6709 + }, + { + "epoch": 2.680783060327607, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6710 + }, + { + "epoch": 2.6811825809029166, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6711 + }, + { + "epoch": 2.6815821014782264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6712 + }, + { + "epoch": 2.6819816220535357, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 6713 + }, + { + "epoch": 2.6823811426288455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6714 + }, + { + "epoch": 2.682780663204155, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6715 + }, + { + "epoch": 2.6831801837794647, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6716 + }, + { + "epoch": 2.6835797043547744, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6717 + }, + { + "epoch": 2.683979224930084, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6718 + }, + { + "epoch": 2.6843787455053936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6719 + }, + { + "epoch": 2.684778266080703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 6720 + }, + { + "epoch": 2.6851777866560127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6721 + }, + { + "epoch": 2.6855773072313225, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6722 + }, + { + "epoch": 2.6859768278066323, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6723 + }, + { + "epoch": 2.6863763483819416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6724 + }, + { + "epoch": 2.6867758689572514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6725 + }, + { + "epoch": 2.6871753895325607, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6726 + }, + { + "epoch": 2.6875749101078705, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6727 + }, + { + "epoch": 2.6879744306831803, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 6728 + }, + { + "epoch": 2.6883739512584897, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6729 + }, + { + "epoch": 2.6887734718337994, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6730 + }, + { + "epoch": 2.6891729924091092, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6731 + }, + { + "epoch": 2.6895725129844186, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6732 + }, + { + "epoch": 2.6899720335597284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6733 + }, + { + "epoch": 2.690371554135038, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6734 + }, + { + "epoch": 2.6907710747103475, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6735 + }, + { + "epoch": 2.6911705952856573, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6736 + }, + { + "epoch": 2.6915701158609666, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6737 + }, + { + "epoch": 2.6919696364362764, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6738 + }, + { + "epoch": 2.692369157011586, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6739 + }, + { + "epoch": 2.692768677586896, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 6740 + }, + { + "epoch": 2.6931681981622053, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6741 + }, + { + "epoch": 2.693567718737515, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6742 + }, + { + "epoch": 2.6939672393128244, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6743 + }, + { + "epoch": 2.6943667598881342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6744 + }, + { + "epoch": 2.694766280463444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6745 + }, + { + "epoch": 2.6951658010387534, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 6746 + }, + { + "epoch": 2.695565321614063, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6747 + }, + { + "epoch": 2.6959648421893725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6748 + }, + { + "epoch": 2.6963643627646823, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6749 + }, + { + "epoch": 2.696763883339992, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 6750 + }, + { + "epoch": 2.697163403915302, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6751 + }, + { + "epoch": 2.697562924490611, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6752 + }, + { + "epoch": 2.697962445065921, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6753 + }, + { + "epoch": 2.6983619656412303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6754 + }, + { + "epoch": 2.69876148621654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6755 + }, + { + "epoch": 2.69916100679185, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6756 + }, + { + "epoch": 2.6995605273671597, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6757 + }, + { + "epoch": 2.699960047942469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6758 + }, + { + "epoch": 2.700359568517779, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6759 + }, + { + "epoch": 2.700759089093088, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6760 + }, + { + "epoch": 2.701158609668398, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6761 + }, + { + "epoch": 2.7015581302437077, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6762 + }, + { + "epoch": 2.701957650819017, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6763 + }, + { + "epoch": 2.702357171394327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6764 + }, + { + "epoch": 2.702756691969636, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6765 + }, + { + "epoch": 2.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6766 + }, + { + "epoch": 2.7035557331202558, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6767 + }, + { + "epoch": 2.7039552536955656, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6768 + }, + { + "epoch": 2.704354774270875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6769 + }, + { + "epoch": 2.7047542948461847, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 6770 + }, + { + "epoch": 2.705153815421494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6771 + }, + { + "epoch": 2.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6772 + }, + { + "epoch": 2.7059528565721136, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6773 + }, + { + "epoch": 2.706352377147423, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6774 + }, + { + "epoch": 2.7067518977227327, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6775 + }, + { + "epoch": 2.7071514182980425, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6776 + }, + { + "epoch": 2.707550938873352, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6777 + }, + { + "epoch": 2.7079504594486616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6778 + }, + { + "epoch": 2.7083499800239714, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6779 + }, + { + "epoch": 2.7087495005992808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6780 + }, + { + "epoch": 2.7091490211745906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 6781 + }, + { + "epoch": 2.7095485417499, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6782 + }, + { + "epoch": 2.7099480623252097, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6783 + }, + { + "epoch": 2.7103475829005195, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 6784 + }, + { + "epoch": 2.7107471034758293, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6785 + }, + { + "epoch": 2.7111466240511386, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6786 + }, + { + "epoch": 2.7115461446264484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 6787 + }, + { + "epoch": 2.7119456652017577, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6788 + }, + { + "epoch": 2.7123451857770675, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6789 + }, + { + "epoch": 2.7127447063523773, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6790 + }, + { + "epoch": 2.7131442269276866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 6791 + }, + { + "epoch": 2.7135437475029964, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6792 + }, + { + "epoch": 2.7139432680783058, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6793 + }, + { + "epoch": 2.7143427886536156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6794 + }, + { + "epoch": 2.7147423092289253, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6795 + }, + { + "epoch": 2.715141829804235, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6796 + }, + { + "epoch": 2.7155413503795445, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6797 + }, + { + "epoch": 2.7159408709548543, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 6798 + }, + { + "epoch": 2.7163403915301636, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6799 + }, + { + "epoch": 2.7167399121054734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6800 + }, + { + "epoch": 2.717139432680783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6801 + }, + { + "epoch": 2.717538953256093, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6802 + }, + { + "epoch": 2.7179384738314023, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6803 + }, + { + "epoch": 2.718337994406712, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6804 + }, + { + "epoch": 2.7187375149820214, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6805 + }, + { + "epoch": 2.719137035557331, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6806 + }, + { + "epoch": 2.719536556132641, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6807 + }, + { + "epoch": 2.7199360767079503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6808 + }, + { + "epoch": 2.72033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6809 + }, + { + "epoch": 2.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6810 + }, + { + "epoch": 2.7211346384338793, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6811 + }, + { + "epoch": 2.721534159009189, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6812 + }, + { + "epoch": 2.721933679584499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6813 + }, + { + "epoch": 2.722333200159808, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6814 + }, + { + "epoch": 2.722732720735118, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6815 + }, + { + "epoch": 2.7231322413104273, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6816 + }, + { + "epoch": 2.723531761885737, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6817 + }, + { + "epoch": 2.723931282461047, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6818 + }, + { + "epoch": 2.724330803036356, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6819 + }, + { + "epoch": 2.724730323611666, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6820 + }, + { + "epoch": 2.725129844186976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6821 + }, + { + "epoch": 2.725529364762285, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6822 + }, + { + "epoch": 2.725928885337595, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6823 + }, + { + "epoch": 2.7263284059129047, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6824 + }, + { + "epoch": 2.726727926488214, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6825 + }, + { + "epoch": 2.727127447063524, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6826 + }, + { + "epoch": 2.727526967638833, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6827 + }, + { + "epoch": 2.727926488214143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6828 + }, + { + "epoch": 2.7283260087894528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6829 + }, + { + "epoch": 2.7287255293647625, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6830 + }, + { + "epoch": 2.729125049940072, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6831 + }, + { + "epoch": 2.7295245705153817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6832 + }, + { + "epoch": 2.729924091090691, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6833 + }, + { + "epoch": 2.730323611666001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6834 + }, + { + "epoch": 2.7307231322413106, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6835 + }, + { + "epoch": 2.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 6836 + }, + { + "epoch": 2.7315221733919297, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6837 + }, + { + "epoch": 2.731921693967239, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6838 + }, + { + "epoch": 2.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6839 + }, + { + "epoch": 2.7327207351178586, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6840 + }, + { + "epoch": 2.7331202556931684, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6841 + }, + { + "epoch": 2.7335197762684778, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6842 + }, + { + "epoch": 2.7339192968437875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6843 + }, + { + "epoch": 2.734318817419097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6844 + }, + { + "epoch": 2.7347183379944067, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6845 + }, + { + "epoch": 2.7351178585697165, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6846 + }, + { + "epoch": 2.7355173791450262, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6847 + }, + { + "epoch": 2.7359168997203356, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6848 + }, + { + "epoch": 2.7363164202956454, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6849 + }, + { + "epoch": 2.7367159408709547, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6850 + }, + { + "epoch": 2.7371154614462645, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6851 + }, + { + "epoch": 2.7375149820215743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6852 + }, + { + "epoch": 2.7379145025968836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6853 + }, + { + "epoch": 2.7383140231721934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6854 + }, + { + "epoch": 2.7387135437475028, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6855 + }, + { + "epoch": 2.7391130643228125, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6856 + }, + { + "epoch": 2.7395125848981223, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6857 + }, + { + "epoch": 2.739912105473432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6858 + }, + { + "epoch": 2.7403116260487415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6859 + }, + { + "epoch": 2.7407111466240512, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6860 + }, + { + "epoch": 2.7411106671993606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6861 + }, + { + "epoch": 2.7415101877746704, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6862 + }, + { + "epoch": 2.74190970834998, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6863 + }, + { + "epoch": 2.7423092289252895, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6864 + }, + { + "epoch": 2.7427087495005993, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6865 + }, + { + "epoch": 2.743108270075909, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6866 + }, + { + "epoch": 2.7435077906512184, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6867 + }, + { + "epoch": 2.743907311226528, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6868 + }, + { + "epoch": 2.744306831801838, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6869 + }, + { + "epoch": 2.7447063523771473, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6870 + }, + { + "epoch": 2.745105872952457, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6871 + }, + { + "epoch": 2.7455053935277665, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6872 + }, + { + "epoch": 2.7459049141030762, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6873 + }, + { + "epoch": 2.746304434678386, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6874 + }, + { + "epoch": 2.746703955253696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6875 + }, + { + "epoch": 2.747103475829005, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6876 + }, + { + "epoch": 2.747502996404315, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 6877 + }, + { + "epoch": 2.7479025169796243, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6878 + }, + { + "epoch": 2.748302037554934, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6879 + }, + { + "epoch": 2.748701558130244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6880 + }, + { + "epoch": 2.749101078705553, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6881 + }, + { + "epoch": 2.749500599280863, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6882 + }, + { + "epoch": 2.7499001198561723, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6883 + }, + { + "epoch": 2.750299640431482, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6884 + }, + { + "epoch": 2.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6885 + }, + { + "epoch": 2.7510986815821017, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6886 + }, + { + "epoch": 2.751498202157411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6887 + }, + { + "epoch": 2.751897722732721, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6888 + }, + { + "epoch": 2.75229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6889 + }, + { + "epoch": 2.75269676388334, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6890 + }, + { + "epoch": 2.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6891 + }, + { + "epoch": 2.7534958050339595, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6892 + }, + { + "epoch": 2.753895325609269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6893 + }, + { + "epoch": 2.7542948461845786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6894 + }, + { + "epoch": 2.754694366759888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6895 + }, + { + "epoch": 2.7550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6896 + }, + { + "epoch": 2.7554934079105076, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6897 + }, + { + "epoch": 2.755892928485817, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6898 + }, + { + "epoch": 2.7562924490611267, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6899 + }, + { + "epoch": 2.756691969636436, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6900 + }, + { + "epoch": 2.757091490211746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6901 + }, + { + "epoch": 2.7574910107870556, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6902 + }, + { + "epoch": 2.7578905313623654, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6903 + }, + { + "epoch": 2.7582900519376747, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6904 + }, + { + "epoch": 2.7586895725129845, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6905 + }, + { + "epoch": 2.759089093088294, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6906 + }, + { + "epoch": 2.7594886136636037, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6907 + }, + { + "epoch": 2.7598881342389134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6908 + }, + { + "epoch": 2.760287654814223, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6909 + }, + { + "epoch": 2.7606871753895326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6910 + }, + { + "epoch": 2.7610866959648424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6911 + }, + { + "epoch": 2.7614862165401517, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6912 + }, + { + "epoch": 2.7618857371154615, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6913 + }, + { + "epoch": 2.7622852576907713, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6914 + }, + { + "epoch": 2.7626847782660806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6915 + }, + { + "epoch": 2.7630842988413904, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6916 + }, + { + "epoch": 2.7634838194166997, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6917 + }, + { + "epoch": 2.7638833399920095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6918 + }, + { + "epoch": 2.7642828605673193, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6919 + }, + { + "epoch": 2.764682381142629, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6920 + }, + { + "epoch": 2.7650819017179384, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6921 + }, + { + "epoch": 2.7654814222932482, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6922 + }, + { + "epoch": 2.7658809428685576, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6923 + }, + { + "epoch": 2.7662804634438674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6924 + }, + { + "epoch": 2.766679984019177, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6925 + }, + { + "epoch": 2.7670795045944865, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6926 + }, + { + "epoch": 2.7674790251697963, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6927 + }, + { + "epoch": 2.7678785457451056, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6928 + }, + { + "epoch": 2.7682780663204154, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6929 + }, + { + "epoch": 2.768677586895725, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6930 + }, + { + "epoch": 2.769077107471035, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6931 + }, + { + "epoch": 2.7694766280463443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6932 + }, + { + "epoch": 2.769876148621654, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6933 + }, + { + "epoch": 2.7702756691969634, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6934 + }, + { + "epoch": 2.7706751897722732, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6935 + }, + { + "epoch": 2.771074710347583, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6936 + }, + { + "epoch": 2.771474230922893, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6937 + }, + { + "epoch": 2.771873751498202, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6938 + }, + { + "epoch": 2.772273272073512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6939 + }, + { + "epoch": 2.7726727926488213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6940 + }, + { + "epoch": 2.773072313224131, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6941 + }, + { + "epoch": 2.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6942 + }, + { + "epoch": 2.77387135437475, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6943 + }, + { + "epoch": 2.77427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6944 + }, + { + "epoch": 2.7746703955253693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 6945 + }, + { + "epoch": 2.775069916100679, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6946 + }, + { + "epoch": 2.775469436675989, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6947 + }, + { + "epoch": 2.7758689572512987, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 6948 + }, + { + "epoch": 2.776268477826608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6949 + }, + { + "epoch": 2.776667998401918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6950 + }, + { + "epoch": 2.777067518977227, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6951 + }, + { + "epoch": 2.777467039552537, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6952 + }, + { + "epoch": 2.7778665601278467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6953 + }, + { + "epoch": 2.778266080703156, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6954 + }, + { + "epoch": 2.778665601278466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6955 + }, + { + "epoch": 2.7790651218537756, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6956 + }, + { + "epoch": 2.779464642429085, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6957 + }, + { + "epoch": 2.7798641630043948, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6958 + }, + { + "epoch": 2.7802636835797045, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6959 + }, + { + "epoch": 2.780663204155014, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6960 + }, + { + "epoch": 2.7810627247303237, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6961 + }, + { + "epoch": 2.781462245305633, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6962 + }, + { + "epoch": 2.781861765880943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6963 + }, + { + "epoch": 2.7822612864562526, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6964 + }, + { + "epoch": 2.7826608070315624, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6965 + }, + { + "epoch": 2.7830603276068717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6966 + }, + { + "epoch": 2.7834598481821815, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6967 + }, + { + "epoch": 2.783859368757491, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6968 + }, + { + "epoch": 2.7842588893328006, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6969 + }, + { + "epoch": 2.7846584099081104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6970 + }, + { + "epoch": 2.7850579304834198, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 6971 + }, + { + "epoch": 2.7854574510587295, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6972 + }, + { + "epoch": 2.785856971634039, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6973 + }, + { + "epoch": 2.7862564922093487, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6974 + }, + { + "epoch": 2.7866560127846585, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6975 + }, + { + "epoch": 2.7870555333599683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6976 + }, + { + "epoch": 2.7874550539352776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6977 + }, + { + "epoch": 2.7878545745105874, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 6978 + }, + { + "epoch": 2.7882540950858967, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6979 + }, + { + "epoch": 2.7886536156612065, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6980 + }, + { + "epoch": 2.7890531362365163, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6981 + }, + { + "epoch": 2.789452656811826, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6982 + }, + { + "epoch": 2.7898521773871354, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6983 + }, + { + "epoch": 2.790251697962445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 6984 + }, + { + "epoch": 2.7906512185377546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6985 + }, + { + "epoch": 2.7910507391130643, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6986 + }, + { + "epoch": 2.791450259688374, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6987 + }, + { + "epoch": 2.7918497802636835, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6988 + }, + { + "epoch": 2.7922493008389933, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6989 + }, + { + "epoch": 2.7926488214143026, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6990 + }, + { + "epoch": 2.7930483419896124, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6991 + }, + { + "epoch": 2.793447862564922, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6992 + }, + { + "epoch": 2.793847383140232, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6993 + }, + { + "epoch": 2.7942469037155413, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6994 + }, + { + "epoch": 2.794646424290851, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6995 + }, + { + "epoch": 2.7950459448661604, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 6996 + }, + { + "epoch": 2.79544546544147, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6997 + }, + { + "epoch": 2.79584498601678, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6998 + }, + { + "epoch": 2.7962445065920893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6999 + }, + { + "epoch": 2.796644027167399, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7000 + }, + { + "epoch": 2.797043547742709, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7001 + }, + { + "epoch": 2.7974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7002 + }, + { + "epoch": 2.797842588893328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7003 + }, + { + "epoch": 2.798242109468638, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7004 + }, + { + "epoch": 2.798641630043947, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7005 + }, + { + "epoch": 2.799041150619257, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7006 + }, + { + "epoch": 2.7994406711945663, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7007 + }, + { + "epoch": 2.799840191769876, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7008 + }, + { + "epoch": 2.800239712345186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 7009 + }, + { + "epoch": 2.8006392329204957, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7010 + }, + { + "epoch": 2.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7011 + }, + { + "epoch": 2.801438274071115, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7012 + }, + { + "epoch": 2.801837794646424, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7013 + }, + { + "epoch": 2.802237315221734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7014 + }, + { + "epoch": 2.8026368357970437, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7015 + }, + { + "epoch": 2.803036356372353, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7016 + }, + { + "epoch": 2.803435876947663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7017 + }, + { + "epoch": 2.803835397522972, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7018 + }, + { + "epoch": 2.804234918098282, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7019 + }, + { + "epoch": 2.8046344386735917, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7020 + }, + { + "epoch": 2.8050339592489015, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7021 + }, + { + "epoch": 2.805433479824211, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7022 + }, + { + "epoch": 2.8058330003995207, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7023 + }, + { + "epoch": 2.80623252097483, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 7024 + }, + { + "epoch": 2.80663204155014, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7025 + }, + { + "epoch": 2.8070315621254496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7026 + }, + { + "epoch": 2.8074310827007594, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7027 + }, + { + "epoch": 2.8078306032760687, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7028 + }, + { + "epoch": 2.8082301238513785, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7029 + }, + { + "epoch": 2.808629644426688, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7030 + }, + { + "epoch": 2.8090291650019976, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7031 + }, + { + "epoch": 2.8094286855773074, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7032 + }, + { + "epoch": 2.8098282061526167, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7033 + }, + { + "epoch": 2.8102277267279265, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7034 + }, + { + "epoch": 2.810627247303236, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7035 + }, + { + "epoch": 2.8110267678785457, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7036 + }, + { + "epoch": 2.8114262884538554, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7037 + }, + { + "epoch": 2.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7038 + }, + { + "epoch": 2.8122253296044746, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7039 + }, + { + "epoch": 2.8126248501797844, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7040 + }, + { + "epoch": 2.8130243707550937, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7041 + }, + { + "epoch": 2.8134238913304035, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7042 + }, + { + "epoch": 2.8138234119057133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7043 + }, + { + "epoch": 2.8142229324810226, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7044 + }, + { + "epoch": 2.8146224530563324, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7045 + }, + { + "epoch": 2.815021973631642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 7046 + }, + { + "epoch": 2.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7047 + }, + { + "epoch": 2.8158210147822613, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 7048 + }, + { + "epoch": 2.816220535357571, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7049 + }, + { + "epoch": 2.8166200559328805, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7050 + }, + { + "epoch": 2.8170195765081902, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7051 + }, + { + "epoch": 2.8174190970834996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7052 + }, + { + "epoch": 2.8178186176588094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 7053 + }, + { + "epoch": 2.818218138234119, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7054 + }, + { + "epoch": 2.818617658809429, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7055 + }, + { + "epoch": 2.8190171793847383, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7056 + }, + { + "epoch": 2.819416699960048, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7057 + }, + { + "epoch": 2.8198162205353574, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7058 + }, + { + "epoch": 2.820215741110667, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7059 + }, + { + "epoch": 2.820615261685977, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7060 + }, + { + "epoch": 2.8210147822612863, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7061 + }, + { + "epoch": 2.821414302836596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7062 + }, + { + "epoch": 2.8218138234119055, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7063 + }, + { + "epoch": 2.8222133439872152, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7064 + }, + { + "epoch": 2.822612864562525, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7065 + }, + { + "epoch": 2.823012385137835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7066 + }, + { + "epoch": 2.823411905713144, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7067 + }, + { + "epoch": 2.823811426288454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7068 + }, + { + "epoch": 2.8242109468637633, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7069 + }, + { + "epoch": 2.824610467439073, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7070 + }, + { + "epoch": 2.825009988014383, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7071 + }, + { + "epoch": 2.8254095085896926, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7072 + }, + { + "epoch": 2.825809029165002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7073 + }, + { + "epoch": 2.8262085497403118, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 7074 + }, + { + "epoch": 2.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7075 + }, + { + "epoch": 2.827007590890931, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 7076 + }, + { + "epoch": 2.8274071114662407, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7077 + }, + { + "epoch": 2.82780663204155, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 7078 + }, + { + "epoch": 2.82820615261686, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7079 + }, + { + "epoch": 2.828605673192169, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7080 + }, + { + "epoch": 2.829005193767479, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7081 + }, + { + "epoch": 2.8294047143427887, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7082 + }, + { + "epoch": 2.8298042349180985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7083 + }, + { + "epoch": 2.830203755493408, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7084 + }, + { + "epoch": 2.8306032760687176, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 7085 + }, + { + "epoch": 2.831002796644027, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7086 + }, + { + "epoch": 2.8314023172193368, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7087 + }, + { + "epoch": 2.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7088 + }, + { + "epoch": 2.832201358369956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7089 + }, + { + "epoch": 2.8326008789452657, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7090 + }, + { + "epoch": 2.8330003995205755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 7091 + }, + { + "epoch": 2.833399920095885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7092 + }, + { + "epoch": 2.8337994406711946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7093 + }, + { + "epoch": 2.8341989612465044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7094 + }, + { + "epoch": 2.8345984818218137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 7095 + }, + { + "epoch": 2.8349980023971235, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7096 + }, + { + "epoch": 2.835397522972433, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7097 + }, + { + "epoch": 2.8357970435477426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7098 + }, + { + "epoch": 2.8361965641230524, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7099 + }, + { + "epoch": 2.836596084698362, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7100 + }, + { + "epoch": 2.8369956052736716, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7101 + }, + { + "epoch": 2.8373951258489813, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7102 + }, + { + "epoch": 2.8377946464242907, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7103 + }, + { + "epoch": 2.8381941669996005, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7104 + }, + { + "epoch": 2.8385936875749103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7105 + }, + { + "epoch": 2.8389932081502196, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7106 + }, + { + "epoch": 2.8393927287255294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7107 + }, + { + "epoch": 2.8397922493008387, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7108 + }, + { + "epoch": 2.8401917698761485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7109 + }, + { + "epoch": 2.8405912904514583, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7110 + }, + { + "epoch": 2.840990811026768, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7111 + }, + { + "epoch": 2.8413903316020774, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7112 + }, + { + "epoch": 2.841789852177387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7113 + }, + { + "epoch": 2.8421893727526966, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7114 + }, + { + "epoch": 2.8425888933280063, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7115 + }, + { + "epoch": 2.842988413903316, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 7116 + }, + { + "epoch": 2.843387934478626, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7117 + }, + { + "epoch": 2.8437874550539353, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7118 + }, + { + "epoch": 2.844186975629245, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7119 + }, + { + "epoch": 2.8445864962045544, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7120 + }, + { + "epoch": 2.844986016779864, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7121 + }, + { + "epoch": 2.845385537355174, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7122 + }, + { + "epoch": 2.8457850579304833, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7123 + }, + { + "epoch": 2.846184578505793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 7124 + }, + { + "epoch": 2.8465840990811024, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7125 + }, + { + "epoch": 2.8469836196564122, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7126 + }, + { + "epoch": 2.847383140231722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7127 + }, + { + "epoch": 2.847782660807032, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7128 + }, + { + "epoch": 2.848182181382341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7129 + }, + { + "epoch": 2.848581701957651, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7130 + }, + { + "epoch": 2.8489812225329603, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7131 + }, + { + "epoch": 2.84938074310827, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7132 + }, + { + "epoch": 2.84978026368358, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7133 + }, + { + "epoch": 2.850179784258889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7134 + }, + { + "epoch": 2.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7135 + }, + { + "epoch": 2.8509788254095088, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7136 + }, + { + "epoch": 2.851378345984818, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 7137 + }, + { + "epoch": 2.851777866560128, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7138 + }, + { + "epoch": 2.8521773871354377, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7139 + }, + { + "epoch": 2.852576907710747, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7140 + }, + { + "epoch": 2.852976428286057, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7141 + }, + { + "epoch": 2.853375948861366, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7142 + }, + { + "epoch": 2.853775469436676, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 7143 + }, + { + "epoch": 2.8541749900119857, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7144 + }, + { + "epoch": 2.8545745105872955, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7145 + }, + { + "epoch": 2.854974031162605, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7146 + }, + { + "epoch": 2.8553735517379146, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7147 + }, + { + "epoch": 2.855773072313224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7148 + }, + { + "epoch": 2.8561725928885338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7149 + }, + { + "epoch": 2.8565721134638435, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7150 + }, + { + "epoch": 2.856971634039153, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7151 + }, + { + "epoch": 2.8573711546144627, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7152 + }, + { + "epoch": 2.857770675189772, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7153 + }, + { + "epoch": 2.858170195765082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7154 + }, + { + "epoch": 2.8585697163403916, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7155 + }, + { + "epoch": 2.8589692369157014, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7156 + }, + { + "epoch": 2.8593687574910107, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7157 + }, + { + "epoch": 2.8597682780663205, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7158 + }, + { + "epoch": 2.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7159 + }, + { + "epoch": 2.8605673192169396, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7160 + }, + { + "epoch": 2.8609668397922494, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7161 + }, + { + "epoch": 2.861366360367559, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7162 + }, + { + "epoch": 2.8617658809428685, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7163 + }, + { + "epoch": 2.8621654015181783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7164 + }, + { + "epoch": 2.8625649220934877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 7165 + }, + { + "epoch": 2.8629644426687975, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7166 + }, + { + "epoch": 2.8633639632441072, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7167 + }, + { + "epoch": 2.8637634838194166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7168 + }, + { + "epoch": 2.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7169 + }, + { + "epoch": 2.8645625249700357, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7170 + }, + { + "epoch": 2.8649620455453455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7171 + }, + { + "epoch": 2.8653615661206553, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 7172 + }, + { + "epoch": 2.865761086695965, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7173 + }, + { + "epoch": 2.8661606072712744, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7174 + }, + { + "epoch": 2.866560127846584, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7175 + }, + { + "epoch": 2.8669596484218935, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7176 + }, + { + "epoch": 2.8673591689972033, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7177 + }, + { + "epoch": 2.867758689572513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7178 + }, + { + "epoch": 2.8681582101478225, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7179 + }, + { + "epoch": 2.8685577307231322, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7180 + }, + { + "epoch": 2.868957251298442, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7181 + }, + { + "epoch": 2.8693567718737514, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7182 + }, + { + "epoch": 2.869756292449061, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7183 + }, + { + "epoch": 2.870155813024371, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7184 + }, + { + "epoch": 2.8705553335996803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7185 + }, + { + "epoch": 2.87095485417499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7186 + }, + { + "epoch": 2.8713543747502994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7187 + }, + { + "epoch": 2.871753895325609, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7188 + }, + { + "epoch": 2.872153415900919, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7189 + }, + { + "epoch": 2.872552936476229, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7190 + }, + { + "epoch": 2.872952457051538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7191 + }, + { + "epoch": 2.873351977626848, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7192 + }, + { + "epoch": 2.8737514982021573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7193 + }, + { + "epoch": 2.874151018777467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7194 + }, + { + "epoch": 2.874550539352777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7195 + }, + { + "epoch": 2.874950059928086, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 7196 + }, + { + "epoch": 2.875349580503396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7197 + }, + { + "epoch": 2.8757491010787053, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7198 + }, + { + "epoch": 2.876148621654015, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7199 + }, + { + "epoch": 2.876548142229325, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 7200 + }, + { + "epoch": 2.8769476628046347, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7201 + }, + { + "epoch": 2.877347183379944, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7202 + }, + { + "epoch": 2.877746703955254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7203 + }, + { + "epoch": 2.878146224530563, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7204 + }, + { + "epoch": 2.878545745105873, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7205 + }, + { + "epoch": 2.8789452656811827, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 7206 + }, + { + "epoch": 2.8793447862564925, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7207 + }, + { + "epoch": 2.879744306831802, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7208 + }, + { + "epoch": 2.8801438274071116, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7209 + }, + { + "epoch": 2.880543347982421, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 7210 + }, + { + "epoch": 2.8809428685577307, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7211 + }, + { + "epoch": 2.8813423891330405, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7212 + }, + { + "epoch": 2.88174190970835, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7213 + }, + { + "epoch": 2.8821414302836597, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7214 + }, + { + "epoch": 2.882540950858969, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7215 + }, + { + "epoch": 2.882940471434279, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7216 + }, + { + "epoch": 2.8833399920095886, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7217 + }, + { + "epoch": 2.8837395125848984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 7218 + }, + { + "epoch": 2.8841390331602077, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7219 + }, + { + "epoch": 2.8845385537355175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7220 + }, + { + "epoch": 2.884938074310827, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7221 + }, + { + "epoch": 2.8853375948861366, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7222 + }, + { + "epoch": 2.8857371154614464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7223 + }, + { + "epoch": 2.8861366360367557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7224 + }, + { + "epoch": 2.8865361566120655, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7225 + }, + { + "epoch": 2.8869356771873753, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7226 + }, + { + "epoch": 2.8873351977626847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7227 + }, + { + "epoch": 2.8877347183379944, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7228 + }, + { + "epoch": 2.8881342389133042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7229 + }, + { + "epoch": 2.8885337594886136, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7230 + }, + { + "epoch": 2.8889332800639234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7231 + }, + { + "epoch": 2.8893328006392327, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7232 + }, + { + "epoch": 2.8897323212145425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 7233 + }, + { + "epoch": 2.8901318417898523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7234 + }, + { + "epoch": 2.890531362365162, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7235 + }, + { + "epoch": 2.8909308829404714, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7236 + }, + { + "epoch": 2.891330403515781, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7237 + }, + { + "epoch": 2.8917299240910905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7238 + }, + { + "epoch": 2.8921294446664003, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7239 + }, + { + "epoch": 2.89252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7240 + }, + { + "epoch": 2.8929284858170194, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7241 + }, + { + "epoch": 2.8933280063923292, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7242 + }, + { + "epoch": 2.8937275269676386, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7243 + }, + { + "epoch": 2.8941270475429484, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7244 + }, + { + "epoch": 2.894526568118258, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 7245 + }, + { + "epoch": 2.894926088693568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7246 + }, + { + "epoch": 2.8953256092688773, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7247 + }, + { + "epoch": 2.895725129844187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7248 + }, + { + "epoch": 2.8961246504194964, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7249 + }, + { + "epoch": 2.896524170994806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 7250 + }, + { + "epoch": 2.896923691570116, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 7251 + }, + { + "epoch": 2.8973232121454258, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7252 + }, + { + "epoch": 2.897722732720735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7253 + }, + { + "epoch": 2.898122253296045, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7254 + }, + { + "epoch": 2.8985217738713542, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7255 + }, + { + "epoch": 2.898921294446664, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 7256 + }, + { + "epoch": 2.899320815021974, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7257 + }, + { + "epoch": 2.899720335597283, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7258 + }, + { + "epoch": 2.900119856172593, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7259 + }, + { + "epoch": 2.9005193767479023, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7260 + }, + { + "epoch": 2.900918897323212, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7261 + }, + { + "epoch": 2.901318417898522, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7262 + }, + { + "epoch": 2.9017179384738316, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7263 + }, + { + "epoch": 2.902117459049141, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7264 + }, + { + "epoch": 2.9025169796244508, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7265 + }, + { + "epoch": 2.90291650019976, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 7266 + }, + { + "epoch": 2.90331602077507, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7267 + }, + { + "epoch": 2.9037155413503797, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7268 + }, + { + "epoch": 2.904115061925689, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7269 + }, + { + "epoch": 2.904514582500999, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7270 + }, + { + "epoch": 2.9049141030763086, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7271 + }, + { + "epoch": 2.905313623651618, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7272 + }, + { + "epoch": 2.9057131442269277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7273 + }, + { + "epoch": 2.9061126648022375, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7274 + }, + { + "epoch": 2.906512185377547, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7275 + }, + { + "epoch": 2.9069117059528566, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7276 + }, + { + "epoch": 2.907311226528166, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7277 + }, + { + "epoch": 2.9077107471034758, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7278 + }, + { + "epoch": 2.9081102676787856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7279 + }, + { + "epoch": 2.9085097882540953, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7280 + }, + { + "epoch": 2.9089093088294047, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 7281 + }, + { + "epoch": 2.9093088294047145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7282 + }, + { + "epoch": 2.909708349980024, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7283 + }, + { + "epoch": 2.9101078705553336, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7284 + }, + { + "epoch": 2.9105073911306434, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7285 + }, + { + "epoch": 2.9109069117059527, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7286 + }, + { + "epoch": 2.9113064322812625, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7287 + }, + { + "epoch": 2.911705952856572, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 7288 + }, + { + "epoch": 2.9121054734318816, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7289 + }, + { + "epoch": 2.9125049940071914, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7290 + }, + { + "epoch": 2.912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7291 + }, + { + "epoch": 2.9133040351578106, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7292 + }, + { + "epoch": 2.9137035557331203, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7293 + }, + { + "epoch": 2.9141030763084297, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7294 + }, + { + "epoch": 2.9145025968837395, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7295 + }, + { + "epoch": 2.9149021174590493, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7296 + }, + { + "epoch": 2.915301638034359, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7297 + }, + { + "epoch": 2.9157011586096684, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7298 + }, + { + "epoch": 2.916100679184978, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7299 + }, + { + "epoch": 2.9165001997602875, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7300 + }, + { + "epoch": 2.9168997203355973, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7301 + }, + { + "epoch": 2.917299240910907, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7302 + }, + { + "epoch": 2.9176987614862164, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7303 + }, + { + "epoch": 2.918098282061526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7304 + }, + { + "epoch": 2.9184978026368356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7305 + }, + { + "epoch": 2.9188973232121453, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7306 + }, + { + "epoch": 2.919296843787455, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7307 + }, + { + "epoch": 2.919696364362765, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7308 + }, + { + "epoch": 2.9200958849380743, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7309 + }, + { + "epoch": 2.920495405513384, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7310 + }, + { + "epoch": 2.9208949260886934, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 7311 + }, + { + "epoch": 2.921294446664003, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7312 + }, + { + "epoch": 2.921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7313 + }, + { + "epoch": 2.9220934878146223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 7314 + }, + { + "epoch": 2.922493008389932, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7315 + }, + { + "epoch": 2.922892528965242, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7316 + }, + { + "epoch": 2.923292049540551, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7317 + }, + { + "epoch": 2.923691570115861, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7318 + }, + { + "epoch": 2.924091090691171, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7319 + }, + { + "epoch": 2.92449061126648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7320 + }, + { + "epoch": 2.92489013184179, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7321 + }, + { + "epoch": 2.9252896524170993, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7322 + }, + { + "epoch": 2.925689172992409, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7323 + }, + { + "epoch": 2.926088693567719, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7324 + }, + { + "epoch": 2.9264882141430286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7325 + }, + { + "epoch": 2.926887734718338, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7326 + }, + { + "epoch": 2.9272872552936477, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7327 + }, + { + "epoch": 2.927686775868957, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7328 + }, + { + "epoch": 2.928086296444267, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7329 + }, + { + "epoch": 2.9284858170195767, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7330 + }, + { + "epoch": 2.928885337594886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 7331 + }, + { + "epoch": 2.929284858170196, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7332 + }, + { + "epoch": 2.929684378745505, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7333 + }, + { + "epoch": 2.930083899320815, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7334 + }, + { + "epoch": 2.9304834198961247, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7335 + }, + { + "epoch": 2.9308829404714345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7336 + }, + { + "epoch": 2.931282461046744, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 7337 + }, + { + "epoch": 2.9316819816220536, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7338 + }, + { + "epoch": 2.932081502197363, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7339 + }, + { + "epoch": 2.9324810227726728, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7340 + }, + { + "epoch": 2.9328805433479825, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7341 + }, + { + "epoch": 2.9332800639232923, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7342 + }, + { + "epoch": 2.9336795844986017, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7343 + }, + { + "epoch": 2.9340791050739115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7344 + }, + { + "epoch": 2.934478625649221, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7345 + }, + { + "epoch": 2.9348781462245306, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7346 + }, + { + "epoch": 2.9352776667998404, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7347 + }, + { + "epoch": 2.9356771873751497, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7348 + }, + { + "epoch": 2.9360767079504595, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7349 + }, + { + "epoch": 2.936476228525769, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7350 + }, + { + "epoch": 2.9368757491010786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7351 + }, + { + "epoch": 2.9372752696763884, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7352 + }, + { + "epoch": 2.937674790251698, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7353 + }, + { + "epoch": 2.9380743108270075, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7354 + }, + { + "epoch": 2.9384738314023173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7355 + }, + { + "epoch": 2.9388733519776267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7356 + }, + { + "epoch": 2.9392728725529365, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7357 + }, + { + "epoch": 2.9396723931282462, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7358 + }, + { + "epoch": 2.9400719137035556, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7359 + }, + { + "epoch": 2.9404714342788654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7360 + }, + { + "epoch": 2.9408709548541747, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7361 + }, + { + "epoch": 2.9412704754294845, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7362 + }, + { + "epoch": 2.9416699960047943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7363 + }, + { + "epoch": 2.942069516580104, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7364 + }, + { + "epoch": 2.9424690371554134, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7365 + }, + { + "epoch": 2.942868557730723, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7366 + }, + { + "epoch": 2.9432680783060325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7367 + }, + { + "epoch": 2.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7368 + }, + { + "epoch": 2.944067119456652, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7369 + }, + { + "epoch": 2.944466640031962, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7370 + }, + { + "epoch": 2.9448661606072712, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7371 + }, + { + "epoch": 2.945265681182581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7372 + }, + { + "epoch": 2.9456652017578904, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7373 + }, + { + "epoch": 2.9460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7374 + }, + { + "epoch": 2.94646424290851, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7375 + }, + { + "epoch": 2.9468637634838193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7376 + }, + { + "epoch": 2.947263284059129, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7377 + }, + { + "epoch": 2.9476628046344384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7378 + }, + { + "epoch": 2.948062325209748, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7379 + }, + { + "epoch": 2.948461845785058, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7380 + }, + { + "epoch": 2.9488613663603678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7381 + }, + { + "epoch": 2.949260886935677, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7382 + }, + { + "epoch": 2.949660407510987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 7383 + }, + { + "epoch": 2.9500599280862962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7384 + }, + { + "epoch": 2.950459448661606, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7385 + }, + { + "epoch": 2.950858969236916, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7386 + }, + { + "epoch": 2.9512584898122256, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7387 + }, + { + "epoch": 2.951658010387535, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7388 + }, + { + "epoch": 2.9520575309628447, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7389 + }, + { + "epoch": 2.952457051538154, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7390 + }, + { + "epoch": 2.952856572113464, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7391 + }, + { + "epoch": 2.9532560926887736, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7392 + }, + { + "epoch": 2.953655613264083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7393 + }, + { + "epoch": 2.9540551338393928, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7394 + }, + { + "epoch": 2.954454654414702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7395 + }, + { + "epoch": 2.954854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7396 + }, + { + "epoch": 2.9552536955653217, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7397 + }, + { + "epoch": 2.9556532161406315, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7398 + }, + { + "epoch": 2.956052736715941, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7399 + }, + { + "epoch": 2.9564522572912506, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7400 + }, + { + "epoch": 2.95685177786656, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7401 + }, + { + "epoch": 2.9572512984418697, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7402 + }, + { + "epoch": 2.9576508190171795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7403 + }, + { + "epoch": 2.958050339592489, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7404 + }, + { + "epoch": 2.9584498601677987, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7405 + }, + { + "epoch": 2.958849380743108, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7406 + }, + { + "epoch": 2.959248901318418, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 7407 + }, + { + "epoch": 2.9596484218937276, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7408 + }, + { + "epoch": 2.9600479424690374, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7409 + }, + { + "epoch": 2.9604474630443467, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7410 + }, + { + "epoch": 2.9608469836196565, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 7411 + }, + { + "epoch": 2.961246504194966, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7412 + }, + { + "epoch": 2.9616460247702756, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7413 + }, + { + "epoch": 2.9620455453455854, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7414 + }, + { + "epoch": 2.962445065920895, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7415 + }, + { + "epoch": 2.9628445864962045, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7416 + }, + { + "epoch": 2.9632441070715143, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7417 + }, + { + "epoch": 2.9636436276468237, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7418 + }, + { + "epoch": 2.9640431482221334, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7419 + }, + { + "epoch": 2.9644426687974432, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7420 + }, + { + "epoch": 2.9648421893727526, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7421 + }, + { + "epoch": 2.9652417099480624, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7422 + }, + { + "epoch": 2.9656412305233717, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7423 + }, + { + "epoch": 2.9660407510986815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7424 + }, + { + "epoch": 2.9664402716739913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7425 + }, + { + "epoch": 2.966839792249301, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7426 + }, + { + "epoch": 2.9672393128246104, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7427 + }, + { + "epoch": 2.96763883339992, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7428 + }, + { + "epoch": 2.9680383539752295, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7429 + }, + { + "epoch": 2.9684378745505393, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7430 + }, + { + "epoch": 2.968837395125849, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7431 + }, + { + "epoch": 2.969236915701159, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7432 + }, + { + "epoch": 2.9696364362764682, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7433 + }, + { + "epoch": 2.970035956851778, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7434 + }, + { + "epoch": 2.9704354774270874, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7435 + }, + { + "epoch": 2.970834998002397, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7436 + }, + { + "epoch": 2.971234518577707, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 7437 + }, + { + "epoch": 2.9716340391530163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7438 + }, + { + "epoch": 2.972033559728326, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7439 + }, + { + "epoch": 2.9724330803036354, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7440 + }, + { + "epoch": 2.972832600878945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7441 + }, + { + "epoch": 2.973232121454255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7442 + }, + { + "epoch": 2.9736316420295648, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7443 + }, + { + "epoch": 2.974031162604874, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7444 + }, + { + "epoch": 2.974430683180184, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7445 + }, + { + "epoch": 2.9748302037554932, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7446 + }, + { + "epoch": 2.975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7447 + }, + { + "epoch": 2.975629244906113, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7448 + }, + { + "epoch": 2.976028765481422, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7449 + }, + { + "epoch": 2.976428286056732, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7450 + }, + { + "epoch": 2.9768278066320413, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7451 + }, + { + "epoch": 2.977227327207351, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7452 + }, + { + "epoch": 2.977626847782661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7453 + }, + { + "epoch": 2.9780263683579706, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7454 + }, + { + "epoch": 2.97842588893328, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7455 + }, + { + "epoch": 2.9788254095085898, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7456 + }, + { + "epoch": 2.979224930083899, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7457 + }, + { + "epoch": 2.979624450659209, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7458 + }, + { + "epoch": 2.9800239712345187, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7459 + }, + { + "epoch": 2.9804234918098285, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7460 + }, + { + "epoch": 2.980823012385138, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7461 + }, + { + "epoch": 2.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7462 + }, + { + "epoch": 2.981622053535757, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7463 + }, + { + "epoch": 2.9820215741110667, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7464 + }, + { + "epoch": 2.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7465 + }, + { + "epoch": 2.982820615261686, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7466 + }, + { + "epoch": 2.9832201358369956, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7467 + }, + { + "epoch": 2.983619656412305, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7468 + }, + { + "epoch": 2.9840191769876148, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7469 + }, + { + "epoch": 2.9844186975629245, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7470 + }, + { + "epoch": 2.9848182181382343, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7471 + }, + { + "epoch": 2.9852177387135437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7472 + }, + { + "epoch": 2.9856172592888535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7473 + }, + { + "epoch": 2.986016779864163, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7474 + }, + { + "epoch": 2.9864163004394726, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7475 + }, + { + "epoch": 2.9868158210147824, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7476 + }, + { + "epoch": 2.987215341590092, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7477 + }, + { + "epoch": 2.9876148621654015, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7478 + }, + { + "epoch": 2.9880143827407113, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7479 + }, + { + "epoch": 2.9884139033160206, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 7480 + }, + { + "epoch": 2.9888134238913304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7481 + }, + { + "epoch": 2.98921294446664, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7482 + }, + { + "epoch": 2.9896124650419496, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7483 + }, + { + "epoch": 2.9900119856172593, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7484 + }, + { + "epoch": 2.9904115061925687, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7485 + }, + { + "epoch": 2.9908110267678785, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 7486 + }, + { + "epoch": 2.9912105473431883, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7487 + }, + { + "epoch": 2.991610067918498, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7488 + }, + { + "epoch": 2.9920095884938074, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7489 + }, + { + "epoch": 2.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7490 + }, + { + "epoch": 2.9928086296444265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7491 + }, + { + "epoch": 2.9932081502197363, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7492 + }, + { + "epoch": 2.993607670795046, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7493 + }, + { + "epoch": 2.9940071913703554, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7494 + }, + { + "epoch": 2.994406711945665, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7495 + }, + { + "epoch": 2.9948062325209746, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7496 + }, + { + "epoch": 2.9952057530962843, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7497 + }, + { + "epoch": 2.995605273671594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7498 + }, + { + "epoch": 2.996004794246904, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7499 + }, + { + "epoch": 2.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7500 + }, + { + "epoch": 2.996803835397523, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7501 + }, + { + "epoch": 2.9972033559728324, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7502 + }, + { + "epoch": 2.997602876548142, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7503 + }, + { + "epoch": 2.998002397123452, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7504 + }, + { + "epoch": 2.9984019176987617, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7505 + }, + { + "epoch": 2.998801438274071, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7506 + }, + { + "epoch": 2.999200958849381, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7507 + }, + { + "epoch": 2.99960047942469, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7508 + }, + { + "epoch": 3.0, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7509 + }, + { + "epoch": 3.00039952057531, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7510 + }, + { + "epoch": 3.000799041150619, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7511 + }, + { + "epoch": 3.001198561725929, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7512 + }, + { + "epoch": 3.0015980823012387, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7513 + }, + { + "epoch": 3.001997602876548, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7514 + }, + { + "epoch": 3.002397123451858, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7515 + }, + { + "epoch": 3.0027966440271676, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7516 + }, + { + "epoch": 3.003196164602477, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7517 + }, + { + "epoch": 3.0035956851777867, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7518 + }, + { + "epoch": 3.003995205753096, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7519 + }, + { + "epoch": 3.004394726328406, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7520 + }, + { + "epoch": 3.0047942469037157, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7521 + }, + { + "epoch": 3.005193767479025, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7522 + }, + { + "epoch": 3.005593288054335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7523 + }, + { + "epoch": 3.0059928086296446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7524 + }, + { + "epoch": 3.006392329204954, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 7525 + }, + { + "epoch": 3.0067918497802637, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7526 + }, + { + "epoch": 3.0071913703555735, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7527 + }, + { + "epoch": 3.007590890930883, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7528 + }, + { + "epoch": 3.0079904115061926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7529 + }, + { + "epoch": 3.0083899320815024, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7530 + }, + { + "epoch": 3.0087894526568117, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7531 + }, + { + "epoch": 3.0091889732321215, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7532 + }, + { + "epoch": 3.009588493807431, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7533 + }, + { + "epoch": 3.0099880143827407, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7534 + }, + { + "epoch": 3.0103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7535 + }, + { + "epoch": 3.01078705553336, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7536 + }, + { + "epoch": 3.0111865761086696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7537 + }, + { + "epoch": 3.0115860966839794, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7538 + }, + { + "epoch": 3.0119856172592887, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7539 + }, + { + "epoch": 3.0123851378345985, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7540 + }, + { + "epoch": 3.0127846584099083, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7541 + }, + { + "epoch": 3.0131841789852176, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7542 + }, + { + "epoch": 3.0135836995605274, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7543 + }, + { + "epoch": 3.013983220135837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7544 + }, + { + "epoch": 3.0143827407111465, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7545 + }, + { + "epoch": 3.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 7546 + }, + { + "epoch": 3.0151817818617657, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 7547 + }, + { + "epoch": 3.0155813024370755, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7548 + }, + { + "epoch": 3.0159808230123852, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7549 + }, + { + "epoch": 3.0163803435876946, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7550 + }, + { + "epoch": 3.0167798641630044, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7551 + }, + { + "epoch": 3.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7552 + }, + { + "epoch": 3.0175789053136235, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7553 + }, + { + "epoch": 3.0179784258889333, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7554 + }, + { + "epoch": 3.018377946464243, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7555 + }, + { + "epoch": 3.0187774670395524, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7556 + }, + { + "epoch": 3.019176987614862, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7557 + }, + { + "epoch": 3.019576508190172, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7558 + }, + { + "epoch": 3.0199760287654813, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7559 + }, + { + "epoch": 3.020375549340791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7560 + }, + { + "epoch": 3.020775069916101, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7561 + }, + { + "epoch": 3.0211745904914102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7562 + }, + { + "epoch": 3.02157411106672, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7563 + }, + { + "epoch": 3.0219736316420294, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7564 + }, + { + "epoch": 3.022373152217339, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7565 + }, + { + "epoch": 3.022772672792649, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7566 + }, + { + "epoch": 3.0231721933679583, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7567 + }, + { + "epoch": 3.023571713943268, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7568 + }, + { + "epoch": 3.023971234518578, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7569 + }, + { + "epoch": 3.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7570 + }, + { + "epoch": 3.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7571 + }, + { + "epoch": 3.0251697962445068, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7572 + }, + { + "epoch": 3.025569316819816, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7573 + }, + { + "epoch": 3.025968837395126, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7574 + }, + { + "epoch": 3.0263683579704357, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7575 + }, + { + "epoch": 3.026767878545745, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7576 + }, + { + "epoch": 3.027167399121055, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7577 + }, + { + "epoch": 3.027566919696364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7578 + }, + { + "epoch": 3.027966440271674, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7579 + }, + { + "epoch": 3.0283659608469837, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7580 + }, + { + "epoch": 3.028765481422293, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7581 + }, + { + "epoch": 3.029165001997603, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7582 + }, + { + "epoch": 3.0295645225729126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 7583 + }, + { + "epoch": 3.029964043148222, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7584 + }, + { + "epoch": 3.0303635637235318, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7585 + }, + { + "epoch": 3.0307630842988416, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7586 + }, + { + "epoch": 3.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7587 + }, + { + "epoch": 3.0315621254494607, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7588 + }, + { + "epoch": 3.0319616460247705, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7589 + }, + { + "epoch": 3.03236116660008, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7590 + }, + { + "epoch": 3.0327606871753896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 7591 + }, + { + "epoch": 3.033160207750699, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 7592 + }, + { + "epoch": 3.0335597283260087, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7593 + }, + { + "epoch": 3.0339592489013185, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7594 + }, + { + "epoch": 3.034358769476628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7595 + }, + { + "epoch": 3.0347582900519376, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7596 + }, + { + "epoch": 3.0351578106272474, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7597 + }, + { + "epoch": 3.0355573312025568, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 7598 + }, + { + "epoch": 3.0359568517778666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7599 + }, + { + "epoch": 3.0363563723531763, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7600 + }, + { + "epoch": 3.0367558929284857, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 7601 + }, + { + "epoch": 3.0371554135037955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7602 + }, + { + "epoch": 3.0375549340791053, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7603 + }, + { + "epoch": 3.0379544546544146, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7604 + }, + { + "epoch": 3.0383539752297244, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7605 + }, + { + "epoch": 3.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7606 + }, + { + "epoch": 3.0391530163803435, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7607 + }, + { + "epoch": 3.0395525369556533, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7608 + }, + { + "epoch": 3.0399520575309626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7609 + }, + { + "epoch": 3.0403515781062724, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7610 + }, + { + "epoch": 3.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7611 + }, + { + "epoch": 3.0411506192568916, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7612 + }, + { + "epoch": 3.0415501398322013, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7613 + }, + { + "epoch": 3.041949660407511, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7614 + }, + { + "epoch": 3.0423491809828205, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7615 + }, + { + "epoch": 3.0427487015581303, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7616 + }, + { + "epoch": 3.04314822213344, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7617 + }, + { + "epoch": 3.0435477427087494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7618 + }, + { + "epoch": 3.043947263284059, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7619 + }, + { + "epoch": 3.044346783859369, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7620 + }, + { + "epoch": 3.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7621 + }, + { + "epoch": 3.045145825009988, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7622 + }, + { + "epoch": 3.0455453455852974, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7623 + }, + { + "epoch": 3.0459448661606072, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7624 + }, + { + "epoch": 3.046344386735917, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 7625 + }, + { + "epoch": 3.0467439073112264, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7626 + }, + { + "epoch": 3.047143427886536, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7627 + }, + { + "epoch": 3.047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7628 + }, + { + "epoch": 3.0479424690371553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7629 + }, + { + "epoch": 3.048341989612465, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 7630 + }, + { + "epoch": 3.048741510187775, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7631 + }, + { + "epoch": 3.049141030763084, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7632 + }, + { + "epoch": 3.049540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7633 + }, + { + "epoch": 3.0499400719137038, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 7634 + }, + { + "epoch": 3.050339592489013, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7635 + }, + { + "epoch": 3.050739113064323, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7636 + }, + { + "epoch": 3.0511386336396322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7637 + }, + { + "epoch": 3.051538154214942, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7638 + }, + { + "epoch": 3.051937674790252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7639 + }, + { + "epoch": 3.052337195365561, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7640 + }, + { + "epoch": 3.052736715940871, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7641 + }, + { + "epoch": 3.0531362365161807, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7642 + }, + { + "epoch": 3.05353575709149, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7643 + }, + { + "epoch": 3.0539352776668, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7644 + }, + { + "epoch": 3.0543347982421096, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7645 + }, + { + "epoch": 3.054734318817419, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7646 + }, + { + "epoch": 3.0551338393927288, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7647 + }, + { + "epoch": 3.0555333599680385, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7648 + }, + { + "epoch": 3.055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7649 + }, + { + "epoch": 3.0563324011186577, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 7650 + }, + { + "epoch": 3.0567319216939675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7651 + }, + { + "epoch": 3.057131442269277, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7652 + }, + { + "epoch": 3.0575309628445866, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7653 + }, + { + "epoch": 3.057930483419896, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7654 + }, + { + "epoch": 3.0583300039952057, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7655 + }, + { + "epoch": 3.0587295245705155, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7656 + }, + { + "epoch": 3.059129045145825, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7657 + }, + { + "epoch": 3.0595285657211346, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7658 + }, + { + "epoch": 3.0599280862964444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7659 + }, + { + "epoch": 3.0603276068717538, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7660 + }, + { + "epoch": 3.0607271274470635, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7661 + }, + { + "epoch": 3.0611266480223733, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7662 + }, + { + "epoch": 3.0615261685976827, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7663 + }, + { + "epoch": 3.0619256891729925, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7664 + }, + { + "epoch": 3.0623252097483022, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7665 + }, + { + "epoch": 3.0627247303236116, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7666 + }, + { + "epoch": 3.0631242508989214, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7667 + }, + { + "epoch": 3.0635237714742307, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7668 + }, + { + "epoch": 3.0639232920495405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7669 + }, + { + "epoch": 3.0643228126248503, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7670 + }, + { + "epoch": 3.0647223332001596, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 7671 + }, + { + "epoch": 3.0651218537754694, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 7672 + }, + { + "epoch": 3.065521374350779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7673 + }, + { + "epoch": 3.0659208949260885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7674 + }, + { + "epoch": 3.0663204155013983, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7675 + }, + { + "epoch": 3.066719936076708, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7676 + }, + { + "epoch": 3.0671194566520175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7677 + }, + { + "epoch": 3.0675189772273272, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7678 + }, + { + "epoch": 3.067918497802637, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7679 + }, + { + "epoch": 3.0683180183779464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7680 + }, + { + "epoch": 3.068717538953256, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7681 + }, + { + "epoch": 3.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7682 + }, + { + "epoch": 3.0695165801038753, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7683 + }, + { + "epoch": 3.069916100679185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7684 + }, + { + "epoch": 3.0703156212544944, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7685 + }, + { + "epoch": 3.070715141829804, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7686 + }, + { + "epoch": 3.071114662405114, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7687 + }, + { + "epoch": 3.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7688 + }, + { + "epoch": 3.071913703555733, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 7689 + }, + { + "epoch": 3.072313224131043, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7690 + }, + { + "epoch": 3.0727127447063523, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7691 + }, + { + "epoch": 3.073112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7692 + }, + { + "epoch": 3.073511785856972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7693 + }, + { + "epoch": 3.073911306432281, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7694 + }, + { + "epoch": 3.074310827007591, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 7695 + }, + { + "epoch": 3.0747103475829007, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7696 + }, + { + "epoch": 3.07510986815821, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7697 + }, + { + "epoch": 3.07550938873352, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7698 + }, + { + "epoch": 3.075908909308829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7699 + }, + { + "epoch": 3.076308429884139, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7700 + }, + { + "epoch": 3.076707950459449, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7701 + }, + { + "epoch": 3.077107471034758, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7702 + }, + { + "epoch": 3.077506991610068, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7703 + }, + { + "epoch": 3.0779065121853777, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7704 + }, + { + "epoch": 3.078306032760687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7705 + }, + { + "epoch": 3.078705553335997, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7706 + }, + { + "epoch": 3.0791050739113066, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7707 + }, + { + "epoch": 3.079504594486616, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7708 + }, + { + "epoch": 3.0799041150619257, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7709 + }, + { + "epoch": 3.0803036356372355, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7710 + }, + { + "epoch": 3.080703156212545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7711 + }, + { + "epoch": 3.0811026767878547, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7712 + }, + { + "epoch": 3.081502197363164, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7713 + }, + { + "epoch": 3.081901717938474, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7714 + }, + { + "epoch": 3.0823012385137836, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7715 + }, + { + "epoch": 3.082700759089093, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7716 + }, + { + "epoch": 3.0831002796644027, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7717 + }, + { + "epoch": 3.0834998002397125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7718 + }, + { + "epoch": 3.083899320815022, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7719 + }, + { + "epoch": 3.0842988413903316, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7720 + }, + { + "epoch": 3.0846983619656414, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7721 + }, + { + "epoch": 3.0850978825409507, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7722 + }, + { + "epoch": 3.0854974031162605, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7723 + }, + { + "epoch": 3.0858969236915703, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 7724 + }, + { + "epoch": 3.0862964442668797, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7725 + }, + { + "epoch": 3.0866959648421894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7726 + }, + { + "epoch": 3.087095485417499, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7727 + }, + { + "epoch": 3.0874950059928086, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7728 + }, + { + "epoch": 3.0878945265681184, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7729 + }, + { + "epoch": 3.0882940471434277, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7730 + }, + { + "epoch": 3.0886935677187375, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7731 + }, + { + "epoch": 3.0890930882940473, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7732 + }, + { + "epoch": 3.0894926088693566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7733 + }, + { + "epoch": 3.0898921294446664, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7734 + }, + { + "epoch": 3.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7735 + }, + { + "epoch": 3.0906911705952855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7736 + }, + { + "epoch": 3.0910906911705953, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7737 + }, + { + "epoch": 3.091490211745905, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7738 + }, + { + "epoch": 3.0918897323212144, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7739 + }, + { + "epoch": 3.0922892528965242, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7740 + }, + { + "epoch": 3.092688773471834, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7741 + }, + { + "epoch": 3.0930882940471434, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7742 + }, + { + "epoch": 3.093487814622453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7743 + }, + { + "epoch": 3.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7744 + }, + { + "epoch": 3.0942868557730723, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7745 + }, + { + "epoch": 3.094686376348382, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7746 + }, + { + "epoch": 3.0950858969236914, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7747 + }, + { + "epoch": 3.095485417499001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 7748 + }, + { + "epoch": 3.095884938074311, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7749 + }, + { + "epoch": 3.0962844586496203, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7750 + }, + { + "epoch": 3.09668397922493, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7751 + }, + { + "epoch": 3.09708349980024, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7752 + }, + { + "epoch": 3.0974830203755492, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7753 + }, + { + "epoch": 3.097882540950859, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 7754 + }, + { + "epoch": 3.098282061526169, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7755 + }, + { + "epoch": 3.098681582101478, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7756 + }, + { + "epoch": 3.099081102676788, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7757 + }, + { + "epoch": 3.0994806232520973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7758 + }, + { + "epoch": 3.099880143827407, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7759 + }, + { + "epoch": 3.100279664402717, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7760 + }, + { + "epoch": 3.100679184978026, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7761 + }, + { + "epoch": 3.101078705553336, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7762 + }, + { + "epoch": 3.1014782261286458, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7763 + }, + { + "epoch": 3.101877746703955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7764 + }, + { + "epoch": 3.102277267279265, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7765 + }, + { + "epoch": 3.1026767878545747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 7766 + }, + { + "epoch": 3.103076308429884, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7767 + }, + { + "epoch": 3.103475829005194, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7768 + }, + { + "epoch": 3.1038753495805036, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7769 + }, + { + "epoch": 3.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7770 + }, + { + "epoch": 3.1046743907311227, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7771 + }, + { + "epoch": 3.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7772 + }, + { + "epoch": 3.105473431881742, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7773 + }, + { + "epoch": 3.1058729524570516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7774 + }, + { + "epoch": 3.106272473032361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7775 + }, + { + "epoch": 3.1066719936076708, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7776 + }, + { + "epoch": 3.1070715141829806, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7777 + }, + { + "epoch": 3.10747103475829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7778 + }, + { + "epoch": 3.1078705553335997, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7779 + }, + { + "epoch": 3.1082700759089095, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7780 + }, + { + "epoch": 3.108669596484219, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7781 + }, + { + "epoch": 3.1090691170595286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 7782 + }, + { + "epoch": 3.1094686376348384, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7783 + }, + { + "epoch": 3.1098681582101477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7784 + }, + { + "epoch": 3.1102676787854575, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 7785 + }, + { + "epoch": 3.1106671993607673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7786 + }, + { + "epoch": 3.1110667199360766, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7787 + }, + { + "epoch": 3.1114662405113864, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7788 + }, + { + "epoch": 3.1118657610866958, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7789 + }, + { + "epoch": 3.1122652816620056, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7790 + }, + { + "epoch": 3.1126648022373153, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7791 + }, + { + "epoch": 3.1130643228126247, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7792 + }, + { + "epoch": 3.1134638433879345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 7793 + }, + { + "epoch": 3.1138633639632443, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7794 + }, + { + "epoch": 3.1142628845385536, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7795 + }, + { + "epoch": 3.1146624051138634, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7796 + }, + { + "epoch": 3.115061925689173, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7797 + }, + { + "epoch": 3.1154614462644825, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7798 + }, + { + "epoch": 3.1158609668397923, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7799 + }, + { + "epoch": 3.116260487415102, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7800 + }, + { + "epoch": 3.1166600079904114, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7801 + }, + { + "epoch": 3.117059528565721, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 7802 + }, + { + "epoch": 3.1174590491410306, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7803 + }, + { + "epoch": 3.1178585697163403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7804 + }, + { + "epoch": 3.11825809029165, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7805 + }, + { + "epoch": 3.1186576108669595, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 7806 + }, + { + "epoch": 3.1190571314422693, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7807 + }, + { + "epoch": 3.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7808 + }, + { + "epoch": 3.1198561725928884, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7809 + }, + { + "epoch": 3.120255693168198, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7810 + }, + { + "epoch": 3.120655213743508, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7811 + }, + { + "epoch": 3.1210547343188173, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7812 + }, + { + "epoch": 3.121454254894127, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7813 + }, + { + "epoch": 3.121853775469437, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7814 + }, + { + "epoch": 3.122253296044746, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7815 + }, + { + "epoch": 3.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7816 + }, + { + "epoch": 3.1230523371953653, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7817 + }, + { + "epoch": 3.123451857770675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7818 + }, + { + "epoch": 3.123851378345985, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7819 + }, + { + "epoch": 3.1242508989212943, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7820 + }, + { + "epoch": 3.124650419496604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7821 + }, + { + "epoch": 3.125049940071914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7822 + }, + { + "epoch": 3.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7823 + }, + { + "epoch": 3.125848981222533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7824 + }, + { + "epoch": 3.1262485017978427, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7825 + }, + { + "epoch": 3.126648022373152, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7826 + }, + { + "epoch": 3.127047542948462, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7827 + }, + { + "epoch": 3.1274470635237717, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7828 + }, + { + "epoch": 3.127846584099081, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7829 + }, + { + "epoch": 3.128246104674391, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7830 + }, + { + "epoch": 3.1286456252497006, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7831 + }, + { + "epoch": 3.12904514582501, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7832 + }, + { + "epoch": 3.1294446664003197, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7833 + }, + { + "epoch": 3.129844186975629, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7834 + }, + { + "epoch": 3.130243707550939, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7835 + }, + { + "epoch": 3.1306432281262486, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7836 + }, + { + "epoch": 3.131042748701558, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7837 + }, + { + "epoch": 3.1314422692768678, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7838 + }, + { + "epoch": 3.1318417898521775, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7839 + }, + { + "epoch": 3.132241310427487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7840 + }, + { + "epoch": 3.1326408310027967, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7841 + }, + { + "epoch": 3.1330403515781065, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7842 + }, + { + "epoch": 3.133439872153416, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7843 + }, + { + "epoch": 3.1338393927287256, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7844 + }, + { + "epoch": 3.134238913304035, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7845 + }, + { + "epoch": 3.1346384338793447, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7846 + }, + { + "epoch": 3.1350379544546545, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7847 + }, + { + "epoch": 3.135437475029964, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7848 + }, + { + "epoch": 3.1358369956052736, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7849 + }, + { + "epoch": 3.1362365161805834, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7850 + }, + { + "epoch": 3.1366360367558928, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7851 + }, + { + "epoch": 3.1370355573312025, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7852 + }, + { + "epoch": 3.1374350779065123, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7853 + }, + { + "epoch": 3.1378345984818217, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7854 + }, + { + "epoch": 3.1382341190571315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7855 + }, + { + "epoch": 3.1386336396324412, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7856 + }, + { + "epoch": 3.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7857 + }, + { + "epoch": 3.1394326807830604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7858 + }, + { + "epoch": 3.13983220135837, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7859 + }, + { + "epoch": 3.1402317219336795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7860 + }, + { + "epoch": 3.1406312425089893, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7861 + }, + { + "epoch": 3.1410307630842986, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7862 + }, + { + "epoch": 3.1414302836596084, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7863 + }, + { + "epoch": 3.141829804234918, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7864 + }, + { + "epoch": 3.1422293248102275, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7865 + }, + { + "epoch": 3.1426288453855373, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7866 + }, + { + "epoch": 3.143028365960847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7867 + }, + { + "epoch": 3.1434278865361565, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7868 + }, + { + "epoch": 3.1438274071114662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7869 + }, + { + "epoch": 3.144226927686776, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7870 + }, + { + "epoch": 3.1446264482620854, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7871 + }, + { + "epoch": 3.145025968837395, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7872 + }, + { + "epoch": 3.145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7873 + }, + { + "epoch": 3.1458250099880143, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 7874 + }, + { + "epoch": 3.146224530563324, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7875 + }, + { + "epoch": 3.146624051138634, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7876 + }, + { + "epoch": 3.147023571713943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7877 + }, + { + "epoch": 3.147423092289253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7878 + }, + { + "epoch": 3.1478226128645623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7879 + }, + { + "epoch": 3.148222133439872, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7880 + }, + { + "epoch": 3.148621654015182, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7881 + }, + { + "epoch": 3.1490211745904912, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7882 + }, + { + "epoch": 3.149420695165801, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7883 + }, + { + "epoch": 3.149820215741111, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7884 + }, + { + "epoch": 3.15021973631642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7885 + }, + { + "epoch": 3.15061925689173, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7886 + }, + { + "epoch": 3.1510187774670397, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 7887 + }, + { + "epoch": 3.151418298042349, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7888 + }, + { + "epoch": 3.151817818617659, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7889 + }, + { + "epoch": 3.152217339192968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7890 + }, + { + "epoch": 3.152616859768278, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7891 + }, + { + "epoch": 3.1530163803435878, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7892 + }, + { + "epoch": 3.153415900918897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7893 + }, + { + "epoch": 3.153815421494207, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7894 + }, + { + "epoch": 3.1542149420695167, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7895 + }, + { + "epoch": 3.154614462644826, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7896 + }, + { + "epoch": 3.155013983220136, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7897 + }, + { + "epoch": 3.1554135037954456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7898 + }, + { + "epoch": 3.155813024370755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7899 + }, + { + "epoch": 3.1562125449460647, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7900 + }, + { + "epoch": 3.1566120655213745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7901 + }, + { + "epoch": 3.157011586096684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7902 + }, + { + "epoch": 3.1574111066719937, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7903 + }, + { + "epoch": 3.1578106272473034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7904 + }, + { + "epoch": 3.158210147822613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7905 + }, + { + "epoch": 3.1586096683979226, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7906 + }, + { + "epoch": 3.159009188973232, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7907 + }, + { + "epoch": 3.1594087095485417, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7908 + }, + { + "epoch": 3.1598082301238515, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 7909 + }, + { + "epoch": 3.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7910 + }, + { + "epoch": 3.1606072712744706, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7911 + }, + { + "epoch": 3.1610067918497804, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7912 + }, + { + "epoch": 3.1614063124250897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7913 + }, + { + "epoch": 3.1618058330003995, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 7914 + }, + { + "epoch": 3.1622053535757093, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7915 + }, + { + "epoch": 3.1626048741510187, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7916 + }, + { + "epoch": 3.1630043947263284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 7917 + }, + { + "epoch": 3.1634039153016382, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7918 + }, + { + "epoch": 3.1638034358769476, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7919 + }, + { + "epoch": 3.1642029564522574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7920 + }, + { + "epoch": 3.164602477027567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7921 + }, + { + "epoch": 3.1650019976028765, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7922 + }, + { + "epoch": 3.1654015181781863, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7923 + }, + { + "epoch": 3.1658010387534956, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7924 + }, + { + "epoch": 3.1662005593288054, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 7925 + }, + { + "epoch": 3.166600079904115, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7926 + }, + { + "epoch": 3.1669996004794245, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7927 + }, + { + "epoch": 3.1673991210547343, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7928 + }, + { + "epoch": 3.167798641630044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7929 + }, + { + "epoch": 3.1681981622053534, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 7930 + }, + { + "epoch": 3.1685976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7931 + }, + { + "epoch": 3.168997203355973, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7932 + }, + { + "epoch": 3.1693967239312824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7933 + }, + { + "epoch": 3.169796244506592, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7934 + }, + { + "epoch": 3.1701957650819015, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7935 + }, + { + "epoch": 3.1705952856572113, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7936 + }, + { + "epoch": 3.170994806232521, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7937 + }, + { + "epoch": 3.1713943268078304, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7938 + }, + { + "epoch": 3.17179384738314, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7939 + }, + { + "epoch": 3.17219336795845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 7940 + }, + { + "epoch": 3.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7941 + }, + { + "epoch": 3.172992409109069, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7942 + }, + { + "epoch": 3.173391929684379, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7943 + }, + { + "epoch": 3.1737914502596882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7944 + }, + { + "epoch": 3.174190970834998, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7945 + }, + { + "epoch": 3.174590491410308, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7946 + }, + { + "epoch": 3.174990011985617, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 7947 + }, + { + "epoch": 3.175389532560927, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7948 + }, + { + "epoch": 3.1757890531362367, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7949 + }, + { + "epoch": 3.176188573711546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 7950 + }, + { + "epoch": 3.176588094286856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7951 + }, + { + "epoch": 3.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7952 + }, + { + "epoch": 3.177387135437475, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7953 + }, + { + "epoch": 3.1777866560127848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7954 + }, + { + "epoch": 3.178186176588094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 7955 + }, + { + "epoch": 3.178585697163404, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7956 + }, + { + "epoch": 3.1789852177387137, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7957 + }, + { + "epoch": 3.179384738314023, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7958 + }, + { + "epoch": 3.179784258889333, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7959 + }, + { + "epoch": 3.1801837794646426, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 7960 + }, + { + "epoch": 3.180583300039952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 7961 + }, + { + "epoch": 3.1809828206152617, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7962 + }, + { + "epoch": 3.1813823411905715, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7963 + }, + { + "epoch": 3.181781861765881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7964 + }, + { + "epoch": 3.1821813823411906, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7965 + }, + { + "epoch": 3.1825809029165004, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7966 + }, + { + "epoch": 3.1829804234918098, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7967 + }, + { + "epoch": 3.1833799440671195, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7968 + }, + { + "epoch": 3.183779464642429, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7969 + }, + { + "epoch": 3.1841789852177387, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7970 + }, + { + "epoch": 3.1845785057930485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7971 + }, + { + "epoch": 3.184978026368358, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7972 + }, + { + "epoch": 3.1853775469436676, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7973 + }, + { + "epoch": 3.1857770675189774, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 7974 + }, + { + "epoch": 3.1861765880942867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7975 + }, + { + "epoch": 3.1865761086695965, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7976 + }, + { + "epoch": 3.1869756292449063, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7977 + }, + { + "epoch": 3.1873751498202156, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7978 + }, + { + "epoch": 3.1877746703955254, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7979 + }, + { + "epoch": 3.1881741909708348, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7980 + }, + { + "epoch": 3.1885737115461446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7981 + }, + { + "epoch": 3.1889732321214543, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7982 + }, + { + "epoch": 3.1893727526967637, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7983 + }, + { + "epoch": 3.1897722732720735, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7984 + }, + { + "epoch": 3.1901717938473833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7985 + }, + { + "epoch": 3.1905713144226926, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7986 + }, + { + "epoch": 3.1909708349980024, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7987 + }, + { + "epoch": 3.191370355573312, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7988 + }, + { + "epoch": 3.1917698761486215, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7989 + }, + { + "epoch": 3.1921693967239313, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7990 + }, + { + "epoch": 3.192568917299241, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7991 + }, + { + "epoch": 3.1929684378745504, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 7992 + }, + { + "epoch": 3.19336795844986, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7993 + }, + { + "epoch": 3.19376747902517, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7994 + }, + { + "epoch": 3.1941669996004793, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 7995 + }, + { + "epoch": 3.194566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7996 + }, + { + "epoch": 3.1949660407510985, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7997 + }, + { + "epoch": 3.1953655613264083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7998 + }, + { + "epoch": 3.195765081901718, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7999 + }, + { + "epoch": 3.1961646024770274, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8000 + }, + { + "epoch": 3.196564123052337, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8001 + }, + { + "epoch": 3.196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8002 + }, + { + "epoch": 3.1973631642029563, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8003 + }, + { + "epoch": 3.197762684778266, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8004 + }, + { + "epoch": 3.198162205353576, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8005 + }, + { + "epoch": 3.198561725928885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 8006 + }, + { + "epoch": 3.198961246504195, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8007 + }, + { + "epoch": 3.199360767079505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8008 + }, + { + "epoch": 3.199760287654814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8009 + }, + { + "epoch": 3.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 8010 + }, + { + "epoch": 3.2005593288054337, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8011 + }, + { + "epoch": 3.200958849380743, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8012 + }, + { + "epoch": 3.201358369956053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8013 + }, + { + "epoch": 3.201757890531362, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8014 + }, + { + "epoch": 3.202157411106672, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8015 + }, + { + "epoch": 3.2025569316819817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8016 + }, + { + "epoch": 3.202956452257291, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8017 + }, + { + "epoch": 3.203355972832601, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8018 + }, + { + "epoch": 3.2037554934079107, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8019 + }, + { + "epoch": 3.20415501398322, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 8020 + }, + { + "epoch": 3.20455453455853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8021 + }, + { + "epoch": 3.2049540551338396, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 8022 + }, + { + "epoch": 3.205353575709149, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8023 + }, + { + "epoch": 3.2057530962844587, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8024 + }, + { + "epoch": 3.206152616859768, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8025 + }, + { + "epoch": 3.206552137435078, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8026 + }, + { + "epoch": 3.2069516580103876, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8027 + }, + { + "epoch": 3.207351178585697, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8028 + }, + { + "epoch": 3.2077506991610067, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8029 + }, + { + "epoch": 3.2081502197363165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 8030 + }, + { + "epoch": 3.208549740311626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8031 + }, + { + "epoch": 3.2089492608869357, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8032 + }, + { + "epoch": 3.2093487814622454, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8033 + }, + { + "epoch": 3.209748302037555, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8034 + }, + { + "epoch": 3.2101478226128646, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8035 + }, + { + "epoch": 3.2105473431881744, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8036 + }, + { + "epoch": 3.2109468637634837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 8037 + }, + { + "epoch": 3.2113463843387935, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8038 + }, + { + "epoch": 3.2117459049141033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8039 + }, + { + "epoch": 3.2121454254894126, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8040 + }, + { + "epoch": 3.2125449460647224, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8041 + }, + { + "epoch": 3.2129444666400317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8042 + }, + { + "epoch": 3.2133439872153415, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8043 + }, + { + "epoch": 3.2137435077906513, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8044 + }, + { + "epoch": 3.2141430283659607, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8045 + }, + { + "epoch": 3.2145425489412705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8046 + }, + { + "epoch": 3.2149420695165802, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8047 + }, + { + "epoch": 3.2153415900918896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8048 + }, + { + "epoch": 3.2157411106671994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8049 + }, + { + "epoch": 3.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8050 + }, + { + "epoch": 3.2165401518178185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8051 + }, + { + "epoch": 3.2169396723931283, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8052 + }, + { + "epoch": 3.217339192968438, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8053 + }, + { + "epoch": 3.2177387135437474, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8054 + }, + { + "epoch": 3.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8055 + }, + { + "epoch": 3.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8056 + }, + { + "epoch": 3.2189372752696763, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8057 + }, + { + "epoch": 3.219336795844986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8058 + }, + { + "epoch": 3.2197363164202955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8059 + }, + { + "epoch": 3.2201358369956052, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 8060 + }, + { + "epoch": 3.220535357570915, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8061 + }, + { + "epoch": 3.2209348781462244, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8062 + }, + { + "epoch": 3.221334398721534, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8063 + }, + { + "epoch": 3.221733919296844, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8064 + }, + { + "epoch": 3.2221334398721533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8065 + }, + { + "epoch": 3.222532960447463, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 8066 + }, + { + "epoch": 3.222932481022773, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8067 + }, + { + "epoch": 3.223332001598082, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 8068 + }, + { + "epoch": 3.223731522173392, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8069 + }, + { + "epoch": 3.2241310427487013, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8070 + }, + { + "epoch": 3.224530563324011, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8071 + }, + { + "epoch": 3.224930083899321, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8072 + }, + { + "epoch": 3.2253296044746302, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8073 + }, + { + "epoch": 3.22572912504994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8074 + }, + { + "epoch": 3.22612864562525, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8075 + }, + { + "epoch": 3.226528166200559, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8076 + }, + { + "epoch": 3.226927686775869, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8077 + }, + { + "epoch": 3.2273272073511787, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8078 + }, + { + "epoch": 3.227726727926488, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8079 + }, + { + "epoch": 3.228126248501798, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8080 + }, + { + "epoch": 3.2285257690771076, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8081 + }, + { + "epoch": 3.228925289652417, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8082 + }, + { + "epoch": 3.2293248102277268, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8083 + }, + { + "epoch": 3.2297243308030366, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8084 + }, + { + "epoch": 3.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8085 + }, + { + "epoch": 3.2305233719536557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8086 + }, + { + "epoch": 3.230922892528965, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8087 + }, + { + "epoch": 3.231322413104275, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8088 + }, + { + "epoch": 3.2317219336795846, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8089 + }, + { + "epoch": 3.232121454254894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8090 + }, + { + "epoch": 3.2325209748302037, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8091 + }, + { + "epoch": 3.2329204954055135, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8092 + }, + { + "epoch": 3.233320015980823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8093 + }, + { + "epoch": 3.2337195365561326, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8094 + }, + { + "epoch": 3.2341190571314424, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8095 + }, + { + "epoch": 3.2345185777067518, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8096 + }, + { + "epoch": 3.2349180982820616, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 8097 + }, + { + "epoch": 3.2353176188573713, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8098 + }, + { + "epoch": 3.2357171394326807, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8099 + }, + { + "epoch": 3.2361166600079905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 8100 + }, + { + "epoch": 3.2365161805833003, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8101 + }, + { + "epoch": 3.2369157011586096, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 8102 + }, + { + "epoch": 3.2373152217339194, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8103 + }, + { + "epoch": 3.2377147423092287, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8104 + }, + { + "epoch": 3.2381142628845385, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8105 + }, + { + "epoch": 3.2385137834598483, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8106 + }, + { + "epoch": 3.2389133040351576, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 8107 + }, + { + "epoch": 3.2393128246104674, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8108 + }, + { + "epoch": 3.239712345185777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8109 + }, + { + "epoch": 3.2401118657610866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8110 + }, + { + "epoch": 3.2405113863363963, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 8111 + }, + { + "epoch": 3.240910906911706, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8112 + }, + { + "epoch": 3.2413104274870155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8113 + }, + { + "epoch": 3.2417099480623253, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8114 + }, + { + "epoch": 3.2421094686376346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8115 + }, + { + "epoch": 3.2425089892129444, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8116 + }, + { + "epoch": 3.242908509788254, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8117 + }, + { + "epoch": 3.2433080303635635, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8118 + }, + { + "epoch": 3.2437075509388733, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 8119 + }, + { + "epoch": 3.244107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8120 + }, + { + "epoch": 3.2445065920894924, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8121 + }, + { + "epoch": 3.2449061126648022, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8122 + }, + { + "epoch": 3.245305633240112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8123 + }, + { + "epoch": 3.2457051538154214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8124 + }, + { + "epoch": 3.246104674390731, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8125 + }, + { + "epoch": 3.246504194966041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8126 + }, + { + "epoch": 3.2469037155413503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8127 + }, + { + "epoch": 3.24730323611666, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8128 + }, + { + "epoch": 3.24770275669197, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8129 + }, + { + "epoch": 3.248102277267279, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8130 + }, + { + "epoch": 3.248501797842589, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8131 + }, + { + "epoch": 3.2489013184178983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8132 + }, + { + "epoch": 3.249300838993208, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8133 + }, + { + "epoch": 3.249700359568518, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8134 + }, + { + "epoch": 3.2500998801438272, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8135 + }, + { + "epoch": 3.250499400719137, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8136 + }, + { + "epoch": 3.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8137 + }, + { + "epoch": 3.251298441869756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8138 + }, + { + "epoch": 3.251697962445066, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 8139 + }, + { + "epoch": 3.2520974830203757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8140 + }, + { + "epoch": 3.252497003595685, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8141 + }, + { + "epoch": 3.252896524170995, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 8142 + }, + { + "epoch": 3.2532960447463046, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8143 + }, + { + "epoch": 3.253695565321614, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8144 + }, + { + "epoch": 3.2540950858969238, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8145 + }, + { + "epoch": 3.2544946064722335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8146 + }, + { + "epoch": 3.254894127047543, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8147 + }, + { + "epoch": 3.2552936476228527, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8148 + }, + { + "epoch": 3.255693168198162, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8149 + }, + { + "epoch": 3.256092688773472, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8150 + }, + { + "epoch": 3.2564922093487816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8151 + }, + { + "epoch": 3.256891729924091, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8152 + }, + { + "epoch": 3.2572912504994007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8153 + }, + { + "epoch": 3.2576907710747105, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8154 + }, + { + "epoch": 3.25809029165002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8155 + }, + { + "epoch": 3.2584898122253296, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8156 + }, + { + "epoch": 3.2588893328006394, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8157 + }, + { + "epoch": 3.2592888533759488, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8158 + }, + { + "epoch": 3.2596883739512585, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8159 + }, + { + "epoch": 3.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8160 + }, + { + "epoch": 3.2604874151018777, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8161 + }, + { + "epoch": 3.2608869356771875, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8162 + }, + { + "epoch": 3.2612864562524972, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8163 + }, + { + "epoch": 3.2616859768278066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8164 + }, + { + "epoch": 3.2620854974031164, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 8165 + }, + { + "epoch": 3.2624850179784257, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8166 + }, + { + "epoch": 3.2628845385537355, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8167 + }, + { + "epoch": 3.2632840591290453, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8168 + }, + { + "epoch": 3.2636835797043546, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8169 + }, + { + "epoch": 3.2640831002796644, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8170 + }, + { + "epoch": 3.264482620854974, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8171 + }, + { + "epoch": 3.2648821414302835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8172 + }, + { + "epoch": 3.2652816620055933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8173 + }, + { + "epoch": 3.265681182580903, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8174 + }, + { + "epoch": 3.2660807031562125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8175 + }, + { + "epoch": 3.2664802237315222, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8176 + }, + { + "epoch": 3.2668797443068316, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8177 + }, + { + "epoch": 3.2672792648821414, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8178 + }, + { + "epoch": 3.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8179 + }, + { + "epoch": 3.2680783060327605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8180 + }, + { + "epoch": 3.2684778266080703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8181 + }, + { + "epoch": 3.26887734718338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8182 + }, + { + "epoch": 3.2692768677586894, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8183 + }, + { + "epoch": 3.269676388333999, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8184 + }, + { + "epoch": 3.270075908909309, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8185 + }, + { + "epoch": 3.2704754294846183, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8186 + }, + { + "epoch": 3.270874950059928, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 8187 + }, + { + "epoch": 3.271274470635238, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8188 + }, + { + "epoch": 3.2716739912105472, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8189 + }, + { + "epoch": 3.272073511785857, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8190 + }, + { + "epoch": 3.272473032361167, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8191 + }, + { + "epoch": 3.272872552936476, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 8192 + }, + { + "epoch": 3.273272073511786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8193 + }, + { + "epoch": 3.2736715940870953, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8194 + }, + { + "epoch": 3.274071114662405, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8195 + }, + { + "epoch": 3.274470635237715, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8196 + }, + { + "epoch": 3.274870155813024, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 8197 + }, + { + "epoch": 3.275269676388334, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8198 + }, + { + "epoch": 3.275669196963644, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8199 + }, + { + "epoch": 3.276068717538953, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8200 + }, + { + "epoch": 3.276468238114263, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8201 + }, + { + "epoch": 3.2768677586895727, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8202 + }, + { + "epoch": 3.277267279264882, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8203 + }, + { + "epoch": 3.277666799840192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8204 + }, + { + "epoch": 3.278066320415501, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8205 + }, + { + "epoch": 3.278465840990811, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8206 + }, + { + "epoch": 3.2788653615661207, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8207 + }, + { + "epoch": 3.2792648821414305, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8208 + }, + { + "epoch": 3.27966440271674, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8209 + }, + { + "epoch": 3.2800639232920497, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8210 + }, + { + "epoch": 3.280463443867359, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 8211 + }, + { + "epoch": 3.280862964442669, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8212 + }, + { + "epoch": 3.2812624850179786, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8213 + }, + { + "epoch": 3.281662005593288, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 8214 + }, + { + "epoch": 3.2820615261685977, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8215 + }, + { + "epoch": 3.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8216 + }, + { + "epoch": 3.282860567319217, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8217 + }, + { + "epoch": 3.2832600878945266, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8218 + }, + { + "epoch": 3.2836596084698364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8219 + }, + { + "epoch": 3.2840591290451457, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8220 + }, + { + "epoch": 3.2844586496204555, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8221 + }, + { + "epoch": 3.284858170195765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8222 + }, + { + "epoch": 3.2852576907710747, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8223 + }, + { + "epoch": 3.2856572113463844, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8224 + }, + { + "epoch": 3.286056731921694, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8225 + }, + { + "epoch": 3.2864562524970036, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8226 + }, + { + "epoch": 3.2868557730723134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8227 + }, + { + "epoch": 3.2872552936476227, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 8228 + }, + { + "epoch": 3.2876548142229325, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8229 + }, + { + "epoch": 3.2880543347982423, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8230 + }, + { + "epoch": 3.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8231 + }, + { + "epoch": 3.2888533759488614, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8232 + }, + { + "epoch": 3.289252896524171, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8233 + }, + { + "epoch": 3.2896524170994805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8234 + }, + { + "epoch": 3.2900519376747903, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8235 + }, + { + "epoch": 3.2904514582501, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8236 + }, + { + "epoch": 3.2908509788254094, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8237 + }, + { + "epoch": 3.2912504994007192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8238 + }, + { + "epoch": 3.2916500199760286, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8239 + }, + { + "epoch": 3.2920495405513384, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8240 + }, + { + "epoch": 3.292449061126648, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8241 + }, + { + "epoch": 3.2928485817019575, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8242 + }, + { + "epoch": 3.2932481022772673, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8243 + }, + { + "epoch": 3.293647622852577, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8244 + }, + { + "epoch": 3.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8245 + }, + { + "epoch": 3.294446664003196, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8246 + }, + { + "epoch": 3.294846184578506, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8247 + }, + { + "epoch": 3.2952457051538153, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8248 + }, + { + "epoch": 3.295645225729125, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8249 + }, + { + "epoch": 3.2960447463044344, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8250 + }, + { + "epoch": 3.2964442668797442, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8251 + }, + { + "epoch": 3.296843787455054, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8252 + }, + { + "epoch": 3.297243308030364, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8253 + }, + { + "epoch": 3.297642828605673, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8254 + }, + { + "epoch": 3.298042349180983, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8255 + }, + { + "epoch": 3.2984418697562923, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8256 + }, + { + "epoch": 3.298841390331602, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8257 + }, + { + "epoch": 3.299240910906912, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8258 + }, + { + "epoch": 3.299640431482221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8259 + }, + { + "epoch": 3.300039952057531, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8260 + }, + { + "epoch": 3.3004394726328408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8261 + }, + { + "epoch": 3.30083899320815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8262 + }, + { + "epoch": 3.30123851378346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8263 + }, + { + "epoch": 3.3016380343587697, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8264 + }, + { + "epoch": 3.302037554934079, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8265 + }, + { + "epoch": 3.302437075509389, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8266 + }, + { + "epoch": 3.302836596084698, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8267 + }, + { + "epoch": 3.303236116660008, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8268 + }, + { + "epoch": 3.3036356372353177, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8269 + }, + { + "epoch": 3.304035157810627, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8270 + }, + { + "epoch": 3.304434678385937, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8271 + }, + { + "epoch": 3.3048341989612466, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8272 + }, + { + "epoch": 3.305233719536556, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8273 + }, + { + "epoch": 3.3056332401118658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8274 + }, + { + "epoch": 3.3060327606871756, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8275 + }, + { + "epoch": 3.306432281262485, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8276 + }, + { + "epoch": 3.3068318018377947, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8277 + }, + { + "epoch": 3.3072313224131045, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8278 + }, + { + "epoch": 3.307630842988414, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 8279 + }, + { + "epoch": 3.3080303635637236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8280 + }, + { + "epoch": 3.3084298841390334, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8281 + }, + { + "epoch": 3.3088294047143427, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8282 + }, + { + "epoch": 3.3092289252896525, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8283 + }, + { + "epoch": 3.309628445864962, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8284 + }, + { + "epoch": 3.3100279664402716, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8285 + }, + { + "epoch": 3.3104274870155814, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8286 + }, + { + "epoch": 3.3108270075908908, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8287 + }, + { + "epoch": 3.3112265281662006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8288 + }, + { + "epoch": 3.3116260487415103, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8289 + }, + { + "epoch": 3.3120255693168197, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8290 + }, + { + "epoch": 3.3124250898921295, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8291 + }, + { + "epoch": 3.3128246104674393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8292 + }, + { + "epoch": 3.3132241310427486, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8293 + }, + { + "epoch": 3.3136236516180584, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8294 + }, + { + "epoch": 3.3140231721933677, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8295 + }, + { + "epoch": 3.3144226927686775, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8296 + }, + { + "epoch": 3.3148222133439873, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8297 + }, + { + "epoch": 3.315221733919297, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8298 + }, + { + "epoch": 3.3156212544946064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8299 + }, + { + "epoch": 3.316020775069916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 8300 + }, + { + "epoch": 3.3164202956452256, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8301 + }, + { + "epoch": 3.3168198162205353, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8302 + }, + { + "epoch": 3.317219336795845, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8303 + }, + { + "epoch": 3.3176188573711545, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8304 + }, + { + "epoch": 3.3180183779464643, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8305 + }, + { + "epoch": 3.318417898521774, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8306 + }, + { + "epoch": 3.3188174190970834, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8307 + }, + { + "epoch": 3.319216939672393, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8308 + }, + { + "epoch": 3.319616460247703, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8309 + }, + { + "epoch": 3.3200159808230123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8310 + }, + { + "epoch": 3.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8311 + }, + { + "epoch": 3.3208150219736314, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8312 + }, + { + "epoch": 3.321214542548941, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8313 + }, + { + "epoch": 3.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8314 + }, + { + "epoch": 3.3220135836995603, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8315 + }, + { + "epoch": 3.32241310427487, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8316 + }, + { + "epoch": 3.32281262485018, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8317 + }, + { + "epoch": 3.3232121454254893, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8318 + }, + { + "epoch": 3.323611666000799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8319 + }, + { + "epoch": 3.324011186576109, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8320 + }, + { + "epoch": 3.324410707151418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8321 + }, + { + "epoch": 3.324810227726728, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8322 + }, + { + "epoch": 3.3252097483020377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8323 + }, + { + "epoch": 3.325609268877347, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8324 + }, + { + "epoch": 3.326008789452657, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8325 + }, + { + "epoch": 3.3264083100279667, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 8326 + }, + { + "epoch": 3.326807830603276, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8327 + }, + { + "epoch": 3.327207351178586, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8328 + }, + { + "epoch": 3.327606871753895, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8329 + }, + { + "epoch": 3.328006392329205, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8330 + }, + { + "epoch": 3.3284059129045147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8331 + }, + { + "epoch": 3.328805433479824, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8332 + }, + { + "epoch": 3.329204954055134, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8333 + }, + { + "epoch": 3.3296044746304436, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8334 + }, + { + "epoch": 3.330003995205753, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8335 + }, + { + "epoch": 3.3304035157810628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8336 + }, + { + "epoch": 3.3308030363563725, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8337 + }, + { + "epoch": 3.331202556931682, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8338 + }, + { + "epoch": 3.3316020775069917, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8339 + }, + { + "epoch": 3.332001598082301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8340 + }, + { + "epoch": 3.332401118657611, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 8341 + }, + { + "epoch": 3.3328006392329206, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8342 + }, + { + "epoch": 3.3332001598082304, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8343 + }, + { + "epoch": 3.3335996803835397, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8344 + }, + { + "epoch": 3.3339992009588495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8345 + }, + { + "epoch": 3.334398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8346 + }, + { + "epoch": 3.3347982421094686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8347 + }, + { + "epoch": 3.3351977626847784, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8348 + }, + { + "epoch": 3.3355972832600878, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8349 + }, + { + "epoch": 3.3359968038353975, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8350 + }, + { + "epoch": 3.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8351 + }, + { + "epoch": 3.3367958449860167, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8352 + }, + { + "epoch": 3.3371953655613265, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8353 + }, + { + "epoch": 3.3375948861366362, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8354 + }, + { + "epoch": 3.3379944067119456, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8355 + }, + { + "epoch": 3.3383939272872554, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8356 + }, + { + "epoch": 3.3387934478625647, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8357 + }, + { + "epoch": 3.3391929684378745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8358 + }, + { + "epoch": 3.3395924890131843, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 8359 + }, + { + "epoch": 3.3399920095884936, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8360 + }, + { + "epoch": 3.3403915301638034, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8361 + }, + { + "epoch": 3.340791050739113, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8362 + }, + { + "epoch": 3.3411905713144225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 8363 + }, + { + "epoch": 3.3415900918897323, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8364 + }, + { + "epoch": 3.341989612465042, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8365 + }, + { + "epoch": 3.3423891330403515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8366 + }, + { + "epoch": 3.3427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8367 + }, + { + "epoch": 3.343188174190971, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8368 + }, + { + "epoch": 3.3435876947662804, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8369 + }, + { + "epoch": 3.34398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8370 + }, + { + "epoch": 3.3443867359169, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8371 + }, + { + "epoch": 3.3447862564922093, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8372 + }, + { + "epoch": 3.345185777067519, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8373 + }, + { + "epoch": 3.3455852976428284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8374 + }, + { + "epoch": 3.345984818218138, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8375 + }, + { + "epoch": 3.346384338793448, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8376 + }, + { + "epoch": 3.3467838593687573, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8377 + }, + { + "epoch": 3.347183379944067, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8378 + }, + { + "epoch": 3.347582900519377, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8379 + }, + { + "epoch": 3.3479824210946862, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8380 + }, + { + "epoch": 3.348381941669996, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8381 + }, + { + "epoch": 3.348781462245306, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8382 + }, + { + "epoch": 3.349180982820615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8383 + }, + { + "epoch": 3.349580503395925, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8384 + }, + { + "epoch": 3.3499800239712343, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8385 + }, + { + "epoch": 3.350379544546544, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8386 + }, + { + "epoch": 3.350779065121854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8387 + }, + { + "epoch": 3.3511785856971636, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8388 + }, + { + "epoch": 3.351578106272473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8389 + }, + { + "epoch": 3.3519776268477828, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8390 + }, + { + "epoch": 3.352377147423092, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8391 + }, + { + "epoch": 3.352776667998402, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8392 + }, + { + "epoch": 3.3531761885737117, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8393 + }, + { + "epoch": 3.353575709149021, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8394 + }, + { + "epoch": 3.353975229724331, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8395 + }, + { + "epoch": 3.3543747502996406, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8396 + }, + { + "epoch": 3.35477427087495, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8397 + }, + { + "epoch": 3.3551737914502597, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8398 + }, + { + "epoch": 3.3555733120255695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8399 + }, + { + "epoch": 3.355972832600879, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 8400 + }, + { + "epoch": 3.3563723531761886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8401 + }, + { + "epoch": 3.356771873751498, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8402 + }, + { + "epoch": 3.357171394326808, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8403 + }, + { + "epoch": 3.3575709149021176, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8404 + }, + { + "epoch": 3.357970435477427, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8405 + }, + { + "epoch": 3.3583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 8406 + }, + { + "epoch": 3.3587694766280465, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8407 + }, + { + "epoch": 3.359168997203356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 8408 + }, + { + "epoch": 3.3595685177786656, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8409 + }, + { + "epoch": 3.3599680383539754, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8410 + }, + { + "epoch": 3.3603675589292847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8411 + }, + { + "epoch": 3.3607670795045945, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8412 + }, + { + "epoch": 3.361166600079904, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 8413 + }, + { + "epoch": 3.3615661206552137, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8414 + }, + { + "epoch": 3.3619656412305234, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8415 + }, + { + "epoch": 3.3623651618058332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 8416 + }, + { + "epoch": 3.3627646823811426, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8417 + }, + { + "epoch": 3.3631642029564524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8418 + }, + { + "epoch": 3.3635637235317617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8419 + }, + { + "epoch": 3.3639632441070715, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 8420 + }, + { + "epoch": 3.3643627646823813, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8421 + }, + { + "epoch": 3.3647622852576906, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8422 + }, + { + "epoch": 3.3651618058330004, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8423 + }, + { + "epoch": 3.36556132640831, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8424 + }, + { + "epoch": 3.3659608469836195, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8425 + }, + { + "epoch": 3.3663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8426 + }, + { + "epoch": 3.366759888134239, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8427 + }, + { + "epoch": 3.3671594087095484, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8428 + }, + { + "epoch": 3.3675589292848582, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8429 + }, + { + "epoch": 3.3679584498601676, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8430 + }, + { + "epoch": 3.3683579704354774, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8431 + }, + { + "epoch": 3.368757491010787, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8432 + }, + { + "epoch": 3.369157011586097, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8433 + }, + { + "epoch": 3.3695565321614063, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8434 + }, + { + "epoch": 3.369956052736716, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8435 + }, + { + "epoch": 3.3703555733120254, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8436 + }, + { + "epoch": 3.370755093887335, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8437 + }, + { + "epoch": 3.371154614462645, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8438 + }, + { + "epoch": 3.3715541350379543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8439 + }, + { + "epoch": 3.371953655613264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8440 + }, + { + "epoch": 3.372353176188574, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8441 + }, + { + "epoch": 3.3727526967638832, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8442 + }, + { + "epoch": 3.373152217339193, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8443 + }, + { + "epoch": 3.373551737914503, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8444 + }, + { + "epoch": 3.373951258489812, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8445 + }, + { + "epoch": 3.374350779065122, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8446 + }, + { + "epoch": 3.3747502996404313, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8447 + }, + { + "epoch": 3.375149820215741, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8448 + }, + { + "epoch": 3.375549340791051, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8449 + }, + { + "epoch": 3.37594886136636, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8450 + }, + { + "epoch": 3.37634838194167, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8451 + }, + { + "epoch": 3.3767479025169798, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8452 + }, + { + "epoch": 3.377147423092289, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 8453 + }, + { + "epoch": 3.377546943667599, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8454 + }, + { + "epoch": 3.3779464642429087, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8455 + }, + { + "epoch": 3.378345984818218, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8456 + }, + { + "epoch": 3.378745505393528, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8457 + }, + { + "epoch": 3.379145025968837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8458 + }, + { + "epoch": 3.379544546544147, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8459 + }, + { + "epoch": 3.3799440671194567, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8460 + }, + { + "epoch": 3.3803435876947665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8461 + }, + { + "epoch": 3.380743108270076, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8462 + }, + { + "epoch": 3.3811426288453856, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 8463 + }, + { + "epoch": 3.381542149420695, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8464 + }, + { + "epoch": 3.3819416699960048, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8465 + }, + { + "epoch": 3.3823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 8466 + }, + { + "epoch": 3.382740711146624, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 8467 + }, + { + "epoch": 3.3831402317219337, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8468 + }, + { + "epoch": 3.3835397522972435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8469 + }, + { + "epoch": 3.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8470 + }, + { + "epoch": 3.3843387934478626, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8471 + }, + { + "epoch": 3.3847383140231724, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8472 + }, + { + "epoch": 3.3851378345984817, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8473 + }, + { + "epoch": 3.3855373551737915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 8474 + }, + { + "epoch": 3.385936875749101, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8475 + }, + { + "epoch": 3.3863363963244106, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8476 + }, + { + "epoch": 3.3867359168997204, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8477 + }, + { + "epoch": 3.38713543747503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8478 + }, + { + "epoch": 3.3875349580503396, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8479 + }, + { + "epoch": 3.3879344786256493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8480 + }, + { + "epoch": 3.3883339992009587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8481 + }, + { + "epoch": 3.3887335197762685, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8482 + }, + { + "epoch": 3.3891330403515783, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8483 + }, + { + "epoch": 3.3895325609268876, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8484 + }, + { + "epoch": 3.3899320815021974, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8485 + }, + { + "epoch": 3.390331602077507, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8486 + }, + { + "epoch": 3.3907311226528165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8487 + }, + { + "epoch": 3.3911306432281263, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8488 + }, + { + "epoch": 3.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8489 + }, + { + "epoch": 3.3919296843787454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8490 + }, + { + "epoch": 3.392329204954055, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8491 + }, + { + "epoch": 3.3927287255293646, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8492 + }, + { + "epoch": 3.3931282461046743, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8493 + }, + { + "epoch": 3.393527766679984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8494 + }, + { + "epoch": 3.3939272872552935, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8495 + }, + { + "epoch": 3.3943268078306033, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8496 + }, + { + "epoch": 3.394726328405913, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8497 + }, + { + "epoch": 3.3951258489812224, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8498 + }, + { + "epoch": 3.395525369556532, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8499 + }, + { + "epoch": 3.395924890131842, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8500 + }, + { + "epoch": 3.3963244107071513, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8501 + }, + { + "epoch": 3.396723931282461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 8502 + }, + { + "epoch": 3.3971234518577704, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8503 + }, + { + "epoch": 3.39752297243308, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8504 + }, + { + "epoch": 3.39792249300839, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8505 + }, + { + "epoch": 3.3983220135837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8506 + }, + { + "epoch": 3.398721534159009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8507 + }, + { + "epoch": 3.399121054734319, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8508 + }, + { + "epoch": 3.3995205753096283, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8509 + }, + { + "epoch": 3.399920095884938, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8510 + }, + { + "epoch": 3.400319616460248, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8511 + }, + { + "epoch": 3.400719137035557, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8512 + }, + { + "epoch": 3.401118657610867, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8513 + }, + { + "epoch": 3.4015181781861767, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8514 + }, + { + "epoch": 3.401917698761486, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8515 + }, + { + "epoch": 3.402317219336796, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8516 + }, + { + "epoch": 3.4027167399121057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8517 + }, + { + "epoch": 3.403116260487415, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8518 + }, + { + "epoch": 3.403515781062725, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 8519 + }, + { + "epoch": 3.403915301638034, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8520 + }, + { + "epoch": 3.404314822213344, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8521 + }, + { + "epoch": 3.4047143427886537, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8522 + }, + { + "epoch": 3.4051138633639635, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8523 + }, + { + "epoch": 3.405513383939273, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8524 + }, + { + "epoch": 3.4059129045145826, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8525 + }, + { + "epoch": 3.406312425089892, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8526 + }, + { + "epoch": 3.4067119456652017, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8527 + }, + { + "epoch": 3.4071114662405115, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8528 + }, + { + "epoch": 3.407510986815821, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8529 + }, + { + "epoch": 3.4079105073911307, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8530 + }, + { + "epoch": 3.4083100279664404, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8531 + }, + { + "epoch": 3.40870954854175, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8532 + }, + { + "epoch": 3.4091090691170596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8533 + }, + { + "epoch": 3.4095085896923694, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8534 + }, + { + "epoch": 3.4099081102676787, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8535 + }, + { + "epoch": 3.4103076308429885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8536 + }, + { + "epoch": 3.410707151418298, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8537 + }, + { + "epoch": 3.4111066719936076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 8538 + }, + { + "epoch": 3.4115061925689174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8539 + }, + { + "epoch": 3.4119057131442267, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8540 + }, + { + "epoch": 3.4123052337195365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8541 + }, + { + "epoch": 3.4127047542948463, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8542 + }, + { + "epoch": 3.4131042748701557, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8543 + }, + { + "epoch": 3.4135037954454654, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8544 + }, + { + "epoch": 3.4139033160207752, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8545 + }, + { + "epoch": 3.4143028365960846, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8546 + }, + { + "epoch": 3.4147023571713944, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8547 + }, + { + "epoch": 3.4151018777467037, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8548 + }, + { + "epoch": 3.4155013983220135, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8549 + }, + { + "epoch": 3.4159009188973233, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8550 + }, + { + "epoch": 3.416300439472633, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8551 + }, + { + "epoch": 3.4166999600479424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8552 + }, + { + "epoch": 3.417099480623252, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8553 + }, + { + "epoch": 3.4174990011985615, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8554 + }, + { + "epoch": 3.4178985217738713, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8555 + }, + { + "epoch": 3.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8556 + }, + { + "epoch": 3.4186975629244905, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8557 + }, + { + "epoch": 3.4190970834998002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8558 + }, + { + "epoch": 3.41949660407511, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8559 + }, + { + "epoch": 3.4198961246504194, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8560 + }, + { + "epoch": 3.420295645225729, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8561 + }, + { + "epoch": 3.420695165801039, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8562 + }, + { + "epoch": 3.4210946863763483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8563 + }, + { + "epoch": 3.421494206951658, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8564 + }, + { + "epoch": 3.4218937275269674, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8565 + }, + { + "epoch": 3.422293248102277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8566 + }, + { + "epoch": 3.422692768677587, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8567 + }, + { + "epoch": 3.4230922892528968, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8568 + }, + { + "epoch": 3.423491809828206, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8569 + }, + { + "epoch": 3.423891330403516, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8570 + }, + { + "epoch": 3.4242908509788252, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8571 + }, + { + "epoch": 3.424690371554135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8572 + }, + { + "epoch": 3.425089892129445, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 8573 + }, + { + "epoch": 3.425489412704754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8574 + }, + { + "epoch": 3.425888933280064, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 8575 + }, + { + "epoch": 3.4262884538553737, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8576 + }, + { + "epoch": 3.426687974430683, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 8577 + }, + { + "epoch": 3.427087495005993, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8578 + }, + { + "epoch": 3.4274870155813026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8579 + }, + { + "epoch": 3.427886536156612, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8580 + }, + { + "epoch": 3.4282860567319218, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8581 + }, + { + "epoch": 3.428685577307231, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8582 + }, + { + "epoch": 3.429085097882541, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8583 + }, + { + "epoch": 3.4294846184578507, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8584 + }, + { + "epoch": 3.42988413903316, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8585 + }, + { + "epoch": 3.43028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8586 + }, + { + "epoch": 3.4306831801837796, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8587 + }, + { + "epoch": 3.431082700759089, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8588 + }, + { + "epoch": 3.4314822213343987, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8589 + }, + { + "epoch": 3.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8590 + }, + { + "epoch": 3.432281262485018, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8591 + }, + { + "epoch": 3.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 8592 + }, + { + "epoch": 3.433080303635637, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8593 + }, + { + "epoch": 3.4334798242109468, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 8594 + }, + { + "epoch": 3.4338793447862566, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8595 + }, + { + "epoch": 3.4342788653615663, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8596 + }, + { + "epoch": 3.4346783859368757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8597 + }, + { + "epoch": 3.4350779065121855, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 8598 + }, + { + "epoch": 3.435477427087495, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8599 + }, + { + "epoch": 3.4358769476628046, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8600 + }, + { + "epoch": 3.4362764682381144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8601 + }, + { + "epoch": 3.4366759888134237, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8602 + }, + { + "epoch": 3.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8603 + }, + { + "epoch": 3.4374750299640433, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8604 + }, + { + "epoch": 3.4378745505393526, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8605 + }, + { + "epoch": 3.4382740711146624, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8606 + }, + { + "epoch": 3.438673591689972, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8607 + }, + { + "epoch": 3.4390731122652816, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8608 + }, + { + "epoch": 3.4394726328405913, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8609 + }, + { + "epoch": 3.4398721534159007, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8610 + }, + { + "epoch": 3.4402716739912105, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 8611 + }, + { + "epoch": 3.4406711945665203, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8612 + }, + { + "epoch": 3.44107071514183, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8613 + }, + { + "epoch": 3.4414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8614 + }, + { + "epoch": 3.441869756292449, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8615 + }, + { + "epoch": 3.4422692768677585, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8616 + }, + { + "epoch": 3.4426687974430683, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8617 + }, + { + "epoch": 3.443068318018378, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8618 + }, + { + "epoch": 3.4434678385936874, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8619 + }, + { + "epoch": 3.443867359168997, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8620 + }, + { + "epoch": 3.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8621 + }, + { + "epoch": 3.4446664003196164, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8622 + }, + { + "epoch": 3.445065920894926, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8623 + }, + { + "epoch": 3.445465441470236, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8624 + }, + { + "epoch": 3.4458649620455453, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8625 + }, + { + "epoch": 3.446264482620855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8626 + }, + { + "epoch": 3.4466640031961644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8627 + }, + { + "epoch": 3.447063523771474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 8628 + }, + { + "epoch": 3.447463044346784, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8629 + }, + { + "epoch": 3.4478625649220933, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8630 + }, + { + "epoch": 3.448262085497403, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8631 + }, + { + "epoch": 3.448661606072713, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8632 + }, + { + "epoch": 3.4490611266480222, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8633 + }, + { + "epoch": 3.449460647223332, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8634 + }, + { + "epoch": 3.449860167798642, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8635 + }, + { + "epoch": 3.450259688373951, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8636 + }, + { + "epoch": 3.450659208949261, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8637 + }, + { + "epoch": 3.4510587295245703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8638 + }, + { + "epoch": 3.45145825009988, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8639 + }, + { + "epoch": 3.45185777067519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 8640 + }, + { + "epoch": 3.4522572912504996, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8641 + }, + { + "epoch": 3.452656811825809, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8642 + }, + { + "epoch": 3.4530563324011188, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8643 + }, + { + "epoch": 3.453455852976428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8644 + }, + { + "epoch": 3.453855373551738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8645 + }, + { + "epoch": 3.4542548941270477, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8646 + }, + { + "epoch": 3.454654414702357, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8647 + }, + { + "epoch": 3.455053935277667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8648 + }, + { + "epoch": 3.4554534558529766, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8649 + }, + { + "epoch": 3.455852976428286, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8650 + }, + { + "epoch": 3.4562524970035957, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 8651 + }, + { + "epoch": 3.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8652 + }, + { + "epoch": 3.457051538154215, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8653 + }, + { + "epoch": 3.4574510587295246, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8654 + }, + { + "epoch": 3.457850579304834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8655 + }, + { + "epoch": 3.4582500998801438, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8656 + }, + { + "epoch": 3.4586496204554535, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8657 + }, + { + "epoch": 3.4590491410307633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8658 + }, + { + "epoch": 3.4594486616060727, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8659 + }, + { + "epoch": 3.4598481821813825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 8660 + }, + { + "epoch": 3.460247702756692, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8661 + }, + { + "epoch": 3.4606472233320016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8662 + }, + { + "epoch": 3.4610467439073114, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8663 + }, + { + "epoch": 3.4614462644826207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8664 + }, + { + "epoch": 3.4618457850579305, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8665 + }, + { + "epoch": 3.4622453056332403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8666 + }, + { + "epoch": 3.4626448262085496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8667 + }, + { + "epoch": 3.4630443467838594, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8668 + }, + { + "epoch": 3.463443867359169, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8669 + }, + { + "epoch": 3.4638433879344785, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8670 + }, + { + "epoch": 3.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8671 + }, + { + "epoch": 3.4646424290850977, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8672 + }, + { + "epoch": 3.4650419496604075, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 8673 + }, + { + "epoch": 3.4654414702357172, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8674 + }, + { + "epoch": 3.4658409908110266, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 8675 + }, + { + "epoch": 3.4662405113863364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8676 + }, + { + "epoch": 3.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8677 + }, + { + "epoch": 3.4670395525369555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8678 + }, + { + "epoch": 3.4674390731122653, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8679 + }, + { + "epoch": 3.467838593687575, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8680 + }, + { + "epoch": 3.4682381142628844, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8681 + }, + { + "epoch": 3.468637634838194, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8682 + }, + { + "epoch": 3.4690371554135035, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8683 + }, + { + "epoch": 3.4694366759888133, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8684 + }, + { + "epoch": 3.469836196564123, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 8685 + }, + { + "epoch": 3.470235717139433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8686 + }, + { + "epoch": 3.4706352377147422, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8687 + }, + { + "epoch": 3.471034758290052, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8688 + }, + { + "epoch": 3.4714342788653614, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8689 + }, + { + "epoch": 3.471833799440671, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8690 + }, + { + "epoch": 3.472233320015981, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8691 + }, + { + "epoch": 3.4726328405912903, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8692 + }, + { + "epoch": 3.4730323611666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8693 + }, + { + "epoch": 3.47343188174191, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8694 + }, + { + "epoch": 3.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8695 + }, + { + "epoch": 3.474230922892529, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8696 + }, + { + "epoch": 3.474630443467839, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8697 + }, + { + "epoch": 3.475029964043148, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8698 + }, + { + "epoch": 3.475429484618458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 8699 + }, + { + "epoch": 3.4758290051937673, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8700 + }, + { + "epoch": 3.476228525769077, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8701 + }, + { + "epoch": 3.476628046344387, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8702 + }, + { + "epoch": 3.4770275669196966, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8703 + }, + { + "epoch": 3.477427087495006, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8704 + }, + { + "epoch": 3.4778266080703157, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8705 + }, + { + "epoch": 3.478226128645625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8706 + }, + { + "epoch": 3.478625649220935, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8707 + }, + { + "epoch": 3.4790251697962447, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8708 + }, + { + "epoch": 3.479424690371554, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8709 + }, + { + "epoch": 3.479824210946864, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8710 + }, + { + "epoch": 3.4802237315221736, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8711 + }, + { + "epoch": 3.480623252097483, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8712 + }, + { + "epoch": 3.4810227726727927, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8713 + }, + { + "epoch": 3.4814222932481025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8714 + }, + { + "epoch": 3.481821813823412, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8715 + }, + { + "epoch": 3.4822213343987216, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8716 + }, + { + "epoch": 3.482620854974031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8717 + }, + { + "epoch": 3.4830203755493407, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 8718 + }, + { + "epoch": 3.4834198961246505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8719 + }, + { + "epoch": 3.48381941669996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8720 + }, + { + "epoch": 3.4842189372752697, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8721 + }, + { + "epoch": 3.4846184578505794, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8722 + }, + { + "epoch": 3.485017978425889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8723 + }, + { + "epoch": 3.4854174990011986, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8724 + }, + { + "epoch": 3.4858170195765084, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8725 + }, + { + "epoch": 3.4862165401518177, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8726 + }, + { + "epoch": 3.4866160607271275, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 8727 + }, + { + "epoch": 3.487015581302437, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8728 + }, + { + "epoch": 3.4874151018777466, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8729 + }, + { + "epoch": 3.4878146224530564, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8730 + }, + { + "epoch": 3.488214143028366, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8731 + }, + { + "epoch": 3.4886136636036755, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8732 + }, + { + "epoch": 3.4890131841789853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8733 + }, + { + "epoch": 3.4894127047542947, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8734 + }, + { + "epoch": 3.4898122253296044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8735 + }, + { + "epoch": 3.4902117459049142, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8736 + }, + { + "epoch": 3.4906112664802236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8737 + }, + { + "epoch": 3.4910107870555334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8738 + }, + { + "epoch": 3.491410307630843, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8739 + }, + { + "epoch": 3.4918098282061525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8740 + }, + { + "epoch": 3.4922093487814623, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8741 + }, + { + "epoch": 3.492608869356772, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8742 + }, + { + "epoch": 3.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8743 + }, + { + "epoch": 3.493407910507391, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8744 + }, + { + "epoch": 3.4938074310827005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8745 + }, + { + "epoch": 3.4942069516580103, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8746 + }, + { + "epoch": 3.49460647223332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 8747 + }, + { + "epoch": 3.49500599280863, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8748 + }, + { + "epoch": 3.4954055133839392, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8749 + }, + { + "epoch": 3.495805033959249, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8750 + }, + { + "epoch": 3.4962045545345584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8751 + }, + { + "epoch": 3.496604075109868, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8752 + }, + { + "epoch": 3.497003595685178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8753 + }, + { + "epoch": 3.4974031162604873, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 8754 + }, + { + "epoch": 3.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8755 + }, + { + "epoch": 3.498202157411107, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8756 + }, + { + "epoch": 3.498601677986416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8757 + }, + { + "epoch": 3.499001198561726, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8758 + }, + { + "epoch": 3.4994007191370358, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8759 + }, + { + "epoch": 3.499800239712345, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8760 + }, + { + "epoch": 3.500199760287655, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8761 + }, + { + "epoch": 3.5005992808629642, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8762 + }, + { + "epoch": 3.500998801438274, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8763 + }, + { + "epoch": 3.501398322013584, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8764 + }, + { + "epoch": 3.5017978425888936, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8765 + }, + { + "epoch": 3.502197363164203, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8766 + }, + { + "epoch": 3.5025968837395127, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8767 + }, + { + "epoch": 3.502996404314822, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8768 + }, + { + "epoch": 3.503395924890132, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8769 + }, + { + "epoch": 3.5037954454654416, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 8770 + }, + { + "epoch": 3.504194966040751, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8771 + }, + { + "epoch": 3.5045944866160608, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8772 + }, + { + "epoch": 3.50499400719137, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8773 + }, + { + "epoch": 3.50539352776668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8774 + }, + { + "epoch": 3.5057930483419897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8775 + }, + { + "epoch": 3.5061925689172995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8776 + }, + { + "epoch": 3.506592089492609, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8777 + }, + { + "epoch": 3.5069916100679186, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8778 + }, + { + "epoch": 3.507391130643228, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 8779 + }, + { + "epoch": 3.5077906512185377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8780 + }, + { + "epoch": 3.5081901717938475, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8781 + }, + { + "epoch": 3.508589692369157, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8782 + }, + { + "epoch": 3.5089892129444666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8783 + }, + { + "epoch": 3.5093887335197764, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8784 + }, + { + "epoch": 3.5097882540950858, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8785 + }, + { + "epoch": 3.5101877746703956, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8786 + }, + { + "epoch": 3.5105872952457053, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8787 + }, + { + "epoch": 3.5109868158210147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8788 + }, + { + "epoch": 3.5113863363963245, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8789 + }, + { + "epoch": 3.511785856971634, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8790 + }, + { + "epoch": 3.5121853775469436, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8791 + }, + { + "epoch": 3.5125848981222534, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8792 + }, + { + "epoch": 3.512984418697563, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8793 + }, + { + "epoch": 3.5133839392728725, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8794 + }, + { + "epoch": 3.5137834598481823, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8795 + }, + { + "epoch": 3.5141829804234916, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8796 + }, + { + "epoch": 3.5145825009988014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8797 + }, + { + "epoch": 3.514982021574111, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 8798 + }, + { + "epoch": 3.5153815421494206, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8799 + }, + { + "epoch": 3.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8800 + }, + { + "epoch": 3.5161805833000397, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8801 + }, + { + "epoch": 3.5165801038753495, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8802 + }, + { + "epoch": 3.5169796244506593, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8803 + }, + { + "epoch": 3.517379145025969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8804 + }, + { + "epoch": 3.5177786656012784, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8805 + }, + { + "epoch": 3.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8806 + }, + { + "epoch": 3.5185777067518975, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8807 + }, + { + "epoch": 3.5189772273272073, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8808 + }, + { + "epoch": 3.519376747902517, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8809 + }, + { + "epoch": 3.519776268477827, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8810 + }, + { + "epoch": 3.520175789053136, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8811 + }, + { + "epoch": 3.520575309628446, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8812 + }, + { + "epoch": 3.5209748302037553, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 8813 + }, + { + "epoch": 3.521374350779065, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8814 + }, + { + "epoch": 3.521773871354375, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8815 + }, + { + "epoch": 3.5221733919296843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8816 + }, + { + "epoch": 3.522572912504994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8817 + }, + { + "epoch": 3.5229724330803034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8818 + }, + { + "epoch": 3.523371953655613, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8819 + }, + { + "epoch": 3.523771474230923, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8820 + }, + { + "epoch": 3.5241709948062327, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8821 + }, + { + "epoch": 3.524570515381542, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8822 + }, + { + "epoch": 3.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8823 + }, + { + "epoch": 3.525369556532161, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8824 + }, + { + "epoch": 3.525769077107471, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8825 + }, + { + "epoch": 3.526168597682781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8826 + }, + { + "epoch": 3.52656811825809, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8827 + }, + { + "epoch": 3.5269676388334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8828 + }, + { + "epoch": 3.5273671594087097, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8829 + }, + { + "epoch": 3.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8830 + }, + { + "epoch": 3.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8831 + }, + { + "epoch": 3.5285657211346386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8832 + }, + { + "epoch": 3.528965241709948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8833 + }, + { + "epoch": 3.5293647622852578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8834 + }, + { + "epoch": 3.529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8835 + }, + { + "epoch": 3.530163803435877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8836 + }, + { + "epoch": 3.5305633240111867, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8837 + }, + { + "epoch": 3.5309628445864965, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8838 + }, + { + "epoch": 3.531362365161806, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8839 + }, + { + "epoch": 3.5317618857371156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8840 + }, + { + "epoch": 3.532161406312425, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8841 + }, + { + "epoch": 3.5325609268877347, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8842 + }, + { + "epoch": 3.5329604474630445, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8843 + }, + { + "epoch": 3.533359968038354, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8844 + }, + { + "epoch": 3.5337594886136636, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8845 + }, + { + "epoch": 3.534159009188973, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8846 + }, + { + "epoch": 3.5345585297642828, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8847 + }, + { + "epoch": 3.5349580503395925, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8848 + }, + { + "epoch": 3.5353575709149023, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8849 + }, + { + "epoch": 3.5357570914902117, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8850 + }, + { + "epoch": 3.5361566120655215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8851 + }, + { + "epoch": 3.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8852 + }, + { + "epoch": 3.5369556532161406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8853 + }, + { + "epoch": 3.5373551737914504, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8854 + }, + { + "epoch": 3.53775469436676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8855 + }, + { + "epoch": 3.5381542149420695, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8856 + }, + { + "epoch": 3.5385537355173793, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8857 + }, + { + "epoch": 3.5389532560926886, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8858 + }, + { + "epoch": 3.5393527766679984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8859 + }, + { + "epoch": 3.539752297243308, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8860 + }, + { + "epoch": 3.5401518178186175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8861 + }, + { + "epoch": 3.5405513383939273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8862 + }, + { + "epoch": 3.5409508589692367, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8863 + }, + { + "epoch": 3.5413503795445465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8864 + }, + { + "epoch": 3.5417499001198562, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 8865 + }, + { + "epoch": 3.542149420695166, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8866 + }, + { + "epoch": 3.5425489412704754, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8867 + }, + { + "epoch": 3.542948461845785, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8868 + }, + { + "epoch": 3.5433479824210945, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 8869 + }, + { + "epoch": 3.5437475029964043, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8870 + }, + { + "epoch": 3.544147023571714, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8871 + }, + { + "epoch": 3.5445465441470234, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8872 + }, + { + "epoch": 3.544946064722333, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8873 + }, + { + "epoch": 3.545345585297643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8874 + }, + { + "epoch": 3.5457451058729523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8875 + }, + { + "epoch": 3.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8876 + }, + { + "epoch": 3.546544147023572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8877 + }, + { + "epoch": 3.5469436675988812, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8878 + }, + { + "epoch": 3.547343188174191, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8879 + }, + { + "epoch": 3.5477427087495004, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8880 + }, + { + "epoch": 3.54814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8881 + }, + { + "epoch": 3.54854174990012, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8882 + }, + { + "epoch": 3.5489412704754297, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8883 + }, + { + "epoch": 3.549340791050739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8884 + }, + { + "epoch": 3.549740311626049, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8885 + }, + { + "epoch": 3.550139832201358, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8886 + }, + { + "epoch": 3.550539352776668, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8887 + }, + { + "epoch": 3.5509388733519778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8888 + }, + { + "epoch": 3.551338393927287, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8889 + }, + { + "epoch": 3.551737914502597, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8890 + }, + { + "epoch": 3.5521374350779062, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8891 + }, + { + "epoch": 3.552536955653216, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8892 + }, + { + "epoch": 3.552936476228526, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8893 + }, + { + "epoch": 3.5533359968038356, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8894 + }, + { + "epoch": 3.553735517379145, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8895 + }, + { + "epoch": 3.5541350379544547, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8896 + }, + { + "epoch": 3.554534558529764, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8897 + }, + { + "epoch": 3.554934079105074, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8898 + }, + { + "epoch": 3.5553335996803836, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8899 + }, + { + "epoch": 3.5557331202556934, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8900 + }, + { + "epoch": 3.556132640831003, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8901 + }, + { + "epoch": 3.5565321614063126, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8902 + }, + { + "epoch": 3.556931681981622, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8903 + }, + { + "epoch": 3.5573312025569317, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8904 + }, + { + "epoch": 3.5577307231322415, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8905 + }, + { + "epoch": 3.558130243707551, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8906 + }, + { + "epoch": 3.5585297642828606, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8907 + }, + { + "epoch": 3.55892928485817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8908 + }, + { + "epoch": 3.5593288054334797, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8909 + }, + { + "epoch": 3.5597283260087895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8910 + }, + { + "epoch": 3.5601278465840993, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8911 + }, + { + "epoch": 3.5605273671594087, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8912 + }, + { + "epoch": 3.5609268877347184, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8913 + }, + { + "epoch": 3.561326408310028, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8914 + }, + { + "epoch": 3.5617259288853376, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8915 + }, + { + "epoch": 3.5621254494606474, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8916 + }, + { + "epoch": 3.5625249700359567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8917 + }, + { + "epoch": 3.5629244906112665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8918 + }, + { + "epoch": 3.5633240111865763, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8919 + }, + { + "epoch": 3.5637235317618856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8920 + }, + { + "epoch": 3.5641230523371954, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8921 + }, + { + "epoch": 3.564522572912505, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8922 + }, + { + "epoch": 3.5649220934878145, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8923 + }, + { + "epoch": 3.5653216140631243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8924 + }, + { + "epoch": 3.5657211346384337, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 8925 + }, + { + "epoch": 3.5661206552137434, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8926 + }, + { + "epoch": 3.5665201757890532, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8927 + }, + { + "epoch": 3.566919696364363, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8928 + }, + { + "epoch": 3.5673192169396724, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8929 + }, + { + "epoch": 3.567718737514982, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8930 + }, + { + "epoch": 3.5681182580902915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8931 + }, + { + "epoch": 3.5685177786656013, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8932 + }, + { + "epoch": 3.568917299240911, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 8933 + }, + { + "epoch": 3.5693168198162204, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8934 + }, + { + "epoch": 3.56971634039153, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8935 + }, + { + "epoch": 3.5701158609668395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8936 + }, + { + "epoch": 3.5705153815421493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8937 + }, + { + "epoch": 3.570914902117459, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8938 + }, + { + "epoch": 3.571314422692769, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8939 + }, + { + "epoch": 3.5717139432680782, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8940 + }, + { + "epoch": 3.572113463843388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8941 + }, + { + "epoch": 3.5725129844186974, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8942 + }, + { + "epoch": 3.572912504994007, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 8943 + }, + { + "epoch": 3.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8944 + }, + { + "epoch": 3.5737115461446267, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8945 + }, + { + "epoch": 3.574111066719936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8946 + }, + { + "epoch": 3.574510587295246, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8947 + }, + { + "epoch": 3.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8948 + }, + { + "epoch": 3.575309628445865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 8949 + }, + { + "epoch": 3.5757091490211748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8950 + }, + { + "epoch": 3.576108669596484, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8951 + }, + { + "epoch": 3.576508190171794, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8952 + }, + { + "epoch": 3.5769077107471032, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8953 + }, + { + "epoch": 3.577307231322413, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8954 + }, + { + "epoch": 3.577706751897723, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8955 + }, + { + "epoch": 3.5781062724730326, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8956 + }, + { + "epoch": 3.578505793048342, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8957 + }, + { + "epoch": 3.5789053136236517, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8958 + }, + { + "epoch": 3.579304834198961, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8959 + }, + { + "epoch": 3.579704354774271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8960 + }, + { + "epoch": 3.5801038753495806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8961 + }, + { + "epoch": 3.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8962 + }, + { + "epoch": 3.5809029165001998, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8963 + }, + { + "epoch": 3.5813024370755095, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8964 + }, + { + "epoch": 3.581701957650819, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8965 + }, + { + "epoch": 3.5821014782261287, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8966 + }, + { + "epoch": 3.5825009988014385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8967 + }, + { + "epoch": 3.582900519376748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8968 + }, + { + "epoch": 3.5833000399520576, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8969 + }, + { + "epoch": 3.583699560527367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8970 + }, + { + "epoch": 3.5840990811026767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8971 + }, + { + "epoch": 3.5844986016779865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8972 + }, + { + "epoch": 3.5848981222532963, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8973 + }, + { + "epoch": 3.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8974 + }, + { + "epoch": 3.5856971634039154, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 8975 + }, + { + "epoch": 3.5860966839792248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8976 + }, + { + "epoch": 3.5864962045545346, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8977 + }, + { + "epoch": 3.5868957251298443, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8978 + }, + { + "epoch": 3.5872952457051537, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8979 + }, + { + "epoch": 3.5876947662804635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8980 + }, + { + "epoch": 3.588094286855773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8981 + }, + { + "epoch": 3.5884938074310826, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 8982 + }, + { + "epoch": 3.5888933280063924, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8983 + }, + { + "epoch": 3.589292848581702, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8984 + }, + { + "epoch": 3.5896923691570115, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 8985 + }, + { + "epoch": 3.5900918897323213, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8986 + }, + { + "epoch": 3.5904914103076306, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 8987 + }, + { + "epoch": 3.5908909308829404, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8988 + }, + { + "epoch": 3.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8989 + }, + { + "epoch": 3.59168997203356, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8990 + }, + { + "epoch": 3.5920894926088693, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8991 + }, + { + "epoch": 3.592489013184179, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8992 + }, + { + "epoch": 3.5928885337594885, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8993 + }, + { + "epoch": 3.5932880543347983, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8994 + }, + { + "epoch": 3.593687574910108, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8995 + }, + { + "epoch": 3.5940870954854174, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 8996 + }, + { + "epoch": 3.594486616060727, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8997 + }, + { + "epoch": 3.5948861366360365, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8998 + }, + { + "epoch": 3.5952856572113463, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8999 + }, + { + "epoch": 3.595685177786656, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9000 + }, + { + "epoch": 3.596084698361966, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9001 + }, + { + "epoch": 3.596484218937275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9002 + }, + { + "epoch": 3.596883739512585, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 9003 + }, + { + "epoch": 3.5972832600878943, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9004 + }, + { + "epoch": 3.597682780663204, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9005 + }, + { + "epoch": 3.598082301238514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9006 + }, + { + "epoch": 3.5984818218138233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9007 + }, + { + "epoch": 3.598881342389133, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9008 + }, + { + "epoch": 3.599280862964443, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 9009 + }, + { + "epoch": 3.599680383539752, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9010 + }, + { + "epoch": 3.600079904115062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9011 + }, + { + "epoch": 3.6004794246903717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9012 + }, + { + "epoch": 3.600878945265681, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9013 + }, + { + "epoch": 3.601278465840991, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9014 + }, + { + "epoch": 3.6016779864163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9015 + }, + { + "epoch": 3.60207750699161, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9016 + }, + { + "epoch": 3.60247702756692, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9017 + }, + { + "epoch": 3.6028765481422296, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9018 + }, + { + "epoch": 3.603276068717539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9019 + }, + { + "epoch": 3.6036755892928487, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9020 + }, + { + "epoch": 3.604075109868158, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9021 + }, + { + "epoch": 3.604474630443468, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9022 + }, + { + "epoch": 3.6048741510187776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9023 + }, + { + "epoch": 3.605273671594087, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9024 + }, + { + "epoch": 3.6056731921693967, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9025 + }, + { + "epoch": 3.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9026 + }, + { + "epoch": 3.606472233320016, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9027 + }, + { + "epoch": 3.6068717538953257, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9028 + }, + { + "epoch": 3.6072712744706354, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9029 + }, + { + "epoch": 3.607670795045945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9030 + }, + { + "epoch": 3.6080703156212546, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9031 + }, + { + "epoch": 3.608469836196564, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9032 + }, + { + "epoch": 3.6088693567718737, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9033 + }, + { + "epoch": 3.6092688773471835, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 9034 + }, + { + "epoch": 3.6096683979224933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9035 + }, + { + "epoch": 3.6100679184978026, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9036 + }, + { + "epoch": 3.6104674390731124, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9037 + }, + { + "epoch": 3.6108669596484217, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9038 + }, + { + "epoch": 3.6112664802237315, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 9039 + }, + { + "epoch": 3.6116660007990413, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9040 + }, + { + "epoch": 3.6120655213743507, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9041 + }, + { + "epoch": 3.6124650419496604, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9042 + }, + { + "epoch": 3.61286456252497, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9043 + }, + { + "epoch": 3.6132640831002796, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9044 + }, + { + "epoch": 3.6136636036755894, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9045 + }, + { + "epoch": 3.614063124250899, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9046 + }, + { + "epoch": 3.6144626448262085, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9047 + }, + { + "epoch": 3.6148621654015183, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9048 + }, + { + "epoch": 3.6152616859768276, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9049 + }, + { + "epoch": 3.6156612065521374, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9050 + }, + { + "epoch": 3.616060727127447, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9051 + }, + { + "epoch": 3.6164602477027565, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9052 + }, + { + "epoch": 3.6168597682780663, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9053 + }, + { + "epoch": 3.617259288853376, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9054 + }, + { + "epoch": 3.6176588094286855, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9055 + }, + { + "epoch": 3.6180583300039952, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9056 + }, + { + "epoch": 3.618457850579305, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9057 + }, + { + "epoch": 3.6188573711546144, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9058 + }, + { + "epoch": 3.619256891729924, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9059 + }, + { + "epoch": 3.6196564123052335, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9060 + }, + { + "epoch": 3.6200559328805433, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9061 + }, + { + "epoch": 3.620455453455853, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9062 + }, + { + "epoch": 3.620854974031163, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 9063 + }, + { + "epoch": 3.621254494606472, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9064 + }, + { + "epoch": 3.621654015181782, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9065 + }, + { + "epoch": 3.6220535357570913, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9066 + }, + { + "epoch": 3.622453056332401, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9067 + }, + { + "epoch": 3.622852576907711, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9068 + }, + { + "epoch": 3.6232520974830202, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9069 + }, + { + "epoch": 3.62365161805833, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9070 + }, + { + "epoch": 3.6240511386336394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9071 + }, + { + "epoch": 3.624450659208949, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9072 + }, + { + "epoch": 3.624850179784259, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9073 + }, + { + "epoch": 3.6252497003595687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9074 + }, + { + "epoch": 3.625649220934878, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 9075 + }, + { + "epoch": 3.626048741510188, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9076 + }, + { + "epoch": 3.626448262085497, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9077 + }, + { + "epoch": 3.626847782660807, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9078 + }, + { + "epoch": 3.6272473032361168, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9079 + }, + { + "epoch": 3.6276468238114266, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9080 + }, + { + "epoch": 3.628046344386736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9081 + }, + { + "epoch": 3.6284458649620457, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9082 + }, + { + "epoch": 3.628845385537355, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9083 + }, + { + "epoch": 3.629244906112665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9084 + }, + { + "epoch": 3.6296444266879746, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9085 + }, + { + "epoch": 3.630043947263284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9086 + }, + { + "epoch": 3.6304434678385937, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9087 + }, + { + "epoch": 3.630842988413903, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9088 + }, + { + "epoch": 3.631242508989213, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9089 + }, + { + "epoch": 3.6316420295645226, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9090 + }, + { + "epoch": 3.6320415501398324, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9091 + }, + { + "epoch": 3.6324410707151418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 9092 + }, + { + "epoch": 3.6328405912904516, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9093 + }, + { + "epoch": 3.633240111865761, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9094 + }, + { + "epoch": 3.6336396324410707, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9095 + }, + { + "epoch": 3.6340391530163805, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9096 + }, + { + "epoch": 3.63443867359169, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9097 + }, + { + "epoch": 3.6348381941669996, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9098 + }, + { + "epoch": 3.6352377147423094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9099 + }, + { + "epoch": 3.6356372353176187, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9100 + }, + { + "epoch": 3.6360367558929285, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9101 + }, + { + "epoch": 3.6364362764682383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9102 + }, + { + "epoch": 3.6368357970435476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9103 + }, + { + "epoch": 3.6372353176188574, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9104 + }, + { + "epoch": 3.6376348381941668, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9105 + }, + { + "epoch": 3.6380343587694766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9106 + }, + { + "epoch": 3.6384338793447863, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9107 + }, + { + "epoch": 3.638833399920096, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9108 + }, + { + "epoch": 3.6392329204954055, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9109 + }, + { + "epoch": 3.6396324410707153, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9110 + }, + { + "epoch": 3.6400319616460246, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9111 + }, + { + "epoch": 3.6404314822213344, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9112 + }, + { + "epoch": 3.640831002796644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9113 + }, + { + "epoch": 3.6412305233719535, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9114 + }, + { + "epoch": 3.6416300439472633, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9115 + }, + { + "epoch": 3.6420295645225726, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 9116 + }, + { + "epoch": 3.6424290850978824, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9117 + }, + { + "epoch": 3.642828605673192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9118 + }, + { + "epoch": 3.643228126248502, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9119 + }, + { + "epoch": 3.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9120 + }, + { + "epoch": 3.644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9121 + }, + { + "epoch": 3.6444266879744305, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9122 + }, + { + "epoch": 3.6448262085497403, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9123 + }, + { + "epoch": 3.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9124 + }, + { + "epoch": 3.64562524970036, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9125 + }, + { + "epoch": 3.646024770275669, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9126 + }, + { + "epoch": 3.646424290850979, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9127 + }, + { + "epoch": 3.6468238114262883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9128 + }, + { + "epoch": 3.647223332001598, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9129 + }, + { + "epoch": 3.647622852576908, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9130 + }, + { + "epoch": 3.6480223731522172, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9131 + }, + { + "epoch": 3.648421893727527, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9132 + }, + { + "epoch": 3.6488214143028364, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 9133 + }, + { + "epoch": 3.649220934878146, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9134 + }, + { + "epoch": 3.649620455453456, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9135 + }, + { + "epoch": 3.6500199760287657, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9136 + }, + { + "epoch": 3.650419496604075, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9137 + }, + { + "epoch": 3.650819017179385, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9138 + }, + { + "epoch": 3.651218537754694, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9139 + }, + { + "epoch": 3.651618058330004, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 9140 + }, + { + "epoch": 3.6520175789053138, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9141 + }, + { + "epoch": 3.652417099480623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9142 + }, + { + "epoch": 3.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9143 + }, + { + "epoch": 3.6532161406312427, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9144 + }, + { + "epoch": 3.653615661206552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 9145 + }, + { + "epoch": 3.654015181781862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9146 + }, + { + "epoch": 3.6544147023571716, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9147 + }, + { + "epoch": 3.654814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9148 + }, + { + "epoch": 3.6552137435077907, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9149 + }, + { + "epoch": 3.6556132640831, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9150 + }, + { + "epoch": 3.65601278465841, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9151 + }, + { + "epoch": 3.6564123052337196, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9152 + }, + { + "epoch": 3.6568118258090294, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9153 + }, + { + "epoch": 3.6572113463843388, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9154 + }, + { + "epoch": 3.6576108669596485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 9155 + }, + { + "epoch": 3.658010387534958, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9156 + }, + { + "epoch": 3.6584099081102677, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9157 + }, + { + "epoch": 3.6588094286855775, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9158 + }, + { + "epoch": 3.659208949260887, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9159 + }, + { + "epoch": 3.6596084698361966, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9160 + }, + { + "epoch": 3.660007990411506, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9161 + }, + { + "epoch": 3.6604075109868157, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9162 + }, + { + "epoch": 3.6608070315621255, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9163 + }, + { + "epoch": 3.6612065521374353, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9164 + }, + { + "epoch": 3.6616060727127446, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9165 + }, + { + "epoch": 3.6620055932880544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9166 + }, + { + "epoch": 3.6624051138633638, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9167 + }, + { + "epoch": 3.6628046344386735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9168 + }, + { + "epoch": 3.6632041550139833, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9169 + }, + { + "epoch": 3.663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9170 + }, + { + "epoch": 3.6640031961646025, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9171 + }, + { + "epoch": 3.6644027167399122, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9172 + }, + { + "epoch": 3.6648022373152216, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9173 + }, + { + "epoch": 3.6652017578905314, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9174 + }, + { + "epoch": 3.665601278465841, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9175 + }, + { + "epoch": 3.6660007990411505, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9176 + }, + { + "epoch": 3.6664003196164603, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9177 + }, + { + "epoch": 3.6667998401917696, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9178 + }, + { + "epoch": 3.6671993607670794, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9179 + }, + { + "epoch": 3.667598881342389, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9180 + }, + { + "epoch": 3.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9181 + }, + { + "epoch": 3.6683979224930083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9182 + }, + { + "epoch": 3.668797443068318, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9183 + }, + { + "epoch": 3.6691969636436275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9184 + }, + { + "epoch": 3.6695964842189372, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9185 + }, + { + "epoch": 3.669996004794247, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9186 + }, + { + "epoch": 3.6703955253695564, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9187 + }, + { + "epoch": 3.670795045944866, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9188 + }, + { + "epoch": 3.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9189 + }, + { + "epoch": 3.6715940870954853, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9190 + }, + { + "epoch": 3.671993607670795, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9191 + }, + { + "epoch": 3.672393128246105, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9192 + }, + { + "epoch": 3.672792648821414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9193 + }, + { + "epoch": 3.673192169396724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9194 + }, + { + "epoch": 3.6735916899720333, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9195 + }, + { + "epoch": 3.673991210547343, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9196 + }, + { + "epoch": 3.674390731122653, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9197 + }, + { + "epoch": 3.6747902516979627, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9198 + }, + { + "epoch": 3.675189772273272, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9199 + }, + { + "epoch": 3.675589292848582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9200 + }, + { + "epoch": 3.675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 9201 + }, + { + "epoch": 3.676388333999201, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9202 + }, + { + "epoch": 3.6767878545745107, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9203 + }, + { + "epoch": 3.67718737514982, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9204 + }, + { + "epoch": 3.67758689572513, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9205 + }, + { + "epoch": 3.677986416300439, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9206 + }, + { + "epoch": 3.678385936875749, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9207 + }, + { + "epoch": 3.678785457451059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9208 + }, + { + "epoch": 3.6791849780263686, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9209 + }, + { + "epoch": 3.679584498601678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9210 + }, + { + "epoch": 3.6799840191769877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9211 + }, + { + "epoch": 3.680383539752297, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9212 + }, + { + "epoch": 3.680783060327607, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9213 + }, + { + "epoch": 3.6811825809029166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9214 + }, + { + "epoch": 3.6815821014782264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9215 + }, + { + "epoch": 3.6819816220535357, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9216 + }, + { + "epoch": 3.6823811426288455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9217 + }, + { + "epoch": 3.682780663204155, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9218 + }, + { + "epoch": 3.6831801837794647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9219 + }, + { + "epoch": 3.6835797043547744, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9220 + }, + { + "epoch": 3.683979224930084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9221 + }, + { + "epoch": 3.6843787455053936, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9222 + }, + { + "epoch": 3.684778266080703, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9223 + }, + { + "epoch": 3.6851777866560127, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9224 + }, + { + "epoch": 3.6855773072313225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9225 + }, + { + "epoch": 3.6859768278066323, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9226 + }, + { + "epoch": 3.6863763483819416, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9227 + }, + { + "epoch": 3.6867758689572514, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9228 + }, + { + "epoch": 3.6871753895325607, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9229 + }, + { + "epoch": 3.6875749101078705, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9230 + }, + { + "epoch": 3.6879744306831803, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9231 + }, + { + "epoch": 3.6883739512584897, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9232 + }, + { + "epoch": 3.6887734718337994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9233 + }, + { + "epoch": 3.6891729924091092, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9234 + }, + { + "epoch": 3.6895725129844186, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9235 + }, + { + "epoch": 3.6899720335597284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9236 + }, + { + "epoch": 3.690371554135038, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9237 + }, + { + "epoch": 3.6907710747103475, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9238 + }, + { + "epoch": 3.6911705952856573, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9239 + }, + { + "epoch": 3.6915701158609666, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9240 + }, + { + "epoch": 3.6919696364362764, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9241 + }, + { + "epoch": 3.692369157011586, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9242 + }, + { + "epoch": 3.692768677586896, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9243 + }, + { + "epoch": 3.6931681981622053, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9244 + }, + { + "epoch": 3.693567718737515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9245 + }, + { + "epoch": 3.6939672393128244, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9246 + }, + { + "epoch": 3.6943667598881342, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9247 + }, + { + "epoch": 3.694766280463444, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9248 + }, + { + "epoch": 3.6951658010387534, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 9249 + }, + { + "epoch": 3.695565321614063, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9250 + }, + { + "epoch": 3.6959648421893725, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 9251 + }, + { + "epoch": 3.6963643627646823, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 9252 + }, + { + "epoch": 3.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9253 + }, + { + "epoch": 3.697163403915302, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9254 + }, + { + "epoch": 3.697562924490611, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9255 + }, + { + "epoch": 3.697962445065921, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9256 + }, + { + "epoch": 3.6983619656412303, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9257 + }, + { + "epoch": 3.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9258 + }, + { + "epoch": 3.69916100679185, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9259 + }, + { + "epoch": 3.6995605273671597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9260 + }, + { + "epoch": 3.699960047942469, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9261 + }, + { + "epoch": 3.700359568517779, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 9262 + }, + { + "epoch": 3.700759089093088, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9263 + }, + { + "epoch": 3.701158609668398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9264 + }, + { + "epoch": 3.7015581302437077, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9265 + }, + { + "epoch": 3.701957650819017, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9266 + }, + { + "epoch": 3.702357171394327, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9267 + }, + { + "epoch": 3.702756691969636, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9268 + }, + { + "epoch": 3.703156212544946, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9269 + }, + { + "epoch": 3.7035557331202558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9270 + }, + { + "epoch": 3.7039552536955656, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9271 + }, + { + "epoch": 3.704354774270875, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9272 + }, + { + "epoch": 3.7047542948461847, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9273 + }, + { + "epoch": 3.705153815421494, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9274 + }, + { + "epoch": 3.705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9275 + }, + { + "epoch": 3.7059528565721136, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9276 + }, + { + "epoch": 3.706352377147423, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9277 + }, + { + "epoch": 3.7067518977227327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9278 + }, + { + "epoch": 3.7071514182980425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9279 + }, + { + "epoch": 3.707550938873352, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9280 + }, + { + "epoch": 3.7079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9281 + }, + { + "epoch": 3.7083499800239714, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9282 + }, + { + "epoch": 3.7087495005992808, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9283 + }, + { + "epoch": 3.7091490211745906, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9284 + }, + { + "epoch": 3.7095485417499, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9285 + }, + { + "epoch": 3.7099480623252097, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9286 + }, + { + "epoch": 3.7103475829005195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9287 + }, + { + "epoch": 3.7107471034758293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9288 + }, + { + "epoch": 3.7111466240511386, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9289 + }, + { + "epoch": 3.7115461446264484, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9290 + }, + { + "epoch": 3.7119456652017577, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9291 + }, + { + "epoch": 3.7123451857770675, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9292 + }, + { + "epoch": 3.7127447063523773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9293 + }, + { + "epoch": 3.7131442269276866, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9294 + }, + { + "epoch": 3.7135437475029964, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9295 + }, + { + "epoch": 3.7139432680783058, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 9296 + }, + { + "epoch": 3.7143427886536156, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9297 + }, + { + "epoch": 3.7147423092289253, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9298 + }, + { + "epoch": 3.715141829804235, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9299 + }, + { + "epoch": 3.7155413503795445, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9300 + }, + { + "epoch": 3.7159408709548543, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 9301 + }, + { + "epoch": 3.7163403915301636, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9302 + }, + { + "epoch": 3.7167399121054734, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 9303 + }, + { + "epoch": 3.717139432680783, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9304 + }, + { + "epoch": 3.717538953256093, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9305 + }, + { + "epoch": 3.7179384738314023, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9306 + }, + { + "epoch": 3.718337994406712, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 9307 + }, + { + "epoch": 3.7187375149820214, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9308 + }, + { + "epoch": 3.719137035557331, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9309 + }, + { + "epoch": 3.719536556132641, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9310 + }, + { + "epoch": 3.7199360767079503, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9311 + }, + { + "epoch": 3.72033559728326, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 9312 + }, + { + "epoch": 3.7207351178585695, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9313 + }, + { + "epoch": 3.7211346384338793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9314 + }, + { + "epoch": 3.721534159009189, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9315 + }, + { + "epoch": 3.721933679584499, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9316 + }, + { + "epoch": 3.722333200159808, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9317 + }, + { + "epoch": 3.722732720735118, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9318 + }, + { + "epoch": 3.7231322413104273, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9319 + }, + { + "epoch": 3.723531761885737, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 9320 + }, + { + "epoch": 3.723931282461047, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9321 + }, + { + "epoch": 3.724330803036356, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9322 + }, + { + "epoch": 3.724730323611666, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9323 + }, + { + "epoch": 3.725129844186976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9324 + }, + { + "epoch": 3.725529364762285, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9325 + }, + { + "epoch": 3.725928885337595, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9326 + }, + { + "epoch": 3.7263284059129047, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9327 + }, + { + "epoch": 3.726727926488214, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9328 + }, + { + "epoch": 3.727127447063524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9329 + }, + { + "epoch": 3.727526967638833, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 9330 + }, + { + "epoch": 3.727926488214143, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9331 + }, + { + "epoch": 3.7283260087894528, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9332 + }, + { + "epoch": 3.7287255293647625, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9333 + }, + { + "epoch": 3.729125049940072, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9334 + }, + { + "epoch": 3.7295245705153817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 9335 + }, + { + "epoch": 3.729924091090691, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9336 + }, + { + "epoch": 3.730323611666001, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9337 + }, + { + "epoch": 3.7307231322413106, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9338 + }, + { + "epoch": 3.73112265281662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9339 + }, + { + "epoch": 3.7315221733919297, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9340 + }, + { + "epoch": 3.731921693967239, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9341 + }, + { + "epoch": 3.732321214542549, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9342 + }, + { + "epoch": 3.7327207351178586, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9343 + }, + { + "epoch": 3.7331202556931684, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9344 + }, + { + "epoch": 3.7335197762684778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9345 + }, + { + "epoch": 3.7339192968437875, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9346 + }, + { + "epoch": 3.734318817419097, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9347 + }, + { + "epoch": 3.7347183379944067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9348 + }, + { + "epoch": 3.7351178585697165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9349 + }, + { + "epoch": 3.7355173791450262, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9350 + }, + { + "epoch": 3.7359168997203356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9351 + }, + { + "epoch": 3.7363164202956454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9352 + }, + { + "epoch": 3.7367159408709547, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9353 + }, + { + "epoch": 3.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 9354 + }, + { + "epoch": 3.7375149820215743, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9355 + }, + { + "epoch": 3.7379145025968836, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9356 + }, + { + "epoch": 3.7383140231721934, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9357 + }, + { + "epoch": 3.7387135437475028, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 9358 + }, + { + "epoch": 3.7391130643228125, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 9359 + }, + { + "epoch": 3.7395125848981223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9360 + }, + { + "epoch": 3.739912105473432, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9361 + }, + { + "epoch": 3.7403116260487415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9362 + }, + { + "epoch": 3.7407111466240512, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 9363 + }, + { + "epoch": 3.7411106671993606, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9364 + }, + { + "epoch": 3.7415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9365 + }, + { + "epoch": 3.74190970834998, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9366 + }, + { + "epoch": 3.7423092289252895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9367 + }, + { + "epoch": 3.7427087495005993, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9368 + }, + { + "epoch": 3.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 9369 + }, + { + "epoch": 3.7435077906512184, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 9370 + }, + { + "epoch": 3.743907311226528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9371 + }, + { + "epoch": 3.744306831801838, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9372 + }, + { + "epoch": 3.7447063523771473, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 9373 + }, + { + "epoch": 3.745105872952457, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9374 + }, + { + "epoch": 3.7455053935277665, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9375 + }, + { + "epoch": 3.7459049141030762, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9376 + }, + { + "epoch": 3.746304434678386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9377 + }, + { + "epoch": 3.746703955253696, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9378 + }, + { + "epoch": 3.747103475829005, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9379 + }, + { + "epoch": 3.747502996404315, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9380 + }, + { + "epoch": 3.7479025169796243, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9381 + }, + { + "epoch": 3.748302037554934, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9382 + }, + { + "epoch": 3.748701558130244, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9383 + }, + { + "epoch": 3.749101078705553, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9384 + }, + { + "epoch": 3.749500599280863, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9385 + }, + { + "epoch": 3.7499001198561723, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9386 + }, + { + "epoch": 3.750299640431482, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9387 + }, + { + "epoch": 3.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9388 + }, + { + "epoch": 3.7510986815821017, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9389 + }, + { + "epoch": 3.751498202157411, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9390 + }, + { + "epoch": 3.751897722732721, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9391 + }, + { + "epoch": 3.75229724330803, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9392 + }, + { + "epoch": 3.75269676388334, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9393 + }, + { + "epoch": 3.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9394 + }, + { + "epoch": 3.7534958050339595, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9395 + }, + { + "epoch": 3.753895325609269, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9396 + }, + { + "epoch": 3.7542948461845786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9397 + }, + { + "epoch": 3.754694366759888, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9398 + }, + { + "epoch": 3.7550938873351978, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9399 + }, + { + "epoch": 3.7554934079105076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9400 + }, + { + "epoch": 3.755892928485817, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9401 + }, + { + "epoch": 3.7562924490611267, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9402 + }, + { + "epoch": 3.756691969636436, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9403 + }, + { + "epoch": 3.757091490211746, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9404 + }, + { + "epoch": 3.7574910107870556, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9405 + }, + { + "epoch": 3.7578905313623654, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9406 + }, + { + "epoch": 3.7582900519376747, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9407 + }, + { + "epoch": 3.7586895725129845, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9408 + }, + { + "epoch": 3.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9409 + }, + { + "epoch": 3.7594886136636037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9410 + }, + { + "epoch": 3.7598881342389134, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9411 + }, + { + "epoch": 3.760287654814223, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9412 + }, + { + "epoch": 3.7606871753895326, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9413 + }, + { + "epoch": 3.7610866959648424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9414 + }, + { + "epoch": 3.7614862165401517, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9415 + }, + { + "epoch": 3.7618857371154615, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9416 + }, + { + "epoch": 3.7622852576907713, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9417 + }, + { + "epoch": 3.7626847782660806, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9418 + }, + { + "epoch": 3.7630842988413904, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9419 + }, + { + "epoch": 3.7634838194166997, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9420 + }, + { + "epoch": 3.7638833399920095, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9421 + }, + { + "epoch": 3.7642828605673193, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9422 + }, + { + "epoch": 3.764682381142629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9423 + }, + { + "epoch": 3.7650819017179384, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9424 + }, + { + "epoch": 3.7654814222932482, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9425 + }, + { + "epoch": 3.7658809428685576, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9426 + }, + { + "epoch": 3.7662804634438674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9427 + }, + { + "epoch": 3.766679984019177, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9428 + }, + { + "epoch": 3.7670795045944865, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9429 + }, + { + "epoch": 3.7674790251697963, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9430 + }, + { + "epoch": 3.7678785457451056, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9431 + }, + { + "epoch": 3.7682780663204154, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9432 + }, + { + "epoch": 3.768677586895725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9433 + }, + { + "epoch": 3.769077107471035, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9434 + }, + { + "epoch": 3.7694766280463443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9435 + }, + { + "epoch": 3.769876148621654, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9436 + }, + { + "epoch": 3.7702756691969634, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9437 + }, + { + "epoch": 3.7706751897722732, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9438 + }, + { + "epoch": 3.771074710347583, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9439 + }, + { + "epoch": 3.771474230922893, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 9440 + }, + { + "epoch": 3.771873751498202, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9441 + }, + { + "epoch": 3.772273272073512, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 9442 + }, + { + "epoch": 3.7726727926488213, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9443 + }, + { + "epoch": 3.773072313224131, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9444 + }, + { + "epoch": 3.773471833799441, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9445 + }, + { + "epoch": 3.77387135437475, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9446 + }, + { + "epoch": 3.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9447 + }, + { + "epoch": 3.7746703955253693, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9448 + }, + { + "epoch": 3.775069916100679, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 9449 + }, + { + "epoch": 3.775469436675989, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9450 + }, + { + "epoch": 3.7758689572512987, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9451 + }, + { + "epoch": 3.776268477826608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9452 + }, + { + "epoch": 3.776667998401918, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9453 + }, + { + "epoch": 3.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9454 + }, + { + "epoch": 3.777467039552537, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9455 + }, + { + "epoch": 3.7778665601278467, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9456 + }, + { + "epoch": 3.778266080703156, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9457 + }, + { + "epoch": 3.778665601278466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9458 + }, + { + "epoch": 3.7790651218537756, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9459 + }, + { + "epoch": 3.779464642429085, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9460 + }, + { + "epoch": 3.7798641630043948, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9461 + }, + { + "epoch": 3.7802636835797045, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9462 + }, + { + "epoch": 3.780663204155014, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9463 + }, + { + "epoch": 3.7810627247303237, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9464 + }, + { + "epoch": 3.781462245305633, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9465 + }, + { + "epoch": 3.781861765880943, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9466 + }, + { + "epoch": 3.7822612864562526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9467 + }, + { + "epoch": 3.7826608070315624, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9468 + }, + { + "epoch": 3.7830603276068717, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9469 + }, + { + "epoch": 3.7834598481821815, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9470 + }, + { + "epoch": 3.783859368757491, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 9471 + }, + { + "epoch": 3.7842588893328006, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9472 + }, + { + "epoch": 3.7846584099081104, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9473 + }, + { + "epoch": 3.7850579304834198, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9474 + }, + { + "epoch": 3.7854574510587295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9475 + }, + { + "epoch": 3.785856971634039, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9476 + }, + { + "epoch": 3.7862564922093487, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9477 + }, + { + "epoch": 3.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9478 + }, + { + "epoch": 3.7870555333599683, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9479 + }, + { + "epoch": 3.7874550539352776, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9480 + }, + { + "epoch": 3.7878545745105874, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9481 + }, + { + "epoch": 3.7882540950858967, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 9482 + }, + { + "epoch": 3.7886536156612065, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9483 + }, + { + "epoch": 3.7890531362365163, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9484 + }, + { + "epoch": 3.789452656811826, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9485 + }, + { + "epoch": 3.7898521773871354, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 9486 + }, + { + "epoch": 3.790251697962445, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9487 + }, + { + "epoch": 3.7906512185377546, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9488 + }, + { + "epoch": 3.7910507391130643, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9489 + }, + { + "epoch": 3.791450259688374, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 9490 + }, + { + "epoch": 3.7918497802636835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9491 + }, + { + "epoch": 3.7922493008389933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9492 + }, + { + "epoch": 3.7926488214143026, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9493 + }, + { + "epoch": 3.7930483419896124, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9494 + }, + { + "epoch": 3.793447862564922, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9495 + }, + { + "epoch": 3.793847383140232, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9496 + }, + { + "epoch": 3.7942469037155413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9497 + }, + { + "epoch": 3.794646424290851, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9498 + }, + { + "epoch": 3.7950459448661604, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9499 + }, + { + "epoch": 3.79544546544147, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9500 + }, + { + "epoch": 3.79584498601678, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9501 + }, + { + "epoch": 3.7962445065920893, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9502 + }, + { + "epoch": 3.796644027167399, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9503 + }, + { + "epoch": 3.797043547742709, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9504 + }, + { + "epoch": 3.7974430683180183, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9505 + }, + { + "epoch": 3.797842588893328, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9506 + }, + { + "epoch": 3.798242109468638, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9507 + }, + { + "epoch": 3.798641630043947, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9508 + }, + { + "epoch": 3.799041150619257, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 9509 + }, + { + "epoch": 3.7994406711945663, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 9510 + }, + { + "epoch": 3.799840191769876, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9511 + }, + { + "epoch": 3.800239712345186, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9512 + }, + { + "epoch": 3.8006392329204957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9513 + }, + { + "epoch": 3.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9514 + }, + { + "epoch": 3.801438274071115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9515 + }, + { + "epoch": 3.801837794646424, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9516 + }, + { + "epoch": 3.802237315221734, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9517 + }, + { + "epoch": 3.8026368357970437, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9518 + }, + { + "epoch": 3.803036356372353, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9519 + }, + { + "epoch": 3.803435876947663, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9520 + }, + { + "epoch": 3.803835397522972, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9521 + }, + { + "epoch": 3.804234918098282, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9522 + }, + { + "epoch": 3.8046344386735917, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9523 + }, + { + "epoch": 3.8050339592489015, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9524 + }, + { + "epoch": 3.805433479824211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9525 + }, + { + "epoch": 3.8058330003995207, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9526 + }, + { + "epoch": 3.80623252097483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9527 + }, + { + "epoch": 3.80663204155014, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9528 + }, + { + "epoch": 3.8070315621254496, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9529 + }, + { + "epoch": 3.8074310827007594, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9530 + }, + { + "epoch": 3.8078306032760687, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9531 + }, + { + "epoch": 3.8082301238513785, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9532 + }, + { + "epoch": 3.808629644426688, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9533 + }, + { + "epoch": 3.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9534 + }, + { + "epoch": 3.8094286855773074, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9535 + }, + { + "epoch": 3.8098282061526167, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9536 + }, + { + "epoch": 3.8102277267279265, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9537 + }, + { + "epoch": 3.810627247303236, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9538 + }, + { + "epoch": 3.8110267678785457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9539 + }, + { + "epoch": 3.8114262884538554, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9540 + }, + { + "epoch": 3.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9541 + }, + { + "epoch": 3.8122253296044746, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9542 + }, + { + "epoch": 3.8126248501797844, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 9543 + }, + { + "epoch": 3.8130243707550937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9544 + }, + { + "epoch": 3.8134238913304035, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9545 + }, + { + "epoch": 3.8138234119057133, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9546 + }, + { + "epoch": 3.8142229324810226, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9547 + }, + { + "epoch": 3.8146224530563324, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9548 + }, + { + "epoch": 3.815021973631642, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9549 + }, + { + "epoch": 3.8154214942069515, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9550 + }, + { + "epoch": 3.8158210147822613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9551 + }, + { + "epoch": 3.816220535357571, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9552 + }, + { + "epoch": 3.8166200559328805, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9553 + }, + { + "epoch": 3.8170195765081902, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 9554 + }, + { + "epoch": 3.8174190970834996, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9555 + }, + { + "epoch": 3.8178186176588094, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9556 + }, + { + "epoch": 3.818218138234119, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9557 + }, + { + "epoch": 3.818617658809429, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9558 + }, + { + "epoch": 3.8190171793847383, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9559 + }, + { + "epoch": 3.819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9560 + }, + { + "epoch": 3.8198162205353574, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9561 + }, + { + "epoch": 3.820215741110667, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9562 + }, + { + "epoch": 3.820615261685977, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9563 + }, + { + "epoch": 3.8210147822612863, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9564 + }, + { + "epoch": 3.821414302836596, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9565 + }, + { + "epoch": 3.8218138234119055, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9566 + }, + { + "epoch": 3.8222133439872152, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9567 + }, + { + "epoch": 3.822612864562525, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9568 + }, + { + "epoch": 3.823012385137835, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 9569 + }, + { + "epoch": 3.823411905713144, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9570 + }, + { + "epoch": 3.823811426288454, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9571 + }, + { + "epoch": 3.8242109468637633, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9572 + }, + { + "epoch": 3.824610467439073, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9573 + }, + { + "epoch": 3.825009988014383, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9574 + }, + { + "epoch": 3.8254095085896926, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9575 + }, + { + "epoch": 3.825809029165002, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9576 + }, + { + "epoch": 3.8262085497403118, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 9577 + }, + { + "epoch": 3.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9578 + }, + { + "epoch": 3.827007590890931, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9579 + }, + { + "epoch": 3.8274071114662407, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9580 + }, + { + "epoch": 3.82780663204155, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9581 + }, + { + "epoch": 3.82820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9582 + }, + { + "epoch": 3.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9583 + }, + { + "epoch": 3.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 9584 + }, + { + "epoch": 3.8294047143427887, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9585 + }, + { + "epoch": 3.8298042349180985, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9586 + }, + { + "epoch": 3.830203755493408, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9587 + }, + { + "epoch": 3.8306032760687176, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9588 + }, + { + "epoch": 3.831002796644027, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9589 + }, + { + "epoch": 3.8314023172193368, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9590 + }, + { + "epoch": 3.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9591 + }, + { + "epoch": 3.832201358369956, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9592 + }, + { + "epoch": 3.8326008789452657, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9593 + }, + { + "epoch": 3.8330003995205755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9594 + }, + { + "epoch": 3.833399920095885, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9595 + }, + { + "epoch": 3.8337994406711946, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9596 + }, + { + "epoch": 3.8341989612465044, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9597 + }, + { + "epoch": 3.8345984818218137, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 9598 + }, + { + "epoch": 3.8349980023971235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9599 + }, + { + "epoch": 3.835397522972433, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9600 + }, + { + "epoch": 3.8357970435477426, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9601 + }, + { + "epoch": 3.8361965641230524, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9602 + }, + { + "epoch": 3.836596084698362, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9603 + }, + { + "epoch": 3.8369956052736716, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9604 + }, + { + "epoch": 3.8373951258489813, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9605 + }, + { + "epoch": 3.8377946464242907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9606 + }, + { + "epoch": 3.8381941669996005, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9607 + }, + { + "epoch": 3.8385936875749103, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9608 + }, + { + "epoch": 3.8389932081502196, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9609 + }, + { + "epoch": 3.8393927287255294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9610 + }, + { + "epoch": 3.8397922493008387, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9611 + }, + { + "epoch": 3.8401917698761485, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9612 + }, + { + "epoch": 3.8405912904514583, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9613 + }, + { + "epoch": 3.840990811026768, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9614 + }, + { + "epoch": 3.8413903316020774, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9615 + }, + { + "epoch": 3.841789852177387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9616 + }, + { + "epoch": 3.8421893727526966, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9617 + }, + { + "epoch": 3.8425888933280063, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9618 + }, + { + "epoch": 3.842988413903316, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 9619 + }, + { + "epoch": 3.843387934478626, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9620 + }, + { + "epoch": 3.8437874550539353, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9621 + }, + { + "epoch": 3.844186975629245, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9622 + }, + { + "epoch": 3.8445864962045544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9623 + }, + { + "epoch": 3.844986016779864, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 9624 + }, + { + "epoch": 3.845385537355174, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9625 + }, + { + "epoch": 3.8457850579304833, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 9626 + }, + { + "epoch": 3.846184578505793, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9627 + }, + { + "epoch": 3.8465840990811024, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9628 + }, + { + "epoch": 3.8469836196564122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9629 + }, + { + "epoch": 3.847383140231722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9630 + }, + { + "epoch": 3.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9631 + }, + { + "epoch": 3.848182181382341, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9632 + }, + { + "epoch": 3.848581701957651, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9633 + }, + { + "epoch": 3.8489812225329603, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 9634 + }, + { + "epoch": 3.84938074310827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9635 + }, + { + "epoch": 3.84978026368358, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 9636 + }, + { + "epoch": 3.850179784258889, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9637 + }, + { + "epoch": 3.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9638 + }, + { + "epoch": 3.8509788254095088, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9639 + }, + { + "epoch": 3.851378345984818, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9640 + }, + { + "epoch": 3.851777866560128, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9641 + }, + { + "epoch": 3.8521773871354377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 9642 + }, + { + "epoch": 3.852576907710747, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9643 + }, + { + "epoch": 3.852976428286057, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9644 + }, + { + "epoch": 3.853375948861366, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9645 + }, + { + "epoch": 3.853775469436676, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9646 + }, + { + "epoch": 3.8541749900119857, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9647 + }, + { + "epoch": 3.8545745105872955, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9648 + }, + { + "epoch": 3.854974031162605, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9649 + }, + { + "epoch": 3.8553735517379146, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9650 + }, + { + "epoch": 3.855773072313224, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9651 + }, + { + "epoch": 3.8561725928885338, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9652 + }, + { + "epoch": 3.8565721134638435, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9653 + }, + { + "epoch": 3.856971634039153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9654 + }, + { + "epoch": 3.8573711546144627, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9655 + }, + { + "epoch": 3.857770675189772, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9656 + }, + { + "epoch": 3.858170195765082, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9657 + }, + { + "epoch": 3.8585697163403916, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 9658 + }, + { + "epoch": 3.8589692369157014, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9659 + }, + { + "epoch": 3.8593687574910107, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9660 + }, + { + "epoch": 3.8597682780663205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9661 + }, + { + "epoch": 3.86016779864163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9662 + }, + { + "epoch": 3.8605673192169396, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9663 + }, + { + "epoch": 3.8609668397922494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9664 + }, + { + "epoch": 3.861366360367559, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9665 + }, + { + "epoch": 3.8617658809428685, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9666 + }, + { + "epoch": 3.8621654015181783, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9667 + }, + { + "epoch": 3.8625649220934877, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9668 + }, + { + "epoch": 3.8629644426687975, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9669 + }, + { + "epoch": 3.8633639632441072, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9670 + }, + { + "epoch": 3.8637634838194166, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9671 + }, + { + "epoch": 3.8641630043947264, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9672 + }, + { + "epoch": 3.8645625249700357, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9673 + }, + { + "epoch": 3.8649620455453455, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9674 + }, + { + "epoch": 3.8653615661206553, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9675 + }, + { + "epoch": 3.865761086695965, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 9676 + }, + { + "epoch": 3.8661606072712744, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 9677 + }, + { + "epoch": 3.866560127846584, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9678 + }, + { + "epoch": 3.8669596484218935, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9679 + }, + { + "epoch": 3.8673591689972033, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9680 + }, + { + "epoch": 3.867758689572513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 9681 + }, + { + "epoch": 3.8681582101478225, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9682 + }, + { + "epoch": 3.8685577307231322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9683 + }, + { + "epoch": 3.868957251298442, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9684 + }, + { + "epoch": 3.8693567718737514, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9685 + }, + { + "epoch": 3.869756292449061, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9686 + }, + { + "epoch": 3.870155813024371, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9687 + }, + { + "epoch": 3.8705553335996803, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 9688 + }, + { + "epoch": 3.87095485417499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9689 + }, + { + "epoch": 3.8713543747502994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9690 + }, + { + "epoch": 3.871753895325609, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9691 + }, + { + "epoch": 3.872153415900919, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9692 + }, + { + "epoch": 3.872552936476229, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9693 + }, + { + "epoch": 3.872952457051538, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9694 + }, + { + "epoch": 3.873351977626848, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9695 + }, + { + "epoch": 3.8737514982021573, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9696 + }, + { + "epoch": 3.874151018777467, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9697 + }, + { + "epoch": 3.874550539352777, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 9698 + }, + { + "epoch": 3.874950059928086, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9699 + }, + { + "epoch": 3.875349580503396, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9700 + }, + { + "epoch": 3.8757491010787053, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9701 + }, + { + "epoch": 3.876148621654015, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9702 + }, + { + "epoch": 3.876548142229325, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9703 + }, + { + "epoch": 3.8769476628046347, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9704 + }, + { + "epoch": 3.877347183379944, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9705 + }, + { + "epoch": 3.877746703955254, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9706 + }, + { + "epoch": 3.878146224530563, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9707 + }, + { + "epoch": 3.878545745105873, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9708 + }, + { + "epoch": 3.8789452656811827, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9709 + }, + { + "epoch": 3.8793447862564925, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9710 + }, + { + "epoch": 3.879744306831802, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9711 + }, + { + "epoch": 3.8801438274071116, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9712 + }, + { + "epoch": 3.880543347982421, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9713 + }, + { + "epoch": 3.8809428685577307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9714 + }, + { + "epoch": 3.8813423891330405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9715 + }, + { + "epoch": 3.88174190970835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9716 + }, + { + "epoch": 3.8821414302836597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9717 + }, + { + "epoch": 3.882540950858969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9718 + }, + { + "epoch": 3.882940471434279, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9719 + }, + { + "epoch": 3.8833399920095886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9720 + }, + { + "epoch": 3.8837395125848984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9721 + }, + { + "epoch": 3.8841390331602077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9722 + }, + { + "epoch": 3.8845385537355175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9723 + }, + { + "epoch": 3.884938074310827, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9724 + }, + { + "epoch": 3.8853375948861366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9725 + }, + { + "epoch": 3.8857371154614464, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9726 + }, + { + "epoch": 3.8861366360367557, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9727 + }, + { + "epoch": 3.8865361566120655, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9728 + }, + { + "epoch": 3.8869356771873753, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9729 + }, + { + "epoch": 3.8873351977626847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9730 + }, + { + "epoch": 3.8877347183379944, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9731 + }, + { + "epoch": 3.8881342389133042, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9732 + }, + { + "epoch": 3.8885337594886136, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9733 + }, + { + "epoch": 3.8889332800639234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9734 + }, + { + "epoch": 3.8893328006392327, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9735 + }, + { + "epoch": 3.8897323212145425, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9736 + }, + { + "epoch": 3.8901318417898523, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9737 + }, + { + "epoch": 3.890531362365162, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9738 + }, + { + "epoch": 3.8909308829404714, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9739 + }, + { + "epoch": 3.891330403515781, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9740 + }, + { + "epoch": 3.8917299240910905, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9741 + }, + { + "epoch": 3.8921294446664003, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9742 + }, + { + "epoch": 3.89252896524171, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9743 + }, + { + "epoch": 3.8929284858170194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9744 + }, + { + "epoch": 3.8933280063923292, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 9745 + }, + { + "epoch": 3.8937275269676386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9746 + }, + { + "epoch": 3.8941270475429484, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9747 + }, + { + "epoch": 3.894526568118258, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9748 + }, + { + "epoch": 3.894926088693568, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9749 + }, + { + "epoch": 3.8953256092688773, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9750 + }, + { + "epoch": 3.895725129844187, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9751 + }, + { + "epoch": 3.8961246504194964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9752 + }, + { + "epoch": 3.896524170994806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9753 + }, + { + "epoch": 3.896923691570116, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9754 + }, + { + "epoch": 3.8973232121454258, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9755 + }, + { + "epoch": 3.897722732720735, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9756 + }, + { + "epoch": 3.898122253296045, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9757 + }, + { + "epoch": 3.8985217738713542, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9758 + }, + { + "epoch": 3.898921294446664, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9759 + }, + { + "epoch": 3.899320815021974, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9760 + }, + { + "epoch": 3.899720335597283, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 9761 + }, + { + "epoch": 3.900119856172593, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9762 + }, + { + "epoch": 3.9005193767479023, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9763 + }, + { + "epoch": 3.900918897323212, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9764 + }, + { + "epoch": 3.901318417898522, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9765 + }, + { + "epoch": 3.9017179384738316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9766 + }, + { + "epoch": 3.902117459049141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9767 + }, + { + "epoch": 3.9025169796244508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9768 + }, + { + "epoch": 3.90291650019976, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9769 + }, + { + "epoch": 3.90331602077507, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9770 + }, + { + "epoch": 3.9037155413503797, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9771 + }, + { + "epoch": 3.904115061925689, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9772 + }, + { + "epoch": 3.904514582500999, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9773 + }, + { + "epoch": 3.9049141030763086, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9774 + }, + { + "epoch": 3.905313623651618, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9775 + }, + { + "epoch": 3.9057131442269277, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9776 + }, + { + "epoch": 3.9061126648022375, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9777 + }, + { + "epoch": 3.906512185377547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9778 + }, + { + "epoch": 3.9069117059528566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9779 + }, + { + "epoch": 3.907311226528166, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9780 + }, + { + "epoch": 3.9077107471034758, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9781 + }, + { + "epoch": 3.9081102676787856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9782 + }, + { + "epoch": 3.9085097882540953, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9783 + }, + { + "epoch": 3.9089093088294047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9784 + }, + { + "epoch": 3.9093088294047145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9785 + }, + { + "epoch": 3.909708349980024, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9786 + }, + { + "epoch": 3.9101078705553336, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9787 + }, + { + "epoch": 3.9105073911306434, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9788 + }, + { + "epoch": 3.9109069117059527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9789 + }, + { + "epoch": 3.9113064322812625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9790 + }, + { + "epoch": 3.911705952856572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9791 + }, + { + "epoch": 3.9121054734318816, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9792 + }, + { + "epoch": 3.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9793 + }, + { + "epoch": 3.912904514582501, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9794 + }, + { + "epoch": 3.9133040351578106, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9795 + }, + { + "epoch": 3.9137035557331203, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9796 + }, + { + "epoch": 3.9141030763084297, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 9797 + }, + { + "epoch": 3.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9798 + }, + { + "epoch": 3.9149021174590493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9799 + }, + { + "epoch": 3.915301638034359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9800 + }, + { + "epoch": 3.9157011586096684, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9801 + }, + { + "epoch": 3.916100679184978, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9802 + }, + { + "epoch": 3.9165001997602875, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9803 + }, + { + "epoch": 3.9168997203355973, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9804 + }, + { + "epoch": 3.917299240910907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9805 + }, + { + "epoch": 3.9176987614862164, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9806 + }, + { + "epoch": 3.918098282061526, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9807 + }, + { + "epoch": 3.9184978026368356, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 9808 + }, + { + "epoch": 3.9188973232121453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 9809 + }, + { + "epoch": 3.919296843787455, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9810 + }, + { + "epoch": 3.919696364362765, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9811 + }, + { + "epoch": 3.9200958849380743, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9812 + }, + { + "epoch": 3.920495405513384, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9813 + }, + { + "epoch": 3.9208949260886934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9814 + }, + { + "epoch": 3.921294446664003, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9815 + }, + { + "epoch": 3.921693967239313, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9816 + }, + { + "epoch": 3.9220934878146223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9817 + }, + { + "epoch": 3.922493008389932, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9818 + }, + { + "epoch": 3.922892528965242, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9819 + }, + { + "epoch": 3.923292049540551, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9820 + }, + { + "epoch": 3.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9821 + }, + { + "epoch": 3.924091090691171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9822 + }, + { + "epoch": 3.92449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9823 + }, + { + "epoch": 3.92489013184179, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9824 + }, + { + "epoch": 3.9252896524170993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9825 + }, + { + "epoch": 3.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9826 + }, + { + "epoch": 3.926088693567719, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9827 + }, + { + "epoch": 3.9264882141430286, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9828 + }, + { + "epoch": 3.926887734718338, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9829 + }, + { + "epoch": 3.9272872552936477, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9830 + }, + { + "epoch": 3.927686775868957, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9831 + }, + { + "epoch": 3.928086296444267, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 9832 + }, + { + "epoch": 3.9284858170195767, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9833 + }, + { + "epoch": 3.928885337594886, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9834 + }, + { + "epoch": 3.929284858170196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9835 + }, + { + "epoch": 3.929684378745505, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9836 + }, + { + "epoch": 3.930083899320815, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9837 + }, + { + "epoch": 3.9304834198961247, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9838 + }, + { + "epoch": 3.9308829404714345, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9839 + }, + { + "epoch": 3.931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9840 + }, + { + "epoch": 3.9316819816220536, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9841 + }, + { + "epoch": 3.932081502197363, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9842 + }, + { + "epoch": 3.9324810227726728, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9843 + }, + { + "epoch": 3.9328805433479825, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9844 + }, + { + "epoch": 3.9332800639232923, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9845 + }, + { + "epoch": 3.9336795844986017, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9846 + }, + { + "epoch": 3.9340791050739115, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9847 + }, + { + "epoch": 3.934478625649221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9848 + }, + { + "epoch": 3.9348781462245306, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9849 + }, + { + "epoch": 3.9352776667998404, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9850 + }, + { + "epoch": 3.9356771873751497, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9851 + }, + { + "epoch": 3.9360767079504595, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9852 + }, + { + "epoch": 3.936476228525769, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9853 + }, + { + "epoch": 3.9368757491010786, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9854 + }, + { + "epoch": 3.9372752696763884, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9855 + }, + { + "epoch": 3.937674790251698, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9856 + }, + { + "epoch": 3.9380743108270075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9857 + }, + { + "epoch": 3.9384738314023173, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9858 + }, + { + "epoch": 3.9388733519776267, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9859 + }, + { + "epoch": 3.9392728725529365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9860 + }, + { + "epoch": 3.9396723931282462, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9861 + }, + { + "epoch": 3.9400719137035556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9862 + }, + { + "epoch": 3.9404714342788654, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9863 + }, + { + "epoch": 3.9408709548541747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9864 + }, + { + "epoch": 3.9412704754294845, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9865 + }, + { + "epoch": 3.9416699960047943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9866 + }, + { + "epoch": 3.942069516580104, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9867 + }, + { + "epoch": 3.9424690371554134, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9868 + }, + { + "epoch": 3.942868557730723, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9869 + }, + { + "epoch": 3.9432680783060325, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9870 + }, + { + "epoch": 3.9436675988813423, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9871 + }, + { + "epoch": 3.944067119456652, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9872 + }, + { + "epoch": 3.944466640031962, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9873 + }, + { + "epoch": 3.9448661606072712, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9874 + }, + { + "epoch": 3.945265681182581, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9875 + }, + { + "epoch": 3.9456652017578904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9876 + }, + { + "epoch": 3.9460647223332, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 9877 + }, + { + "epoch": 3.94646424290851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9878 + }, + { + "epoch": 3.9468637634838193, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9879 + }, + { + "epoch": 3.947263284059129, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9880 + }, + { + "epoch": 3.9476628046344384, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9881 + }, + { + "epoch": 3.948062325209748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9882 + }, + { + "epoch": 3.948461845785058, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9883 + }, + { + "epoch": 3.9488613663603678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9884 + }, + { + "epoch": 3.949260886935677, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9885 + }, + { + "epoch": 3.949660407510987, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9886 + }, + { + "epoch": 3.9500599280862962, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9887 + }, + { + "epoch": 3.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9888 + }, + { + "epoch": 3.950858969236916, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9889 + }, + { + "epoch": 3.9512584898122256, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9890 + }, + { + "epoch": 3.951658010387535, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9891 + }, + { + "epoch": 3.9520575309628447, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9892 + }, + { + "epoch": 3.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9893 + }, + { + "epoch": 3.952856572113464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9894 + }, + { + "epoch": 3.9532560926887736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9895 + }, + { + "epoch": 3.953655613264083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9896 + }, + { + "epoch": 3.9540551338393928, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9897 + }, + { + "epoch": 3.954454654414702, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9898 + }, + { + "epoch": 3.954854174990012, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9899 + }, + { + "epoch": 3.9552536955653217, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9900 + }, + { + "epoch": 3.9556532161406315, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9901 + }, + { + "epoch": 3.956052736715941, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9902 + }, + { + "epoch": 3.9564522572912506, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9903 + }, + { + "epoch": 3.95685177786656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9904 + }, + { + "epoch": 3.9572512984418697, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9905 + }, + { + "epoch": 3.9576508190171795, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9906 + }, + { + "epoch": 3.958050339592489, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9907 + }, + { + "epoch": 3.9584498601677987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9908 + }, + { + "epoch": 3.958849380743108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9909 + }, + { + "epoch": 3.959248901318418, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9910 + }, + { + "epoch": 3.9596484218937276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 9911 + }, + { + "epoch": 3.9600479424690374, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9912 + }, + { + "epoch": 3.9604474630443467, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9913 + }, + { + "epoch": 3.9608469836196565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9914 + }, + { + "epoch": 3.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9915 + }, + { + "epoch": 3.9616460247702756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9916 + }, + { + "epoch": 3.9620455453455854, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9917 + }, + { + "epoch": 3.962445065920895, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9918 + }, + { + "epoch": 3.9628445864962045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9919 + }, + { + "epoch": 3.9632441070715143, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9920 + }, + { + "epoch": 3.9636436276468237, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9921 + }, + { + "epoch": 3.9640431482221334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9922 + }, + { + "epoch": 3.9644426687974432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9923 + }, + { + "epoch": 3.9648421893727526, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9924 + }, + { + "epoch": 3.9652417099480624, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9925 + }, + { + "epoch": 3.9656412305233717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 9926 + }, + { + "epoch": 3.9660407510986815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9927 + }, + { + "epoch": 3.9664402716739913, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9928 + }, + { + "epoch": 3.966839792249301, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9929 + }, + { + "epoch": 3.9672393128246104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9930 + }, + { + "epoch": 3.96763883339992, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9931 + }, + { + "epoch": 3.9680383539752295, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9932 + }, + { + "epoch": 3.9684378745505393, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9933 + }, + { + "epoch": 3.968837395125849, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 9934 + }, + { + "epoch": 3.969236915701159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 9935 + }, + { + "epoch": 3.9696364362764682, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9936 + }, + { + "epoch": 3.970035956851778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 9937 + }, + { + "epoch": 3.9704354774270874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9938 + }, + { + "epoch": 3.970834998002397, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9939 + }, + { + "epoch": 3.971234518577707, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9940 + }, + { + "epoch": 3.9716340391530163, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9941 + }, + { + "epoch": 3.972033559728326, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9942 + }, + { + "epoch": 3.9724330803036354, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9943 + }, + { + "epoch": 3.972832600878945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 9944 + }, + { + "epoch": 3.973232121454255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9945 + }, + { + "epoch": 3.9736316420295648, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9946 + }, + { + "epoch": 3.974031162604874, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9947 + }, + { + "epoch": 3.974430683180184, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9948 + }, + { + "epoch": 3.9748302037554932, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9949 + }, + { + "epoch": 3.975229724330803, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9950 + }, + { + "epoch": 3.975629244906113, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9951 + }, + { + "epoch": 3.976028765481422, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9952 + }, + { + "epoch": 3.976428286056732, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9953 + }, + { + "epoch": 3.9768278066320413, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9954 + }, + { + "epoch": 3.977227327207351, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9955 + }, + { + "epoch": 3.977626847782661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 9956 + }, + { + "epoch": 3.9780263683579706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9957 + }, + { + "epoch": 3.97842588893328, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9958 + }, + { + "epoch": 3.9788254095085898, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9959 + }, + { + "epoch": 3.979224930083899, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9960 + }, + { + "epoch": 3.979624450659209, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9961 + }, + { + "epoch": 3.9800239712345187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9962 + }, + { + "epoch": 3.9804234918098285, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9963 + }, + { + "epoch": 3.980823012385138, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 9964 + }, + { + "epoch": 3.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9965 + }, + { + "epoch": 3.981622053535757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9966 + }, + { + "epoch": 3.9820215741110667, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9967 + }, + { + "epoch": 3.9824210946863765, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9968 + }, + { + "epoch": 3.982820615261686, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9969 + }, + { + "epoch": 3.9832201358369956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9970 + }, + { + "epoch": 3.983619656412305, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9971 + }, + { + "epoch": 3.9840191769876148, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9972 + }, + { + "epoch": 3.9844186975629245, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9973 + }, + { + "epoch": 3.9848182181382343, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9974 + }, + { + "epoch": 3.9852177387135437, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9975 + }, + { + "epoch": 3.9856172592888535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9976 + }, + { + "epoch": 3.986016779864163, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9977 + }, + { + "epoch": 3.9864163004394726, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9978 + }, + { + "epoch": 3.9868158210147824, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9979 + }, + { + "epoch": 3.987215341590092, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9980 + }, + { + "epoch": 3.9876148621654015, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9981 + }, + { + "epoch": 3.9880143827407113, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9982 + }, + { + "epoch": 3.9884139033160206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9983 + }, + { + "epoch": 3.9888134238913304, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9984 + }, + { + "epoch": 3.98921294446664, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9985 + }, + { + "epoch": 3.9896124650419496, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9986 + }, + { + "epoch": 3.9900119856172593, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 9987 + }, + { + "epoch": 3.9904115061925687, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9988 + }, + { + "epoch": 3.9908110267678785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9989 + }, + { + "epoch": 3.9912105473431883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9990 + }, + { + "epoch": 3.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9991 + }, + { + "epoch": 3.9920095884938074, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9992 + }, + { + "epoch": 3.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9993 + }, + { + "epoch": 3.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9994 + }, + { + "epoch": 3.9932081502197363, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9995 + }, + { + "epoch": 3.993607670795046, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9996 + }, + { + "epoch": 3.9940071913703554, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9997 + }, + { + "epoch": 3.994406711945665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9998 + }, + { + "epoch": 3.9948062325209746, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9999 + }, + { + "epoch": 3.9952057530962843, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10000 + }, + { + "epoch": 3.995605273671594, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10001 + }, + { + "epoch": 3.996004794246904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10002 + }, + { + "epoch": 3.9964043148222133, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10003 + }, + { + "epoch": 3.996803835397523, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10004 + }, + { + "epoch": 3.9972033559728324, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10005 + }, + { + "epoch": 3.997602876548142, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10006 + }, + { + "epoch": 3.998002397123452, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10007 + }, + { + "epoch": 3.9984019176987617, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10008 + }, + { + "epoch": 3.998801438274071, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10009 + }, + { + "epoch": 3.999200958849381, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 10010 + }, + { + "epoch": 3.99960047942469, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 10011 + }, + { + "epoch": 4.0, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10012 + }, + { + "epoch": 4.00039952057531, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10013 + }, + { + "epoch": 4.00079904115062, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10014 + }, + { + "epoch": 4.0011985617259285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10015 + }, + { + "epoch": 4.001598082301238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10016 + }, + { + "epoch": 4.001997602876548, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10017 + }, + { + "epoch": 4.002397123451858, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10018 + }, + { + "epoch": 4.002796644027168, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10019 + }, + { + "epoch": 4.003196164602477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10020 + }, + { + "epoch": 4.003595685177786, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10021 + }, + { + "epoch": 4.003995205753096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10022 + }, + { + "epoch": 4.004394726328406, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10023 + }, + { + "epoch": 4.004794246903716, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10024 + }, + { + "epoch": 4.0051937674790254, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10025 + }, + { + "epoch": 4.005593288054335, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10026 + }, + { + "epoch": 4.005992808629644, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10027 + }, + { + "epoch": 4.006392329204954, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10028 + }, + { + "epoch": 4.006791849780264, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10029 + }, + { + "epoch": 4.0071913703555735, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10030 + }, + { + "epoch": 4.007590890930883, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10031 + }, + { + "epoch": 4.007990411506192, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10032 + }, + { + "epoch": 4.008389932081502, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10033 + }, + { + "epoch": 4.008789452656812, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10034 + }, + { + "epoch": 4.0091889732321215, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10035 + }, + { + "epoch": 4.009588493807431, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 10036 + }, + { + "epoch": 4.009988014382741, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10037 + }, + { + "epoch": 4.01038753495805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 10038 + }, + { + "epoch": 4.01078705553336, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10039 + }, + { + "epoch": 4.01118657610867, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10040 + }, + { + "epoch": 4.011586096683979, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10041 + }, + { + "epoch": 4.011985617259289, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10042 + }, + { + "epoch": 4.012385137834598, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10043 + }, + { + "epoch": 4.012784658409908, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10044 + }, + { + "epoch": 4.013184178985218, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10045 + }, + { + "epoch": 4.013583699560527, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10046 + }, + { + "epoch": 4.013983220135837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10047 + }, + { + "epoch": 4.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10048 + }, + { + "epoch": 4.014782261286456, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 10049 + }, + { + "epoch": 4.015181781861766, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10050 + }, + { + "epoch": 4.0155813024370755, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10051 + }, + { + "epoch": 4.015980823012385, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10052 + }, + { + "epoch": 4.016380343587695, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10053 + }, + { + "epoch": 4.016779864163005, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10054 + }, + { + "epoch": 4.017179384738314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10055 + }, + { + "epoch": 4.0175789053136235, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10056 + }, + { + "epoch": 4.017978425888933, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10057 + }, + { + "epoch": 4.018377946464243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10058 + }, + { + "epoch": 4.018777467039553, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10059 + }, + { + "epoch": 4.019176987614862, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 10060 + }, + { + "epoch": 4.0195765081901715, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10061 + }, + { + "epoch": 4.019976028765481, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10062 + }, + { + "epoch": 4.020375549340791, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10063 + }, + { + "epoch": 4.020775069916101, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10064 + }, + { + "epoch": 4.021174590491411, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10065 + }, + { + "epoch": 4.02157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10066 + }, + { + "epoch": 4.021973631642029, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10067 + }, + { + "epoch": 4.022373152217339, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10068 + }, + { + "epoch": 4.022772672792649, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10069 + }, + { + "epoch": 4.023172193367959, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10070 + }, + { + "epoch": 4.0235717139432685, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10071 + }, + { + "epoch": 4.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10072 + }, + { + "epoch": 4.024370755093887, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10073 + }, + { + "epoch": 4.024770275669197, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10074 + }, + { + "epoch": 4.025169796244507, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10075 + }, + { + "epoch": 4.025569316819817, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10076 + }, + { + "epoch": 4.0259688373951255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10077 + }, + { + "epoch": 4.026368357970435, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10078 + }, + { + "epoch": 4.026767878545745, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 10079 + }, + { + "epoch": 4.027167399121055, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10080 + }, + { + "epoch": 4.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10081 + }, + { + "epoch": 4.027966440271674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10082 + }, + { + "epoch": 4.028365960846983, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10083 + }, + { + "epoch": 4.028765481422293, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10084 + }, + { + "epoch": 4.029165001997603, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10085 + }, + { + "epoch": 4.029564522572913, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10086 + }, + { + "epoch": 4.029964043148222, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10087 + }, + { + "epoch": 4.030363563723531, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10088 + }, + { + "epoch": 4.030763084298841, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10089 + }, + { + "epoch": 4.031162604874151, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10090 + }, + { + "epoch": 4.031562125449461, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 10091 + }, + { + "epoch": 4.0319616460247705, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10092 + }, + { + "epoch": 4.03236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10093 + }, + { + "epoch": 4.032760687175389, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10094 + }, + { + "epoch": 4.033160207750699, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10095 + }, + { + "epoch": 4.033559728326009, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10096 + }, + { + "epoch": 4.0339592489013185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10097 + }, + { + "epoch": 4.034358769476628, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10098 + }, + { + "epoch": 4.034758290051938, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10099 + }, + { + "epoch": 4.035157810627247, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10100 + }, + { + "epoch": 4.035557331202557, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10101 + }, + { + "epoch": 4.035956851777867, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10102 + }, + { + "epoch": 4.036356372353176, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10103 + }, + { + "epoch": 4.036755892928486, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10104 + }, + { + "epoch": 4.037155413503795, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10105 + }, + { + "epoch": 4.037554934079105, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10106 + }, + { + "epoch": 4.037954454654415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 10107 + }, + { + "epoch": 4.038353975229724, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10108 + }, + { + "epoch": 4.038753495805034, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10109 + }, + { + "epoch": 4.039153016380344, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10110 + }, + { + "epoch": 4.039552536955653, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10111 + }, + { + "epoch": 4.039952057530963, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10112 + }, + { + "epoch": 4.040351578106272, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10113 + }, + { + "epoch": 4.040751098681582, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10114 + }, + { + "epoch": 4.041150619256892, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10115 + }, + { + "epoch": 4.041550139832202, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10116 + }, + { + "epoch": 4.041949660407511, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10117 + }, + { + "epoch": 4.0423491809828205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10118 + }, + { + "epoch": 4.04274870155813, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10119 + }, + { + "epoch": 4.04314822213344, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10120 + }, + { + "epoch": 4.04354774270875, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10121 + }, + { + "epoch": 4.043947263284059, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10122 + }, + { + "epoch": 4.0443467838593685, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10123 + }, + { + "epoch": 4.044746304434678, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10124 + }, + { + "epoch": 4.045145825009988, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10125 + }, + { + "epoch": 4.045545345585298, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10126 + }, + { + "epoch": 4.045944866160608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10127 + }, + { + "epoch": 4.046344386735917, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10128 + }, + { + "epoch": 4.046743907311226, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10129 + }, + { + "epoch": 4.047143427886536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10130 + }, + { + "epoch": 4.047542948461846, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10131 + }, + { + "epoch": 4.047942469037156, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10132 + }, + { + "epoch": 4.048341989612465, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10133 + }, + { + "epoch": 4.048741510187774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10134 + }, + { + "epoch": 4.049141030763084, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10135 + }, + { + "epoch": 4.049540551338394, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10136 + }, + { + "epoch": 4.049940071913704, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10137 + }, + { + "epoch": 4.0503395924890135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10138 + }, + { + "epoch": 4.050739113064322, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10139 + }, + { + "epoch": 4.051138633639632, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 10140 + }, + { + "epoch": 4.051538154214942, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10141 + }, + { + "epoch": 4.051937674790252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10142 + }, + { + "epoch": 4.052337195365562, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10143 + }, + { + "epoch": 4.052736715940871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10144 + }, + { + "epoch": 4.05313623651618, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10145 + }, + { + "epoch": 4.05353575709149, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10146 + }, + { + "epoch": 4.0539352776668, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10147 + }, + { + "epoch": 4.05433479824211, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10148 + }, + { + "epoch": 4.054734318817419, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10149 + }, + { + "epoch": 4.055133839392728, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10150 + }, + { + "epoch": 4.055533359968038, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10151 + }, + { + "epoch": 4.055932880543348, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10152 + }, + { + "epoch": 4.056332401118658, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10153 + }, + { + "epoch": 4.0567319216939675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10154 + }, + { + "epoch": 4.057131442269277, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10155 + }, + { + "epoch": 4.057530962844586, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10156 + }, + { + "epoch": 4.057930483419896, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10157 + }, + { + "epoch": 4.058330003995206, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10158 + }, + { + "epoch": 4.0587295245705155, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10159 + }, + { + "epoch": 4.059129045145825, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10160 + }, + { + "epoch": 4.059528565721135, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10161 + }, + { + "epoch": 4.059928086296444, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10162 + }, + { + "epoch": 4.060327606871754, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10163 + }, + { + "epoch": 4.0607271274470635, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10164 + }, + { + "epoch": 4.061126648022373, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10165 + }, + { + "epoch": 4.061526168597683, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10166 + }, + { + "epoch": 4.061925689172992, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10167 + }, + { + "epoch": 4.062325209748302, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10168 + }, + { + "epoch": 4.062724730323612, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10169 + }, + { + "epoch": 4.063124250898921, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10170 + }, + { + "epoch": 4.063523771474231, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10171 + }, + { + "epoch": 4.063923292049541, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10172 + }, + { + "epoch": 4.06432281262485, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10173 + }, + { + "epoch": 4.06472233320016, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10174 + }, + { + "epoch": 4.065121853775469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10175 + }, + { + "epoch": 4.065521374350779, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10176 + }, + { + "epoch": 4.065920894926089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10177 + }, + { + "epoch": 4.066320415501398, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10178 + }, + { + "epoch": 4.066719936076708, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10179 + }, + { + "epoch": 4.0671194566520175, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10180 + }, + { + "epoch": 4.067518977227327, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10181 + }, + { + "epoch": 4.067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10182 + }, + { + "epoch": 4.068318018377947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10183 + }, + { + "epoch": 4.068717538953256, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10184 + }, + { + "epoch": 4.0691170595285655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10185 + }, + { + "epoch": 4.069516580103875, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10186 + }, + { + "epoch": 4.069916100679185, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10187 + }, + { + "epoch": 4.070315621254495, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10188 + }, + { + "epoch": 4.070715141829805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10189 + }, + { + "epoch": 4.0711146624051135, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10190 + }, + { + "epoch": 4.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10191 + }, + { + "epoch": 4.071913703555733, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10192 + }, + { + "epoch": 4.072313224131043, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10193 + }, + { + "epoch": 4.072712744706353, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10194 + }, + { + "epoch": 4.073112265281662, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10195 + }, + { + "epoch": 4.073511785856971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10196 + }, + { + "epoch": 4.073911306432281, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10197 + }, + { + "epoch": 4.074310827007591, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10198 + }, + { + "epoch": 4.074710347582901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10199 + }, + { + "epoch": 4.0751098681582105, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10200 + }, + { + "epoch": 4.075509388733519, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10201 + }, + { + "epoch": 4.075908909308829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 10202 + }, + { + "epoch": 4.076308429884139, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10203 + }, + { + "epoch": 4.076707950459449, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 10204 + }, + { + "epoch": 4.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10205 + }, + { + "epoch": 4.077506991610068, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10206 + }, + { + "epoch": 4.077906512185377, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10207 + }, + { + "epoch": 4.078306032760687, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10208 + }, + { + "epoch": 4.078705553335997, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10209 + }, + { + "epoch": 4.079105073911307, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10210 + }, + { + "epoch": 4.079504594486616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10211 + }, + { + "epoch": 4.079904115061925, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10212 + }, + { + "epoch": 4.080303635637235, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10213 + }, + { + "epoch": 4.080703156212545, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10214 + }, + { + "epoch": 4.081102676787855, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10215 + }, + { + "epoch": 4.081502197363164, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10216 + }, + { + "epoch": 4.081901717938474, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10217 + }, + { + "epoch": 4.082301238513783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10218 + }, + { + "epoch": 4.082700759089093, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10219 + }, + { + "epoch": 4.083100279664403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10220 + }, + { + "epoch": 4.0834998002397125, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10221 + }, + { + "epoch": 4.083899320815022, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10222 + }, + { + "epoch": 4.084298841390331, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10223 + }, + { + "epoch": 4.084698361965641, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10224 + }, + { + "epoch": 4.085097882540951, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 10225 + }, + { + "epoch": 4.0854974031162605, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10226 + }, + { + "epoch": 4.08589692369157, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10227 + }, + { + "epoch": 4.08629644426688, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10228 + }, + { + "epoch": 4.086695964842189, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10229 + }, + { + "epoch": 4.087095485417499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10230 + }, + { + "epoch": 4.087495005992809, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10231 + }, + { + "epoch": 4.087894526568118, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10232 + }, + { + "epoch": 4.088294047143428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10233 + }, + { + "epoch": 4.088693567718738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10234 + }, + { + "epoch": 4.089093088294047, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10235 + }, + { + "epoch": 4.089492608869357, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10236 + }, + { + "epoch": 4.089892129444666, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10237 + }, + { + "epoch": 4.090291650019976, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10238 + }, + { + "epoch": 4.090691170595286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10239 + }, + { + "epoch": 4.091090691170595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10240 + }, + { + "epoch": 4.091490211745905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10241 + }, + { + "epoch": 4.0918897323212144, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10242 + }, + { + "epoch": 4.092289252896524, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10243 + }, + { + "epoch": 4.092688773471834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10244 + }, + { + "epoch": 4.093088294047144, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10245 + }, + { + "epoch": 4.093487814622453, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10246 + }, + { + "epoch": 4.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10247 + }, + { + "epoch": 4.094286855773072, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10248 + }, + { + "epoch": 4.094686376348382, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 10249 + }, + { + "epoch": 4.095085896923692, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10250 + }, + { + "epoch": 4.095485417499002, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10251 + }, + { + "epoch": 4.0958849380743105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10252 + }, + { + "epoch": 4.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10253 + }, + { + "epoch": 4.09668397922493, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10254 + }, + { + "epoch": 4.09708349980024, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10255 + }, + { + "epoch": 4.09748302037555, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10256 + }, + { + "epoch": 4.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10257 + }, + { + "epoch": 4.098282061526168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10258 + }, + { + "epoch": 4.098681582101478, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10259 + }, + { + "epoch": 4.099081102676788, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10260 + }, + { + "epoch": 4.099480623252098, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10261 + }, + { + "epoch": 4.0998801438274075, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10262 + }, + { + "epoch": 4.100279664402716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10263 + }, + { + "epoch": 4.100679184978026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10264 + }, + { + "epoch": 4.101078705553336, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10265 + }, + { + "epoch": 4.101478226128646, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10266 + }, + { + "epoch": 4.1018777467039556, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 10267 + }, + { + "epoch": 4.1022772672792644, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10268 + }, + { + "epoch": 4.102676787854574, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10269 + }, + { + "epoch": 4.103076308429884, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10270 + }, + { + "epoch": 4.103475829005194, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10271 + }, + { + "epoch": 4.103875349580504, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10272 + }, + { + "epoch": 4.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 10273 + }, + { + "epoch": 4.104674390731122, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10274 + }, + { + "epoch": 4.105073911306432, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10275 + }, + { + "epoch": 4.105473431881742, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10276 + }, + { + "epoch": 4.105872952457052, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10277 + }, + { + "epoch": 4.106272473032361, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 10278 + }, + { + "epoch": 4.106671993607671, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10279 + }, + { + "epoch": 4.10707151418298, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10280 + }, + { + "epoch": 4.10747103475829, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10281 + }, + { + "epoch": 4.1078705553336, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10282 + }, + { + "epoch": 4.1082700759089095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10283 + }, + { + "epoch": 4.108669596484219, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10284 + }, + { + "epoch": 4.109069117059528, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10285 + }, + { + "epoch": 4.109468637634838, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10286 + }, + { + "epoch": 4.109868158210148, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10287 + }, + { + "epoch": 4.1102676787854575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10288 + }, + { + "epoch": 4.110667199360767, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10289 + }, + { + "epoch": 4.111066719936077, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10290 + }, + { + "epoch": 4.111466240511386, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10291 + }, + { + "epoch": 4.111865761086696, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10292 + }, + { + "epoch": 4.112265281662006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10293 + }, + { + "epoch": 4.112664802237315, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10294 + }, + { + "epoch": 4.113064322812625, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 10295 + }, + { + "epoch": 4.113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10296 + }, + { + "epoch": 4.113863363963244, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10297 + }, + { + "epoch": 4.114262884538554, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10298 + }, + { + "epoch": 4.114662405113863, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10299 + }, + { + "epoch": 4.115061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10300 + }, + { + "epoch": 4.115461446264483, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10301 + }, + { + "epoch": 4.115860966839792, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10302 + }, + { + "epoch": 4.116260487415102, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10303 + }, + { + "epoch": 4.116660007990411, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10304 + }, + { + "epoch": 4.117059528565721, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10305 + }, + { + "epoch": 4.117459049141031, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10306 + }, + { + "epoch": 4.117858569716341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10307 + }, + { + "epoch": 4.11825809029165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10308 + }, + { + "epoch": 4.1186576108669595, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10309 + }, + { + "epoch": 4.119057131442269, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10310 + }, + { + "epoch": 4.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 10311 + }, + { + "epoch": 4.119856172592889, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10312 + }, + { + "epoch": 4.120255693168198, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10313 + }, + { + "epoch": 4.1206552137435075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 10314 + }, + { + "epoch": 4.121054734318817, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10315 + }, + { + "epoch": 4.121454254894127, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10316 + }, + { + "epoch": 4.121853775469437, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10317 + }, + { + "epoch": 4.122253296044747, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10318 + }, + { + "epoch": 4.122652816620056, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10319 + }, + { + "epoch": 4.123052337195365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10320 + }, + { + "epoch": 4.123451857770675, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10321 + }, + { + "epoch": 4.123851378345985, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10322 + }, + { + "epoch": 4.124250898921295, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10323 + }, + { + "epoch": 4.1246504194966045, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10324 + }, + { + "epoch": 4.125049940071913, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10325 + }, + { + "epoch": 4.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10326 + }, + { + "epoch": 4.125848981222533, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10327 + }, + { + "epoch": 4.126248501797843, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 10328 + }, + { + "epoch": 4.1266480223731525, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10329 + }, + { + "epoch": 4.127047542948461, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10330 + }, + { + "epoch": 4.127447063523771, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10331 + }, + { + "epoch": 4.127846584099081, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10332 + }, + { + "epoch": 4.128246104674391, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10333 + }, + { + "epoch": 4.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10334 + }, + { + "epoch": 4.12904514582501, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10335 + }, + { + "epoch": 4.129444666400319, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10336 + }, + { + "epoch": 4.129844186975629, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 10337 + }, + { + "epoch": 4.130243707550939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10338 + }, + { + "epoch": 4.130643228126249, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10339 + }, + { + "epoch": 4.131042748701558, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10340 + }, + { + "epoch": 4.131442269276867, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10341 + }, + { + "epoch": 4.131841789852177, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10342 + }, + { + "epoch": 4.132241310427487, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10343 + }, + { + "epoch": 4.132640831002797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10344 + }, + { + "epoch": 4.1330403515781065, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10345 + }, + { + "epoch": 4.133439872153416, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10346 + }, + { + "epoch": 4.133839392728725, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10347 + }, + { + "epoch": 4.134238913304035, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10348 + }, + { + "epoch": 4.134638433879345, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10349 + }, + { + "epoch": 4.1350379544546545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10350 + }, + { + "epoch": 4.135437475029964, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 10351 + }, + { + "epoch": 4.135836995605274, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10352 + }, + { + "epoch": 4.136236516180583, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10353 + }, + { + "epoch": 4.136636036755893, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10354 + }, + { + "epoch": 4.1370355573312025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10355 + }, + { + "epoch": 4.137435077906512, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10356 + }, + { + "epoch": 4.137834598481822, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10357 + }, + { + "epoch": 4.138234119057131, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10358 + }, + { + "epoch": 4.138633639632441, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10359 + }, + { + "epoch": 4.139033160207751, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10360 + }, + { + "epoch": 4.13943268078306, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10361 + }, + { + "epoch": 4.13983220135837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10362 + }, + { + "epoch": 4.14023172193368, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10363 + }, + { + "epoch": 4.140631242508989, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10364 + }, + { + "epoch": 4.141030763084299, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10365 + }, + { + "epoch": 4.141430283659608, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10366 + }, + { + "epoch": 4.141829804234918, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10367 + }, + { + "epoch": 4.142229324810228, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10368 + }, + { + "epoch": 4.142628845385538, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10369 + }, + { + "epoch": 4.143028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10370 + }, + { + "epoch": 4.1434278865361565, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10371 + }, + { + "epoch": 4.143827407111466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10372 + }, + { + "epoch": 4.144226927686776, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10373 + }, + { + "epoch": 4.144626448262086, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10374 + }, + { + "epoch": 4.145025968837395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10375 + }, + { + "epoch": 4.1454254894127045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10376 + }, + { + "epoch": 4.145825009988014, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10377 + }, + { + "epoch": 4.146224530563324, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10378 + }, + { + "epoch": 4.146624051138634, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10379 + }, + { + "epoch": 4.147023571713944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10380 + }, + { + "epoch": 4.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10381 + }, + { + "epoch": 4.147822612864562, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10382 + }, + { + "epoch": 4.148222133439872, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10383 + }, + { + "epoch": 4.148621654015182, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10384 + }, + { + "epoch": 4.149021174590492, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10385 + }, + { + "epoch": 4.1494206951658015, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 10386 + }, + { + "epoch": 4.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10387 + }, + { + "epoch": 4.15021973631642, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10388 + }, + { + "epoch": 4.15061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10389 + }, + { + "epoch": 4.15101877746704, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10390 + }, + { + "epoch": 4.1514182980423495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10391 + }, + { + "epoch": 4.151817818617658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10392 + }, + { + "epoch": 4.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10393 + }, + { + "epoch": 4.152616859768278, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 10394 + }, + { + "epoch": 4.153016380343588, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10395 + }, + { + "epoch": 4.153415900918898, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 10396 + }, + { + "epoch": 4.153815421494207, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10397 + }, + { + "epoch": 4.154214942069516, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10398 + }, + { + "epoch": 4.154614462644826, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10399 + }, + { + "epoch": 4.155013983220136, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10400 + }, + { + "epoch": 4.155413503795446, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10401 + }, + { + "epoch": 4.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10402 + }, + { + "epoch": 4.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10403 + }, + { + "epoch": 4.156612065521374, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10404 + }, + { + "epoch": 4.157011586096684, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10405 + }, + { + "epoch": 4.157411106671994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10406 + }, + { + "epoch": 4.157810627247303, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10407 + }, + { + "epoch": 4.158210147822613, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10408 + }, + { + "epoch": 4.158609668397922, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10409 + }, + { + "epoch": 4.159009188973232, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10410 + }, + { + "epoch": 4.159408709548542, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10411 + }, + { + "epoch": 4.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10412 + }, + { + "epoch": 4.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10413 + }, + { + "epoch": 4.160607271274471, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10414 + }, + { + "epoch": 4.16100679184978, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10415 + }, + { + "epoch": 4.16140631242509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10416 + }, + { + "epoch": 4.1618058330003995, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10417 + }, + { + "epoch": 4.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10418 + }, + { + "epoch": 4.162604874151019, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10419 + }, + { + "epoch": 4.163004394726328, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10420 + }, + { + "epoch": 4.163403915301638, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10421 + }, + { + "epoch": 4.163803435876948, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10422 + }, + { + "epoch": 4.164202956452257, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10423 + }, + { + "epoch": 4.164602477027567, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10424 + }, + { + "epoch": 4.165001997602877, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10425 + }, + { + "epoch": 4.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10426 + }, + { + "epoch": 4.165801038753496, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10427 + }, + { + "epoch": 4.166200559328805, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10428 + }, + { + "epoch": 4.166600079904115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10429 + }, + { + "epoch": 4.166999600479425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10430 + }, + { + "epoch": 4.167399121054734, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10431 + }, + { + "epoch": 4.167798641630044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10432 + }, + { + "epoch": 4.168198162205353, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 10433 + }, + { + "epoch": 4.168597682780663, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10434 + }, + { + "epoch": 4.168997203355973, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10435 + }, + { + "epoch": 4.169396723931283, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10436 + }, + { + "epoch": 4.169796244506592, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10437 + }, + { + "epoch": 4.1701957650819015, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10438 + }, + { + "epoch": 4.170595285657211, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10439 + }, + { + "epoch": 4.170994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10440 + }, + { + "epoch": 4.171394326807831, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10441 + }, + { + "epoch": 4.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10442 + }, + { + "epoch": 4.1721933679584495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10443 + }, + { + "epoch": 4.172592888533759, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10444 + }, + { + "epoch": 4.172992409109069, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10445 + }, + { + "epoch": 4.173391929684379, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 10446 + }, + { + "epoch": 4.173791450259689, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10447 + }, + { + "epoch": 4.174190970834998, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10448 + }, + { + "epoch": 4.174590491410307, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10449 + }, + { + "epoch": 4.174990011985617, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10450 + }, + { + "epoch": 4.175389532560927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10451 + }, + { + "epoch": 4.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10452 + }, + { + "epoch": 4.1761885737115465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10453 + }, + { + "epoch": 4.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 10454 + }, + { + "epoch": 4.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10455 + }, + { + "epoch": 4.177387135437475, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10456 + }, + { + "epoch": 4.177786656012785, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10457 + }, + { + "epoch": 4.1781861765880945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10458 + }, + { + "epoch": 4.178585697163404, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10459 + }, + { + "epoch": 4.178985217738713, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10460 + }, + { + "epoch": 4.179384738314023, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10461 + }, + { + "epoch": 4.179784258889333, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10462 + }, + { + "epoch": 4.180183779464643, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10463 + }, + { + "epoch": 4.180583300039952, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10464 + }, + { + "epoch": 4.180982820615261, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10465 + }, + { + "epoch": 4.181382341190571, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10466 + }, + { + "epoch": 4.181781861765881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10467 + }, + { + "epoch": 4.182181382341191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10468 + }, + { + "epoch": 4.1825809029165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10469 + }, + { + "epoch": 4.18298042349181, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10470 + }, + { + "epoch": 4.183379944067119, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10471 + }, + { + "epoch": 4.183779464642429, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10472 + }, + { + "epoch": 4.184178985217739, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10473 + }, + { + "epoch": 4.1845785057930485, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10474 + }, + { + "epoch": 4.184978026368358, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10475 + }, + { + "epoch": 4.185377546943668, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10476 + }, + { + "epoch": 4.185777067518977, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10477 + }, + { + "epoch": 4.186176588094287, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 10478 + }, + { + "epoch": 4.1865761086695965, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10479 + }, + { + "epoch": 4.186975629244906, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10480 + }, + { + "epoch": 4.187375149820216, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10481 + }, + { + "epoch": 4.187774670395525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10482 + }, + { + "epoch": 4.188174190970835, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10483 + }, + { + "epoch": 4.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10484 + }, + { + "epoch": 4.188973232121454, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10485 + }, + { + "epoch": 4.189372752696764, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10486 + }, + { + "epoch": 4.189772273272074, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10487 + }, + { + "epoch": 4.190171793847383, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10488 + }, + { + "epoch": 4.190571314422693, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10489 + }, + { + "epoch": 4.190970834998002, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10490 + }, + { + "epoch": 4.191370355573312, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10491 + }, + { + "epoch": 4.191769876148622, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 10492 + }, + { + "epoch": 4.192169396723932, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10493 + }, + { + "epoch": 4.192568917299241, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10494 + }, + { + "epoch": 4.19296843787455, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10495 + }, + { + "epoch": 4.19336795844986, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10496 + }, + { + "epoch": 4.19376747902517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 10497 + }, + { + "epoch": 4.19416699960048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10498 + }, + { + "epoch": 4.194566520175789, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10499 + }, + { + "epoch": 4.1949660407510985, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10500 + }, + { + "epoch": 4.195365561326408, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10501 + }, + { + "epoch": 4.195765081901718, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10502 + }, + { + "epoch": 4.196164602477028, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10503 + }, + { + "epoch": 4.196564123052338, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10504 + }, + { + "epoch": 4.1969636436276465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10505 + }, + { + "epoch": 4.197363164202956, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10506 + }, + { + "epoch": 4.197762684778266, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10507 + }, + { + "epoch": 4.198162205353576, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10508 + }, + { + "epoch": 4.198561725928886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10509 + }, + { + "epoch": 4.1989612465041946, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10510 + }, + { + "epoch": 4.199360767079504, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10511 + }, + { + "epoch": 4.199760287654814, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10512 + }, + { + "epoch": 4.200159808230124, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10513 + }, + { + "epoch": 4.200559328805434, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10514 + }, + { + "epoch": 4.2009588493807435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10515 + }, + { + "epoch": 4.201358369956052, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10516 + }, + { + "epoch": 4.201757890531362, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10517 + }, + { + "epoch": 4.202157411106672, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 10518 + }, + { + "epoch": 4.202556931681982, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10519 + }, + { + "epoch": 4.2029564522572915, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10520 + }, + { + "epoch": 4.2033559728326, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10521 + }, + { + "epoch": 4.20375549340791, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10522 + }, + { + "epoch": 4.20415501398322, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10523 + }, + { + "epoch": 4.20455453455853, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10524 + }, + { + "epoch": 4.20495405513384, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10525 + }, + { + "epoch": 4.205353575709149, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10526 + }, + { + "epoch": 4.205753096284458, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10527 + }, + { + "epoch": 4.206152616859768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10528 + }, + { + "epoch": 4.206552137435078, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10529 + }, + { + "epoch": 4.206951658010388, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10530 + }, + { + "epoch": 4.207351178585697, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10531 + }, + { + "epoch": 4.207750699161007, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 10532 + }, + { + "epoch": 4.208150219736316, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10533 + }, + { + "epoch": 4.208549740311626, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10534 + }, + { + "epoch": 4.208949260886936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10535 + }, + { + "epoch": 4.2093487814622454, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10536 + }, + { + "epoch": 4.209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 10537 + }, + { + "epoch": 4.210147822612864, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10538 + }, + { + "epoch": 4.210547343188174, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10539 + }, + { + "epoch": 4.210946863763484, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10540 + }, + { + "epoch": 4.2113463843387935, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10541 + }, + { + "epoch": 4.211745904914103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10542 + }, + { + "epoch": 4.212145425489413, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10543 + }, + { + "epoch": 4.212544946064722, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10544 + }, + { + "epoch": 4.212944466640032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10545 + }, + { + "epoch": 4.2133439872153415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10546 + }, + { + "epoch": 4.213743507790651, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10547 + }, + { + "epoch": 4.214143028365961, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10548 + }, + { + "epoch": 4.214542548941271, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10549 + }, + { + "epoch": 4.21494206951658, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10550 + }, + { + "epoch": 4.21534159009189, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10551 + }, + { + "epoch": 4.215741110667199, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10552 + }, + { + "epoch": 4.216140631242509, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10553 + }, + { + "epoch": 4.216540151817819, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10554 + }, + { + "epoch": 4.216939672393128, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10555 + }, + { + "epoch": 4.217339192968438, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10556 + }, + { + "epoch": 4.217738713543747, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10557 + }, + { + "epoch": 4.218138234119057, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10558 + }, + { + "epoch": 4.218537754694367, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10559 + }, + { + "epoch": 4.218937275269677, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10560 + }, + { + "epoch": 4.219336795844986, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10561 + }, + { + "epoch": 4.2197363164202955, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10562 + }, + { + "epoch": 4.220135836995605, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10563 + }, + { + "epoch": 4.220535357570915, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10564 + }, + { + "epoch": 4.220934878146225, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10565 + }, + { + "epoch": 4.221334398721535, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10566 + }, + { + "epoch": 4.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10567 + }, + { + "epoch": 4.222133439872153, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10568 + }, + { + "epoch": 4.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10569 + }, + { + "epoch": 4.222932481022773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10570 + }, + { + "epoch": 4.223332001598083, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10571 + }, + { + "epoch": 4.2237315221733915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10572 + }, + { + "epoch": 4.224131042748701, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10573 + }, + { + "epoch": 4.224530563324011, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10574 + }, + { + "epoch": 4.224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10575 + }, + { + "epoch": 4.225329604474631, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10576 + }, + { + "epoch": 4.2257291250499405, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10577 + }, + { + "epoch": 4.226128645625249, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 10578 + }, + { + "epoch": 4.226528166200559, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10579 + }, + { + "epoch": 4.226927686775869, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10580 + }, + { + "epoch": 4.227327207351179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10581 + }, + { + "epoch": 4.2277267279264885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10582 + }, + { + "epoch": 4.228126248501797, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10583 + }, + { + "epoch": 4.228525769077107, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10584 + }, + { + "epoch": 4.228925289652417, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10585 + }, + { + "epoch": 4.229324810227727, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10586 + }, + { + "epoch": 4.229724330803037, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 10587 + }, + { + "epoch": 4.230123851378346, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10588 + }, + { + "epoch": 4.230523371953655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10589 + }, + { + "epoch": 4.230922892528965, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10590 + }, + { + "epoch": 4.231322413104275, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10591 + }, + { + "epoch": 4.231721933679585, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 10592 + }, + { + "epoch": 4.232121454254894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10593 + }, + { + "epoch": 4.232520974830204, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10594 + }, + { + "epoch": 4.232920495405513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10595 + }, + { + "epoch": 4.233320015980823, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10596 + }, + { + "epoch": 4.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10597 + }, + { + "epoch": 4.234119057131442, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10598 + }, + { + "epoch": 4.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10599 + }, + { + "epoch": 4.234918098282061, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10600 + }, + { + "epoch": 4.235317618857371, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10601 + }, + { + "epoch": 4.235717139432681, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10602 + }, + { + "epoch": 4.2361166600079905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10603 + }, + { + "epoch": 4.2365161805833, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10604 + }, + { + "epoch": 4.23691570115861, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 10605 + }, + { + "epoch": 4.237315221733919, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10606 + }, + { + "epoch": 4.237714742309229, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10607 + }, + { + "epoch": 4.2381142628845385, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10608 + }, + { + "epoch": 4.238513783459848, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10609 + }, + { + "epoch": 4.238913304035158, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10610 + }, + { + "epoch": 4.239312824610467, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10611 + }, + { + "epoch": 4.239712345185777, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 10612 + }, + { + "epoch": 4.240111865761087, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10613 + }, + { + "epoch": 4.240511386336396, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10614 + }, + { + "epoch": 4.240910906911706, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 10615 + }, + { + "epoch": 4.241310427487016, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10616 + }, + { + "epoch": 4.241709948062325, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10617 + }, + { + "epoch": 4.242109468637635, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10618 + }, + { + "epoch": 4.242508989212944, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10619 + }, + { + "epoch": 4.242908509788254, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10620 + }, + { + "epoch": 4.243308030363564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10621 + }, + { + "epoch": 4.243707550938874, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 10622 + }, + { + "epoch": 4.244107071514183, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10623 + }, + { + "epoch": 4.244506592089492, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10624 + }, + { + "epoch": 4.244906112664802, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10625 + }, + { + "epoch": 4.245305633240112, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10626 + }, + { + "epoch": 4.245705153815422, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10627 + }, + { + "epoch": 4.246104674390731, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 10628 + }, + { + "epoch": 4.2465041949660405, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10629 + }, + { + "epoch": 4.24690371554135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10630 + }, + { + "epoch": 4.24730323611666, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10631 + }, + { + "epoch": 4.24770275669197, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10632 + }, + { + "epoch": 4.24810227726728, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10633 + }, + { + "epoch": 4.2485017978425885, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10634 + }, + { + "epoch": 4.248901318417898, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10635 + }, + { + "epoch": 4.249300838993208, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10636 + }, + { + "epoch": 4.249700359568518, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10637 + }, + { + "epoch": 4.250099880143828, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10638 + }, + { + "epoch": 4.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10639 + }, + { + "epoch": 4.250898921294446, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10640 + }, + { + "epoch": 4.251298441869756, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10641 + }, + { + "epoch": 4.251697962445066, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10642 + }, + { + "epoch": 4.252097483020376, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10643 + }, + { + "epoch": 4.2524970035956855, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10644 + }, + { + "epoch": 4.252896524170994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10645 + }, + { + "epoch": 4.253296044746304, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10646 + }, + { + "epoch": 4.253695565321614, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10647 + }, + { + "epoch": 4.254095085896924, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10648 + }, + { + "epoch": 4.2544946064722335, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10649 + }, + { + "epoch": 4.254894127047543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 10650 + }, + { + "epoch": 4.255293647622852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10651 + }, + { + "epoch": 4.255693168198162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10652 + }, + { + "epoch": 4.256092688773472, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10653 + }, + { + "epoch": 4.256492209348782, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10654 + }, + { + "epoch": 4.256891729924091, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10655 + }, + { + "epoch": 4.257291250499401, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10656 + }, + { + "epoch": 4.25769077107471, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10657 + }, + { + "epoch": 4.25809029165002, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10658 + }, + { + "epoch": 4.25848981222533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10659 + }, + { + "epoch": 4.258889332800639, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10660 + }, + { + "epoch": 4.259288853375949, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10661 + }, + { + "epoch": 4.259688373951258, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10662 + }, + { + "epoch": 4.260087894526568, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10663 + }, + { + "epoch": 4.260487415101878, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10664 + }, + { + "epoch": 4.2608869356771875, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10665 + }, + { + "epoch": 4.261286456252497, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10666 + }, + { + "epoch": 4.261685976827807, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10667 + }, + { + "epoch": 4.262085497403116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10668 + }, + { + "epoch": 4.262485017978426, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10669 + }, + { + "epoch": 4.2628845385537355, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 10670 + }, + { + "epoch": 4.263284059129045, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10671 + }, + { + "epoch": 4.263683579704355, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10672 + }, + { + "epoch": 4.264083100279665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10673 + }, + { + "epoch": 4.264482620854974, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10674 + }, + { + "epoch": 4.2648821414302835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10675 + }, + { + "epoch": 4.265281662005593, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10676 + }, + { + "epoch": 4.265681182580903, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10677 + }, + { + "epoch": 4.266080703156213, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10678 + }, + { + "epoch": 4.266480223731522, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10679 + }, + { + "epoch": 4.266879744306832, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10680 + }, + { + "epoch": 4.267279264882141, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 10681 + }, + { + "epoch": 4.267678785457451, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10682 + }, + { + "epoch": 4.268078306032761, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 10683 + }, + { + "epoch": 4.26847782660807, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10684 + }, + { + "epoch": 4.26887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 10685 + }, + { + "epoch": 4.269276867758689, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 10686 + }, + { + "epoch": 4.269676388333999, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 10687 + }, + { + "epoch": 4.270075908909309, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10688 + }, + { + "epoch": 4.270475429484619, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10689 + }, + { + "epoch": 4.270874950059928, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10690 + }, + { + "epoch": 4.2712744706352375, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10691 + }, + { + "epoch": 4.271673991210547, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 10692 + }, + { + "epoch": 4.272073511785857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10693 + }, + { + "epoch": 4.272473032361167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10694 + }, + { + "epoch": 4.272872552936477, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10695 + }, + { + "epoch": 4.2732720735117855, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10696 + }, + { + "epoch": 4.273671594087095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10697 + }, + { + "epoch": 4.274071114662405, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10698 + }, + { + "epoch": 4.274470635237715, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10699 + }, + { + "epoch": 4.274870155813025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10700 + }, + { + "epoch": 4.2752696763883336, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10701 + }, + { + "epoch": 4.275669196963643, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10702 + }, + { + "epoch": 4.276068717538953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10703 + }, + { + "epoch": 4.276468238114263, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10704 + }, + { + "epoch": 4.276867758689573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 10705 + }, + { + "epoch": 4.2772672792648825, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10706 + }, + { + "epoch": 4.277666799840191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10707 + }, + { + "epoch": 4.278066320415501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10708 + }, + { + "epoch": 4.278465840990811, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10709 + }, + { + "epoch": 4.278865361566121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10710 + }, + { + "epoch": 4.2792648821414305, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10711 + }, + { + "epoch": 4.27966440271674, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10712 + }, + { + "epoch": 4.280063923292049, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10713 + }, + { + "epoch": 4.280463443867359, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10714 + }, + { + "epoch": 4.280862964442669, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10715 + }, + { + "epoch": 4.281262485017979, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10716 + }, + { + "epoch": 4.281662005593288, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10717 + }, + { + "epoch": 4.282061526168597, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10718 + }, + { + "epoch": 4.282461046743907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10719 + }, + { + "epoch": 4.282860567319217, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10720 + }, + { + "epoch": 4.283260087894527, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10721 + }, + { + "epoch": 4.283659608469836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10722 + }, + { + "epoch": 4.284059129045146, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10723 + }, + { + "epoch": 4.284458649620455, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10724 + }, + { + "epoch": 4.284858170195765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 10725 + }, + { + "epoch": 4.285257690771075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10726 + }, + { + "epoch": 4.285657211346384, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 10727 + }, + { + "epoch": 4.286056731921694, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10728 + }, + { + "epoch": 4.286456252497004, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10729 + }, + { + "epoch": 4.286855773072313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10730 + }, + { + "epoch": 4.287255293647623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10731 + }, + { + "epoch": 4.2876548142229325, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10732 + }, + { + "epoch": 4.288054334798242, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10733 + }, + { + "epoch": 4.288453855373552, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10734 + }, + { + "epoch": 4.288853375948861, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10735 + }, + { + "epoch": 4.289252896524171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 10736 + }, + { + "epoch": 4.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10737 + }, + { + "epoch": 4.29005193767479, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10738 + }, + { + "epoch": 4.2904514582501, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10739 + }, + { + "epoch": 4.29085097882541, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10740 + }, + { + "epoch": 4.291250499400719, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10741 + }, + { + "epoch": 4.291650019976029, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10742 + }, + { + "epoch": 4.292049540551338, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10743 + }, + { + "epoch": 4.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10744 + }, + { + "epoch": 4.292848581701958, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10745 + }, + { + "epoch": 4.293248102277268, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10746 + }, + { + "epoch": 4.293647622852577, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10747 + }, + { + "epoch": 4.294047143427886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10748 + }, + { + "epoch": 4.294446664003196, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10749 + }, + { + "epoch": 4.294846184578506, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10750 + }, + { + "epoch": 4.295245705153816, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10751 + }, + { + "epoch": 4.295645225729125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 10752 + }, + { + "epoch": 4.2960447463044344, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10753 + }, + { + "epoch": 4.296444266879744, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10754 + }, + { + "epoch": 4.296843787455054, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10755 + }, + { + "epoch": 4.297243308030364, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 10756 + }, + { + "epoch": 4.297642828605674, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10757 + }, + { + "epoch": 4.2980423491809825, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10758 + }, + { + "epoch": 4.298441869756292, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10759 + }, + { + "epoch": 4.298841390331602, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10760 + }, + { + "epoch": 4.299240910906912, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10761 + }, + { + "epoch": 4.299640431482222, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 10762 + }, + { + "epoch": 4.300039952057531, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10763 + }, + { + "epoch": 4.30043947263284, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10764 + }, + { + "epoch": 4.30083899320815, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10765 + }, + { + "epoch": 4.30123851378346, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10766 + }, + { + "epoch": 4.30163803435877, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 10767 + }, + { + "epoch": 4.3020375549340795, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10768 + }, + { + "epoch": 4.302437075509388, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10769 + }, + { + "epoch": 4.302836596084698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10770 + }, + { + "epoch": 4.303236116660008, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10771 + }, + { + "epoch": 4.303635637235318, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10772 + }, + { + "epoch": 4.3040351578106275, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10773 + }, + { + "epoch": 4.304434678385936, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10774 + }, + { + "epoch": 4.304834198961246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10775 + }, + { + "epoch": 4.305233719536556, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10776 + }, + { + "epoch": 4.305633240111866, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10777 + }, + { + "epoch": 4.3060327606871756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10778 + }, + { + "epoch": 4.306432281262485, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10779 + }, + { + "epoch": 4.306831801837794, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10780 + }, + { + "epoch": 4.307231322413104, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10781 + }, + { + "epoch": 4.307630842988414, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10782 + }, + { + "epoch": 4.308030363563724, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10783 + }, + { + "epoch": 4.308429884139033, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10784 + }, + { + "epoch": 4.308829404714343, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10785 + }, + { + "epoch": 4.309228925289652, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10786 + }, + { + "epoch": 4.309628445864962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10787 + }, + { + "epoch": 4.310027966440272, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10788 + }, + { + "epoch": 4.310427487015581, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10789 + }, + { + "epoch": 4.310827007590891, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10790 + }, + { + "epoch": 4.3112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10791 + }, + { + "epoch": 4.31162604874151, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10792 + }, + { + "epoch": 4.31202556931682, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10793 + }, + { + "epoch": 4.3124250898921295, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10794 + }, + { + "epoch": 4.312824610467439, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10795 + }, + { + "epoch": 4.313224131042749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10796 + }, + { + "epoch": 4.313623651618058, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10797 + }, + { + "epoch": 4.314023172193368, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10798 + }, + { + "epoch": 4.3144226927686775, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10799 + }, + { + "epoch": 4.314822213343987, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10800 + }, + { + "epoch": 4.315221733919297, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10801 + }, + { + "epoch": 4.315621254494607, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10802 + }, + { + "epoch": 4.316020775069916, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10803 + }, + { + "epoch": 4.316420295645226, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 10804 + }, + { + "epoch": 4.316819816220535, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10805 + }, + { + "epoch": 4.317219336795845, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10806 + }, + { + "epoch": 4.317618857371155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10807 + }, + { + "epoch": 4.318018377946464, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10808 + }, + { + "epoch": 4.318417898521774, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10809 + }, + { + "epoch": 4.318817419097083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10810 + }, + { + "epoch": 4.319216939672393, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10811 + }, + { + "epoch": 4.319616460247703, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10812 + }, + { + "epoch": 4.320015980823013, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10813 + }, + { + "epoch": 4.320415501398322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10814 + }, + { + "epoch": 4.320815021973631, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10815 + }, + { + "epoch": 4.321214542548941, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10816 + }, + { + "epoch": 4.321614063124251, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10817 + }, + { + "epoch": 4.322013583699561, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10818 + }, + { + "epoch": 4.322413104274871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10819 + }, + { + "epoch": 4.3228126248501795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10820 + }, + { + "epoch": 4.323212145425489, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10821 + }, + { + "epoch": 4.323611666000799, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 10822 + }, + { + "epoch": 4.324011186576109, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 10823 + }, + { + "epoch": 4.324410707151419, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10824 + }, + { + "epoch": 4.3248102277267275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10825 + }, + { + "epoch": 4.325209748302037, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10826 + }, + { + "epoch": 4.325609268877347, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10827 + }, + { + "epoch": 4.326008789452657, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10828 + }, + { + "epoch": 4.326408310027967, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10829 + }, + { + "epoch": 4.3268078306032765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10830 + }, + { + "epoch": 4.327207351178585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10831 + }, + { + "epoch": 4.327606871753895, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10832 + }, + { + "epoch": 4.328006392329205, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10833 + }, + { + "epoch": 4.328405912904515, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10834 + }, + { + "epoch": 4.3288054334798245, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10835 + }, + { + "epoch": 4.329204954055134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10836 + }, + { + "epoch": 4.329604474630443, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10837 + }, + { + "epoch": 4.330003995205753, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10838 + }, + { + "epoch": 4.330403515781063, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10839 + }, + { + "epoch": 4.3308030363563725, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10840 + }, + { + "epoch": 4.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10841 + }, + { + "epoch": 4.331602077506991, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10842 + }, + { + "epoch": 4.332001598082301, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10843 + }, + { + "epoch": 4.332401118657611, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10844 + }, + { + "epoch": 4.332800639232921, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10845 + }, + { + "epoch": 4.33320015980823, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10846 + }, + { + "epoch": 4.33359968038354, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10847 + }, + { + "epoch": 4.333999200958849, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10848 + }, + { + "epoch": 4.334398721534159, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10849 + }, + { + "epoch": 4.334798242109469, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10850 + }, + { + "epoch": 4.335197762684778, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10851 + }, + { + "epoch": 4.335597283260088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10852 + }, + { + "epoch": 4.335996803835398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10853 + }, + { + "epoch": 4.336396324410707, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10854 + }, + { + "epoch": 4.336795844986017, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10855 + }, + { + "epoch": 4.3371953655613265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10856 + }, + { + "epoch": 4.337594886136636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10857 + }, + { + "epoch": 4.337994406711946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10858 + }, + { + "epoch": 4.338393927287255, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10859 + }, + { + "epoch": 4.338793447862565, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10860 + }, + { + "epoch": 4.3391929684378745, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10861 + }, + { + "epoch": 4.339592489013184, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10862 + }, + { + "epoch": 4.339992009588494, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 10863 + }, + { + "epoch": 4.340391530163803, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 10864 + }, + { + "epoch": 4.340791050739113, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10865 + }, + { + "epoch": 4.3411905713144225, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10866 + }, + { + "epoch": 4.341590091889732, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10867 + }, + { + "epoch": 4.341989612465042, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10868 + }, + { + "epoch": 4.342389133040352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10869 + }, + { + "epoch": 4.342788653615661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10870 + }, + { + "epoch": 4.343188174190971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10871 + }, + { + "epoch": 4.34358769476628, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 10872 + }, + { + "epoch": 4.34398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10873 + }, + { + "epoch": 4.3443867359169, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10874 + }, + { + "epoch": 4.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10875 + }, + { + "epoch": 4.345185777067519, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10876 + }, + { + "epoch": 4.345585297642828, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10877 + }, + { + "epoch": 4.345984818218138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10878 + }, + { + "epoch": 4.346384338793448, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10879 + }, + { + "epoch": 4.346783859368758, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10880 + }, + { + "epoch": 4.347183379944067, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10881 + }, + { + "epoch": 4.3475829005193765, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10882 + }, + { + "epoch": 4.347982421094686, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10883 + }, + { + "epoch": 4.348381941669996, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10884 + }, + { + "epoch": 4.348781462245306, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10885 + }, + { + "epoch": 4.349180982820616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10886 + }, + { + "epoch": 4.3495805033959245, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10887 + }, + { + "epoch": 4.349980023971234, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10888 + }, + { + "epoch": 4.350379544546544, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10889 + }, + { + "epoch": 4.350779065121854, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10890 + }, + { + "epoch": 4.351178585697164, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10891 + }, + { + "epoch": 4.351578106272473, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10892 + }, + { + "epoch": 4.351977626847782, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 10893 + }, + { + "epoch": 4.352377147423092, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10894 + }, + { + "epoch": 4.352776667998402, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10895 + }, + { + "epoch": 4.353176188573712, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10896 + }, + { + "epoch": 4.3535757091490215, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10897 + }, + { + "epoch": 4.35397522972433, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10898 + }, + { + "epoch": 4.35437475029964, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10899 + }, + { + "epoch": 4.35477427087495, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10900 + }, + { + "epoch": 4.35517379145026, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10901 + }, + { + "epoch": 4.3555733120255695, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10902 + }, + { + "epoch": 4.355972832600879, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10903 + }, + { + "epoch": 4.356372353176188, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10904 + }, + { + "epoch": 4.356771873751498, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 10905 + }, + { + "epoch": 4.357171394326808, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10906 + }, + { + "epoch": 4.357570914902118, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10907 + }, + { + "epoch": 4.357970435477427, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10908 + }, + { + "epoch": 4.358369956052737, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10909 + }, + { + "epoch": 4.358769476628046, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10910 + }, + { + "epoch": 4.359168997203356, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10911 + }, + { + "epoch": 4.359568517778666, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10912 + }, + { + "epoch": 4.359968038353975, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10913 + }, + { + "epoch": 4.360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10914 + }, + { + "epoch": 4.360767079504594, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10915 + }, + { + "epoch": 4.361166600079904, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10916 + }, + { + "epoch": 4.361566120655214, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10917 + }, + { + "epoch": 4.361965641230523, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10918 + }, + { + "epoch": 4.362365161805833, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10919 + }, + { + "epoch": 4.362764682381143, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10920 + }, + { + "epoch": 4.363164202956452, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10921 + }, + { + "epoch": 4.363563723531762, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10922 + }, + { + "epoch": 4.3639632441070715, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10923 + }, + { + "epoch": 4.364362764682381, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10924 + }, + { + "epoch": 4.364762285257691, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10925 + }, + { + "epoch": 4.365161805833001, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10926 + }, + { + "epoch": 4.36556132640831, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10927 + }, + { + "epoch": 4.3659608469836195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10928 + }, + { + "epoch": 4.366360367558929, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10929 + }, + { + "epoch": 4.366759888134239, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10930 + }, + { + "epoch": 4.367159408709549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10931 + }, + { + "epoch": 4.367558929284858, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10932 + }, + { + "epoch": 4.367958449860168, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10933 + }, + { + "epoch": 4.368357970435477, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10934 + }, + { + "epoch": 4.368757491010787, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10935 + }, + { + "epoch": 4.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10936 + }, + { + "epoch": 4.369556532161407, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10937 + }, + { + "epoch": 4.369956052736716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10938 + }, + { + "epoch": 4.370355573312025, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10939 + }, + { + "epoch": 4.370755093887335, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10940 + }, + { + "epoch": 4.371154614462645, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10941 + }, + { + "epoch": 4.371554135037955, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10942 + }, + { + "epoch": 4.3719536556132645, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10943 + }, + { + "epoch": 4.372353176188573, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10944 + }, + { + "epoch": 4.372752696763883, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 10945 + }, + { + "epoch": 4.373152217339193, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10946 + }, + { + "epoch": 4.373551737914503, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10947 + }, + { + "epoch": 4.373951258489813, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10948 + }, + { + "epoch": 4.3743507790651215, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10949 + }, + { + "epoch": 4.374750299640431, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 10950 + }, + { + "epoch": 4.375149820215741, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 10951 + }, + { + "epoch": 4.375549340791051, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10952 + }, + { + "epoch": 4.375948861366361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10953 + }, + { + "epoch": 4.3763483819416695, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 10954 + }, + { + "epoch": 4.376747902516979, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10955 + }, + { + "epoch": 4.377147423092289, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10956 + }, + { + "epoch": 4.377546943667599, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10957 + }, + { + "epoch": 4.377946464242909, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10958 + }, + { + "epoch": 4.3783459848182185, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10959 + }, + { + "epoch": 4.378745505393527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10960 + }, + { + "epoch": 4.379145025968837, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10961 + }, + { + "epoch": 4.379544546544147, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 10962 + }, + { + "epoch": 4.379944067119457, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10963 + }, + { + "epoch": 4.3803435876947665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10964 + }, + { + "epoch": 4.380743108270076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10965 + }, + { + "epoch": 4.381142628845385, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10966 + }, + { + "epoch": 4.381542149420695, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10967 + }, + { + "epoch": 4.381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10968 + }, + { + "epoch": 4.3823411905713145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10969 + }, + { + "epoch": 4.382740711146624, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10970 + }, + { + "epoch": 4.383140231721933, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10971 + }, + { + "epoch": 4.383539752297243, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10972 + }, + { + "epoch": 4.383939272872553, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10973 + }, + { + "epoch": 4.384338793447863, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10974 + }, + { + "epoch": 4.384738314023172, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10975 + }, + { + "epoch": 4.385137834598482, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10976 + }, + { + "epoch": 4.385537355173791, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10977 + }, + { + "epoch": 4.385936875749101, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10978 + }, + { + "epoch": 4.386336396324411, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10979 + }, + { + "epoch": 4.38673591689972, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10980 + }, + { + "epoch": 4.38713543747503, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10981 + }, + { + "epoch": 4.38753495805034, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10982 + }, + { + "epoch": 4.387934478625649, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10983 + }, + { + "epoch": 4.388333999200959, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10984 + }, + { + "epoch": 4.3887335197762685, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10985 + }, + { + "epoch": 4.389133040351578, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 10986 + }, + { + "epoch": 4.389532560926888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 10987 + }, + { + "epoch": 4.389932081502197, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10988 + }, + { + "epoch": 4.390331602077507, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10989 + }, + { + "epoch": 4.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 10990 + }, + { + "epoch": 4.391130643228126, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10991 + }, + { + "epoch": 4.391530163803436, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10992 + }, + { + "epoch": 4.391929684378746, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10993 + }, + { + "epoch": 4.392329204954055, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 10994 + }, + { + "epoch": 4.3927287255293646, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10995 + }, + { + "epoch": 4.393128246104674, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10996 + }, + { + "epoch": 4.393527766679984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10997 + }, + { + "epoch": 4.393927287255294, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10998 + }, + { + "epoch": 4.394326807830604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10999 + }, + { + "epoch": 4.394726328405913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11000 + }, + { + "epoch": 4.395125848981222, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11001 + }, + { + "epoch": 4.395525369556532, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11002 + }, + { + "epoch": 4.395924890131842, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 11003 + }, + { + "epoch": 4.396324410707152, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11004 + }, + { + "epoch": 4.396723931282461, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11005 + }, + { + "epoch": 4.39712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11006 + }, + { + "epoch": 4.39752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11007 + }, + { + "epoch": 4.39792249300839, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11008 + }, + { + "epoch": 4.3983220135837, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11009 + }, + { + "epoch": 4.39872153415901, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11010 + }, + { + "epoch": 4.3991210547343185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11011 + }, + { + "epoch": 4.399520575309628, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11012 + }, + { + "epoch": 4.399920095884938, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11013 + }, + { + "epoch": 4.400319616460248, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11014 + }, + { + "epoch": 4.400719137035558, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11015 + }, + { + "epoch": 4.401118657610867, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11016 + }, + { + "epoch": 4.401518178186176, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11017 + }, + { + "epoch": 4.401917698761486, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11018 + }, + { + "epoch": 4.402317219336796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11019 + }, + { + "epoch": 4.402716739912106, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11020 + }, + { + "epoch": 4.4031162604874154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11021 + }, + { + "epoch": 4.403515781062724, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11022 + }, + { + "epoch": 4.403915301638034, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11023 + }, + { + "epoch": 4.404314822213344, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11024 + }, + { + "epoch": 4.404714342788654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11025 + }, + { + "epoch": 4.4051138633639635, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11026 + }, + { + "epoch": 4.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11027 + }, + { + "epoch": 4.405912904514582, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11028 + }, + { + "epoch": 4.406312425089892, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11029 + }, + { + "epoch": 4.406711945665202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11030 + }, + { + "epoch": 4.4071114662405115, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11031 + }, + { + "epoch": 4.407510986815821, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11032 + }, + { + "epoch": 4.407910507391131, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 11033 + }, + { + "epoch": 4.40831002796644, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11034 + }, + { + "epoch": 4.40870954854175, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11035 + }, + { + "epoch": 4.40910906911706, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11036 + }, + { + "epoch": 4.409508589692369, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11037 + }, + { + "epoch": 4.409908110267679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11038 + }, + { + "epoch": 4.410307630842988, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11039 + }, + { + "epoch": 4.410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11040 + }, + { + "epoch": 4.411106671993608, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11041 + }, + { + "epoch": 4.411506192568917, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11042 + }, + { + "epoch": 4.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11043 + }, + { + "epoch": 4.412305233719536, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11044 + }, + { + "epoch": 4.412704754294846, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11045 + }, + { + "epoch": 4.413104274870156, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11046 + }, + { + "epoch": 4.4135037954454654, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11047 + }, + { + "epoch": 4.413903316020775, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11048 + }, + { + "epoch": 4.414302836596085, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11049 + }, + { + "epoch": 4.414702357171394, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11050 + }, + { + "epoch": 4.415101877746704, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11051 + }, + { + "epoch": 4.4155013983220135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11052 + }, + { + "epoch": 4.415900918897323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11053 + }, + { + "epoch": 4.416300439472633, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11054 + }, + { + "epoch": 4.416699960047943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11055 + }, + { + "epoch": 4.417099480623252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11056 + }, + { + "epoch": 4.4174990011985615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11057 + }, + { + "epoch": 4.417898521773871, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11058 + }, + { + "epoch": 4.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11059 + }, + { + "epoch": 4.418697562924491, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11060 + }, + { + "epoch": 4.4190970834998, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11061 + }, + { + "epoch": 4.41949660407511, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11062 + }, + { + "epoch": 4.419896124650419, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11063 + }, + { + "epoch": 4.420295645225729, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11064 + }, + { + "epoch": 4.420695165801039, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11065 + }, + { + "epoch": 4.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11066 + }, + { + "epoch": 4.421494206951658, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11067 + }, + { + "epoch": 4.421893727526967, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11068 + }, + { + "epoch": 4.422293248102277, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11069 + }, + { + "epoch": 4.422692768677587, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11070 + }, + { + "epoch": 4.423092289252897, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11071 + }, + { + "epoch": 4.4234918098282066, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11072 + }, + { + "epoch": 4.4238913304035155, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11073 + }, + { + "epoch": 4.424290850978825, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11074 + }, + { + "epoch": 4.424690371554135, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11075 + }, + { + "epoch": 4.425089892129445, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11076 + }, + { + "epoch": 4.425489412704755, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11077 + }, + { + "epoch": 4.4258889332800635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11078 + }, + { + "epoch": 4.426288453855373, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11079 + }, + { + "epoch": 4.426687974430683, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11080 + }, + { + "epoch": 4.427087495005993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11081 + }, + { + "epoch": 4.427487015581303, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11082 + }, + { + "epoch": 4.427886536156612, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11083 + }, + { + "epoch": 4.428286056731921, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11084 + }, + { + "epoch": 4.428685577307231, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11085 + }, + { + "epoch": 4.429085097882541, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11086 + }, + { + "epoch": 4.429484618457851, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11087 + }, + { + "epoch": 4.4298841390331605, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 11088 + }, + { + "epoch": 4.43028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11089 + }, + { + "epoch": 4.430683180183779, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11090 + }, + { + "epoch": 4.431082700759089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11091 + }, + { + "epoch": 4.431482221334399, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 11092 + }, + { + "epoch": 4.4318817419097085, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11093 + }, + { + "epoch": 4.432281262485018, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11094 + }, + { + "epoch": 4.432680783060327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11095 + }, + { + "epoch": 4.433080303635637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11096 + }, + { + "epoch": 4.433479824210947, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11097 + }, + { + "epoch": 4.433879344786257, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 11098 + }, + { + "epoch": 4.434278865361566, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11099 + }, + { + "epoch": 4.434678385936876, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11100 + }, + { + "epoch": 4.435077906512185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11101 + }, + { + "epoch": 4.435477427087495, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11102 + }, + { + "epoch": 4.435876947662805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11103 + }, + { + "epoch": 4.436276468238114, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11104 + }, + { + "epoch": 4.436675988813424, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11105 + }, + { + "epoch": 4.437075509388734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11106 + }, + { + "epoch": 4.437475029964043, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 11107 + }, + { + "epoch": 4.437874550539353, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 11108 + }, + { + "epoch": 4.438274071114662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11109 + }, + { + "epoch": 4.438673591689972, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11110 + }, + { + "epoch": 4.439073112265282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11111 + }, + { + "epoch": 4.439472632840591, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11112 + }, + { + "epoch": 4.439872153415901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11113 + }, + { + "epoch": 4.4402716739912105, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11114 + }, + { + "epoch": 4.44067119456652, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11115 + }, + { + "epoch": 4.44107071514183, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11116 + }, + { + "epoch": 4.44147023571714, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11117 + }, + { + "epoch": 4.441869756292449, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11118 + }, + { + "epoch": 4.4422692768677585, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 11119 + }, + { + "epoch": 4.442668797443068, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11120 + }, + { + "epoch": 4.443068318018378, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11121 + }, + { + "epoch": 4.443467838593688, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11122 + }, + { + "epoch": 4.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11123 + }, + { + "epoch": 4.444266879744307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11124 + }, + { + "epoch": 4.444666400319616, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11125 + }, + { + "epoch": 4.445065920894926, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11126 + }, + { + "epoch": 4.445465441470236, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11127 + }, + { + "epoch": 4.445864962045546, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11128 + }, + { + "epoch": 4.446264482620855, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11129 + }, + { + "epoch": 4.446664003196164, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11130 + }, + { + "epoch": 4.447063523771474, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11131 + }, + { + "epoch": 4.447463044346784, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11132 + }, + { + "epoch": 4.447862564922094, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11133 + }, + { + "epoch": 4.448262085497403, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11134 + }, + { + "epoch": 4.448661606072712, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11135 + }, + { + "epoch": 4.449061126648022, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11136 + }, + { + "epoch": 4.449460647223332, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 11137 + }, + { + "epoch": 4.449860167798642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11138 + }, + { + "epoch": 4.450259688373952, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11139 + }, + { + "epoch": 4.4506592089492605, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 11140 + }, + { + "epoch": 4.45105872952457, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11141 + }, + { + "epoch": 4.45145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11142 + }, + { + "epoch": 4.45185777067519, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11143 + }, + { + "epoch": 4.4522572912505, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11144 + }, + { + "epoch": 4.452656811825809, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11145 + }, + { + "epoch": 4.453056332401118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11146 + }, + { + "epoch": 4.453455852976428, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11147 + }, + { + "epoch": 4.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11148 + }, + { + "epoch": 4.454254894127048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11149 + }, + { + "epoch": 4.4546544147023575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11150 + }, + { + "epoch": 4.455053935277666, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11151 + }, + { + "epoch": 4.455453455852976, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11152 + }, + { + "epoch": 4.455852976428286, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11153 + }, + { + "epoch": 4.456252497003596, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11154 + }, + { + "epoch": 4.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11155 + }, + { + "epoch": 4.457051538154215, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11156 + }, + { + "epoch": 4.457451058729524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11157 + }, + { + "epoch": 4.457850579304834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11158 + }, + { + "epoch": 4.458250099880144, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11159 + }, + { + "epoch": 4.4586496204554535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11160 + }, + { + "epoch": 4.459049141030763, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11161 + }, + { + "epoch": 4.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11162 + }, + { + "epoch": 4.459848182181382, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11163 + }, + { + "epoch": 4.460247702756692, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11164 + }, + { + "epoch": 4.460647223332002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11165 + }, + { + "epoch": 4.461046743907311, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11166 + }, + { + "epoch": 4.461446264482621, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11167 + }, + { + "epoch": 4.46184578505793, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11168 + }, + { + "epoch": 4.46224530563324, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11169 + }, + { + "epoch": 4.46264482620855, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11170 + }, + { + "epoch": 4.463044346783859, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11171 + }, + { + "epoch": 4.463443867359169, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 11172 + }, + { + "epoch": 4.463843387934479, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11173 + }, + { + "epoch": 4.464242908509788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11174 + }, + { + "epoch": 4.464642429085098, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11175 + }, + { + "epoch": 4.4650419496604075, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11176 + }, + { + "epoch": 4.465441470235717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11177 + }, + { + "epoch": 4.465840990811027, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11178 + }, + { + "epoch": 4.466240511386337, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11179 + }, + { + "epoch": 4.466640031961646, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11180 + }, + { + "epoch": 4.4670395525369555, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11181 + }, + { + "epoch": 4.467439073112265, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11182 + }, + { + "epoch": 4.467838593687575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11183 + }, + { + "epoch": 4.468238114262885, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11184 + }, + { + "epoch": 4.468637634838194, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 11185 + }, + { + "epoch": 4.4690371554135035, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11186 + }, + { + "epoch": 4.469436675988813, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11187 + }, + { + "epoch": 4.469836196564123, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11188 + }, + { + "epoch": 4.470235717139433, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11189 + }, + { + "epoch": 4.470635237714743, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 11190 + }, + { + "epoch": 4.471034758290052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11191 + }, + { + "epoch": 4.471434278865361, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11192 + }, + { + "epoch": 4.471833799440671, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11193 + }, + { + "epoch": 4.472233320015981, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11194 + }, + { + "epoch": 4.472632840591291, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11195 + }, + { + "epoch": 4.4730323611666005, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11196 + }, + { + "epoch": 4.473431881741909, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11197 + }, + { + "epoch": 4.473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11198 + }, + { + "epoch": 4.474230922892529, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11199 + }, + { + "epoch": 4.474630443467839, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11200 + }, + { + "epoch": 4.475029964043149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11201 + }, + { + "epoch": 4.4754294846184575, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 11202 + }, + { + "epoch": 4.475829005193767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 11203 + }, + { + "epoch": 4.476228525769077, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11204 + }, + { + "epoch": 4.476628046344387, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 11205 + }, + { + "epoch": 4.477027566919697, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11206 + }, + { + "epoch": 4.477427087495006, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11207 + }, + { + "epoch": 4.477826608070315, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11208 + }, + { + "epoch": 4.478226128645625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11209 + }, + { + "epoch": 4.478625649220935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11210 + }, + { + "epoch": 4.479025169796245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11211 + }, + { + "epoch": 4.479424690371554, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11212 + }, + { + "epoch": 4.479824210946864, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11213 + }, + { + "epoch": 4.480223731522173, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11214 + }, + { + "epoch": 4.480623252097483, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11215 + }, + { + "epoch": 4.481022772672793, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11216 + }, + { + "epoch": 4.4814222932481025, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11217 + }, + { + "epoch": 4.481821813823412, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11218 + }, + { + "epoch": 4.482221334398721, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11219 + }, + { + "epoch": 4.482620854974031, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11220 + }, + { + "epoch": 4.483020375549341, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11221 + }, + { + "epoch": 4.4834198961246505, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11222 + }, + { + "epoch": 4.48381941669996, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11223 + }, + { + "epoch": 4.484218937275269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11224 + }, + { + "epoch": 4.484618457850579, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11225 + }, + { + "epoch": 4.485017978425889, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11226 + }, + { + "epoch": 4.485417499001199, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11227 + }, + { + "epoch": 4.485817019576508, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11228 + }, + { + "epoch": 4.486216540151818, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11229 + }, + { + "epoch": 4.486616060727127, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11230 + }, + { + "epoch": 4.487015581302437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11231 + }, + { + "epoch": 4.487415101877747, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11232 + }, + { + "epoch": 4.487814622453056, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11233 + }, + { + "epoch": 4.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11234 + }, + { + "epoch": 4.488613663603676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11235 + }, + { + "epoch": 4.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11236 + }, + { + "epoch": 4.489412704754295, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11237 + }, + { + "epoch": 4.4898122253296044, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11238 + }, + { + "epoch": 4.490211745904914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11239 + }, + { + "epoch": 4.490611266480224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11240 + }, + { + "epoch": 4.491010787055533, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11241 + }, + { + "epoch": 4.491410307630843, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 11242 + }, + { + "epoch": 4.4918098282061525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11243 + }, + { + "epoch": 4.492209348781462, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11244 + }, + { + "epoch": 4.492608869356772, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11245 + }, + { + "epoch": 4.493008389932082, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11246 + }, + { + "epoch": 4.493407910507391, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11247 + }, + { + "epoch": 4.4938074310827005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11248 + }, + { + "epoch": 4.49420695165801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11249 + }, + { + "epoch": 4.49460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11250 + }, + { + "epoch": 4.49500599280863, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11251 + }, + { + "epoch": 4.49540551338394, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11252 + }, + { + "epoch": 4.495805033959249, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11253 + }, + { + "epoch": 4.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11254 + }, + { + "epoch": 4.496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11255 + }, + { + "epoch": 4.497003595685178, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11256 + }, + { + "epoch": 4.497403116260488, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11257 + }, + { + "epoch": 4.497802636835797, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 11258 + }, + { + "epoch": 4.498202157411106, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11259 + }, + { + "epoch": 4.498601677986416, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11260 + }, + { + "epoch": 4.499001198561726, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11261 + }, + { + "epoch": 4.499400719137036, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11262 + }, + { + "epoch": 4.4998002397123456, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11263 + }, + { + "epoch": 4.5001997602876544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11264 + }, + { + "epoch": 4.500599280862964, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11265 + }, + { + "epoch": 4.500998801438274, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11266 + }, + { + "epoch": 4.501398322013584, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11267 + }, + { + "epoch": 4.501797842588894, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11268 + }, + { + "epoch": 4.502197363164203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11269 + }, + { + "epoch": 4.502596883739512, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11270 + }, + { + "epoch": 4.502996404314822, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11271 + }, + { + "epoch": 4.503395924890132, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11272 + }, + { + "epoch": 4.503795445465442, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 11273 + }, + { + "epoch": 4.504194966040751, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11274 + }, + { + "epoch": 4.50459448661606, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11275 + }, + { + "epoch": 4.50499400719137, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11276 + }, + { + "epoch": 4.50539352776668, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11277 + }, + { + "epoch": 4.50579304834199, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11278 + }, + { + "epoch": 4.5061925689172995, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11279 + }, + { + "epoch": 4.506592089492609, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11280 + }, + { + "epoch": 4.506991610067918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11281 + }, + { + "epoch": 4.507391130643228, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11282 + }, + { + "epoch": 4.507790651218538, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11283 + }, + { + "epoch": 4.5081901717938475, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11284 + }, + { + "epoch": 4.508589692369157, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11285 + }, + { + "epoch": 4.508989212944467, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11286 + }, + { + "epoch": 4.509388733519776, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11287 + }, + { + "epoch": 4.509788254095086, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11288 + }, + { + "epoch": 4.5101877746703956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11289 + }, + { + "epoch": 4.510587295245705, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11290 + }, + { + "epoch": 4.510986815821015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11291 + }, + { + "epoch": 4.511386336396324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 11292 + }, + { + "epoch": 4.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11293 + }, + { + "epoch": 4.512185377546944, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11294 + }, + { + "epoch": 4.512584898122253, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11295 + }, + { + "epoch": 4.512984418697563, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11296 + }, + { + "epoch": 4.513383939272872, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11297 + }, + { + "epoch": 4.513783459848182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11298 + }, + { + "epoch": 4.514182980423492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11299 + }, + { + "epoch": 4.514582500998801, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11300 + }, + { + "epoch": 4.514982021574111, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11301 + }, + { + "epoch": 4.515381542149421, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.2975, + "step": 11302 + }, + { + "epoch": 4.515781062724731, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11303 + }, + { + "epoch": 4.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11304 + }, + { + "epoch": 4.5165801038753495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 11305 + }, + { + "epoch": 4.516979624450659, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 11306 + }, + { + "epoch": 4.517379145025969, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11307 + }, + { + "epoch": 4.517778665601279, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 11308 + }, + { + "epoch": 4.518178186176588, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11309 + }, + { + "epoch": 4.5185777067518975, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11310 + }, + { + "epoch": 4.518977227327207, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11311 + }, + { + "epoch": 4.519376747902517, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11312 + }, + { + "epoch": 4.519776268477827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11313 + }, + { + "epoch": 4.520175789053136, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11314 + }, + { + "epoch": 4.520575309628446, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11315 + }, + { + "epoch": 4.520974830203755, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 11316 + }, + { + "epoch": 4.521374350779065, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11317 + }, + { + "epoch": 4.521773871354375, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11318 + }, + { + "epoch": 4.522173391929685, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11319 + }, + { + "epoch": 4.5225729125049945, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 11320 + }, + { + "epoch": 4.522972433080303, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11321 + }, + { + "epoch": 4.523371953655613, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11322 + }, + { + "epoch": 4.523771474230923, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11323 + }, + { + "epoch": 4.524170994806233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11324 + }, + { + "epoch": 4.5245705153815425, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11325 + }, + { + "epoch": 4.524970035956851, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11326 + }, + { + "epoch": 4.525369556532161, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11327 + }, + { + "epoch": 4.525769077107471, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11328 + }, + { + "epoch": 4.526168597682781, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11329 + }, + { + "epoch": 4.526568118258091, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11330 + }, + { + "epoch": 4.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11331 + }, + { + "epoch": 4.527367159408709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11332 + }, + { + "epoch": 4.527766679984019, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11333 + }, + { + "epoch": 4.528166200559329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11334 + }, + { + "epoch": 4.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 11335 + }, + { + "epoch": 4.528965241709948, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11336 + }, + { + "epoch": 4.529364762285257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11337 + }, + { + "epoch": 4.529764282860567, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11338 + }, + { + "epoch": 4.530163803435877, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11339 + }, + { + "epoch": 4.530563324011187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11340 + }, + { + "epoch": 4.5309628445864965, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11341 + }, + { + "epoch": 4.531362365161806, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11342 + }, + { + "epoch": 4.531761885737115, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 11343 + }, + { + "epoch": 4.532161406312425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11344 + }, + { + "epoch": 4.532560926887735, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11345 + }, + { + "epoch": 4.5329604474630445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 11346 + }, + { + "epoch": 4.533359968038354, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11347 + }, + { + "epoch": 4.533759488613663, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11348 + }, + { + "epoch": 4.534159009188973, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11349 + }, + { + "epoch": 4.534558529764283, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 11350 + }, + { + "epoch": 4.5349580503395925, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11351 + }, + { + "epoch": 4.535357570914902, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11352 + }, + { + "epoch": 4.535757091490212, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 11353 + }, + { + "epoch": 4.536156612065521, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11354 + }, + { + "epoch": 4.536556132640831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11355 + }, + { + "epoch": 4.536955653216141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11356 + }, + { + "epoch": 4.53735517379145, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11357 + }, + { + "epoch": 4.53775469436676, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11358 + }, + { + "epoch": 4.53815421494207, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11359 + }, + { + "epoch": 4.538553735517379, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11360 + }, + { + "epoch": 4.538953256092689, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11361 + }, + { + "epoch": 4.539352776667998, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11362 + }, + { + "epoch": 4.539752297243308, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11363 + }, + { + "epoch": 4.540151817818618, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11364 + }, + { + "epoch": 4.540551338393927, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11365 + }, + { + "epoch": 4.540950858969237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11366 + }, + { + "epoch": 4.5413503795445465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11367 + }, + { + "epoch": 4.541749900119856, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11368 + }, + { + "epoch": 4.542149420695166, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11369 + }, + { + "epoch": 4.542548941270476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11370 + }, + { + "epoch": 4.542948461845785, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11371 + }, + { + "epoch": 4.5433479824210945, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11372 + }, + { + "epoch": 4.543747502996404, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11373 + }, + { + "epoch": 4.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11374 + }, + { + "epoch": 4.544546544147024, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11375 + }, + { + "epoch": 4.544946064722334, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11376 + }, + { + "epoch": 4.5453455852976425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11377 + }, + { + "epoch": 4.545745105872952, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11378 + }, + { + "epoch": 4.546144626448262, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11379 + }, + { + "epoch": 4.546544147023572, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11380 + }, + { + "epoch": 4.546943667598882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11381 + }, + { + "epoch": 4.547343188174191, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11382 + }, + { + "epoch": 4.5477427087495, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11383 + }, + { + "epoch": 4.54814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11384 + }, + { + "epoch": 4.54854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11385 + }, + { + "epoch": 4.54894127047543, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11386 + }, + { + "epoch": 4.549340791050739, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11387 + }, + { + "epoch": 4.549740311626048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11388 + }, + { + "epoch": 4.550139832201358, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11389 + }, + { + "epoch": 4.550539352776668, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11390 + }, + { + "epoch": 4.550938873351978, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11391 + }, + { + "epoch": 4.551338393927288, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11392 + }, + { + "epoch": 4.551737914502597, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11393 + }, + { + "epoch": 4.552137435077906, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11394 + }, + { + "epoch": 4.552536955653216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11395 + }, + { + "epoch": 4.552936476228526, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11396 + }, + { + "epoch": 4.553335996803836, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11397 + }, + { + "epoch": 4.553735517379145, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11398 + }, + { + "epoch": 4.554135037954454, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11399 + }, + { + "epoch": 4.554534558529764, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11400 + }, + { + "epoch": 4.554934079105074, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 11401 + }, + { + "epoch": 4.555333599680384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11402 + }, + { + "epoch": 4.555733120255693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11403 + }, + { + "epoch": 4.556132640831002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11404 + }, + { + "epoch": 4.556532161406312, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11405 + }, + { + "epoch": 4.556931681981622, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11406 + }, + { + "epoch": 4.557331202556932, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11407 + }, + { + "epoch": 4.5577307231322415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11408 + }, + { + "epoch": 4.558130243707551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 11409 + }, + { + "epoch": 4.558529764282861, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11410 + }, + { + "epoch": 4.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11411 + }, + { + "epoch": 4.55932880543348, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11412 + }, + { + "epoch": 4.5597283260087895, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11413 + }, + { + "epoch": 4.560127846584099, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11414 + }, + { + "epoch": 4.560527367159409, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 11415 + }, + { + "epoch": 4.560926887734718, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11416 + }, + { + "epoch": 4.561326408310028, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11417 + }, + { + "epoch": 4.561725928885338, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11418 + }, + { + "epoch": 4.562125449460647, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11419 + }, + { + "epoch": 4.562524970035957, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11420 + }, + { + "epoch": 4.562924490611266, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11421 + }, + { + "epoch": 4.563324011186576, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11422 + }, + { + "epoch": 4.563723531761886, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11423 + }, + { + "epoch": 4.564123052337195, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11424 + }, + { + "epoch": 4.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11425 + }, + { + "epoch": 4.564922093487815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11426 + }, + { + "epoch": 4.565321614063124, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11427 + }, + { + "epoch": 4.565721134638434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11428 + }, + { + "epoch": 4.566120655213743, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11429 + }, + { + "epoch": 4.566520175789053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11430 + }, + { + "epoch": 4.566919696364363, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11431 + }, + { + "epoch": 4.567319216939673, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11432 + }, + { + "epoch": 4.567718737514982, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11433 + }, + { + "epoch": 4.5681182580902915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11434 + }, + { + "epoch": 4.568517778665601, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11435 + }, + { + "epoch": 4.568917299240911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11436 + }, + { + "epoch": 4.569316819816221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11437 + }, + { + "epoch": 4.56971634039153, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11438 + }, + { + "epoch": 4.5701158609668395, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11439 + }, + { + "epoch": 4.570515381542149, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11440 + }, + { + "epoch": 4.570914902117459, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11441 + }, + { + "epoch": 4.571314422692769, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11442 + }, + { + "epoch": 4.571713943268079, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11443 + }, + { + "epoch": 4.572113463843388, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11444 + }, + { + "epoch": 4.572512984418697, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11445 + }, + { + "epoch": 4.572912504994007, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11446 + }, + { + "epoch": 4.573312025569317, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11447 + }, + { + "epoch": 4.573711546144627, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11448 + }, + { + "epoch": 4.5741110667199365, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11449 + }, + { + "epoch": 4.574510587295245, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11450 + }, + { + "epoch": 4.574910107870555, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11451 + }, + { + "epoch": 4.575309628445865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11452 + }, + { + "epoch": 4.575709149021175, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11453 + }, + { + "epoch": 4.5761086695964845, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11454 + }, + { + "epoch": 4.5765081901717934, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11455 + }, + { + "epoch": 4.576907710747103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11456 + }, + { + "epoch": 4.577307231322413, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11457 + }, + { + "epoch": 4.577706751897723, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11458 + }, + { + "epoch": 4.578106272473033, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11459 + }, + { + "epoch": 4.578505793048342, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11460 + }, + { + "epoch": 4.578905313623651, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11461 + }, + { + "epoch": 4.579304834198961, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11462 + }, + { + "epoch": 4.579704354774271, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11463 + }, + { + "epoch": 4.580103875349581, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11464 + }, + { + "epoch": 4.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11465 + }, + { + "epoch": 4.5809029165002, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11466 + }, + { + "epoch": 4.581302437075509, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11467 + }, + { + "epoch": 4.581701957650819, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11468 + }, + { + "epoch": 4.582101478226129, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11469 + }, + { + "epoch": 4.5825009988014385, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11470 + }, + { + "epoch": 4.582900519376748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11471 + }, + { + "epoch": 4.583300039952057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 11472 + }, + { + "epoch": 4.583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11473 + }, + { + "epoch": 4.584099081102677, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11474 + }, + { + "epoch": 4.5844986016779865, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11475 + }, + { + "epoch": 4.584898122253296, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11476 + }, + { + "epoch": 4.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11477 + }, + { + "epoch": 4.585697163403915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11478 + }, + { + "epoch": 4.586096683979225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11479 + }, + { + "epoch": 4.5864962045545346, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11480 + }, + { + "epoch": 4.586895725129844, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11481 + }, + { + "epoch": 4.587295245705154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11482 + }, + { + "epoch": 4.587694766280464, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11483 + }, + { + "epoch": 4.588094286855773, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11484 + }, + { + "epoch": 4.588493807431083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 11485 + }, + { + "epoch": 4.588893328006392, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11486 + }, + { + "epoch": 4.589292848581702, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 11487 + }, + { + "epoch": 4.589692369157012, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11488 + }, + { + "epoch": 4.590091889732321, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11489 + }, + { + "epoch": 4.590491410307631, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11490 + }, + { + "epoch": 4.59089093088294, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 11491 + }, + { + "epoch": 4.59129045145825, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11492 + }, + { + "epoch": 4.59168997203356, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11493 + }, + { + "epoch": 4.592089492608869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11494 + }, + { + "epoch": 4.592489013184179, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11495 + }, + { + "epoch": 4.5928885337594885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 11496 + }, + { + "epoch": 4.593288054334798, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11497 + }, + { + "epoch": 4.593687574910108, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11498 + }, + { + "epoch": 4.594087095485418, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11499 + }, + { + "epoch": 4.594486616060728, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11500 + }, + { + "epoch": 4.5948861366360365, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11501 + }, + { + "epoch": 4.595285657211346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11502 + }, + { + "epoch": 4.595685177786656, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11503 + }, + { + "epoch": 4.596084698361966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11504 + }, + { + "epoch": 4.596484218937276, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11505 + }, + { + "epoch": 4.5968837395125846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11506 + }, + { + "epoch": 4.597283260087894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11507 + }, + { + "epoch": 4.597682780663204, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11508 + }, + { + "epoch": 4.598082301238514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11509 + }, + { + "epoch": 4.598481821813824, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11510 + }, + { + "epoch": 4.598881342389133, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11511 + }, + { + "epoch": 4.599280862964442, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11512 + }, + { + "epoch": 4.599680383539752, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11513 + }, + { + "epoch": 4.600079904115062, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11514 + }, + { + "epoch": 4.600479424690372, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11515 + }, + { + "epoch": 4.6008789452656815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11516 + }, + { + "epoch": 4.60127846584099, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11517 + }, + { + "epoch": 4.6016779864163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11518 + }, + { + "epoch": 4.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11519 + }, + { + "epoch": 4.60247702756692, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11520 + }, + { + "epoch": 4.60287654814223, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11521 + }, + { + "epoch": 4.603276068717539, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11522 + }, + { + "epoch": 4.603675589292848, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11523 + }, + { + "epoch": 4.604075109868158, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11524 + }, + { + "epoch": 4.604474630443468, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11525 + }, + { + "epoch": 4.604874151018778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11526 + }, + { + "epoch": 4.605273671594087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11527 + }, + { + "epoch": 4.605673192169396, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11528 + }, + { + "epoch": 4.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11529 + }, + { + "epoch": 4.606472233320016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11530 + }, + { + "epoch": 4.606871753895326, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11531 + }, + { + "epoch": 4.6072712744706354, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 11532 + }, + { + "epoch": 4.607670795045945, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11533 + }, + { + "epoch": 4.608070315621254, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11534 + }, + { + "epoch": 4.608469836196564, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11535 + }, + { + "epoch": 4.608869356771874, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11536 + }, + { + "epoch": 4.6092688773471835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11537 + }, + { + "epoch": 4.609668397922493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11538 + }, + { + "epoch": 4.610067918497803, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11539 + }, + { + "epoch": 4.610467439073112, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11540 + }, + { + "epoch": 4.610866959648422, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11541 + }, + { + "epoch": 4.6112664802237315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11542 + }, + { + "epoch": 4.611666000799041, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11543 + }, + { + "epoch": 4.612065521374351, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11544 + }, + { + "epoch": 4.61246504194966, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11545 + }, + { + "epoch": 4.61286456252497, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11546 + }, + { + "epoch": 4.61326408310028, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11547 + }, + { + "epoch": 4.613663603675589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11548 + }, + { + "epoch": 4.614063124250899, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11549 + }, + { + "epoch": 4.614462644826209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11550 + }, + { + "epoch": 4.614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11551 + }, + { + "epoch": 4.615261685976828, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 11552 + }, + { + "epoch": 4.615661206552137, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11553 + }, + { + "epoch": 4.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11554 + }, + { + "epoch": 4.616460247702757, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11555 + }, + { + "epoch": 4.616859768278067, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11556 + }, + { + "epoch": 4.617259288853376, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11557 + }, + { + "epoch": 4.6176588094286855, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11558 + }, + { + "epoch": 4.618058330003995, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11559 + }, + { + "epoch": 4.618457850579305, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11560 + }, + { + "epoch": 4.618857371154615, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11561 + }, + { + "epoch": 4.619256891729924, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11562 + }, + { + "epoch": 4.6196564123052335, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11563 + }, + { + "epoch": 4.620055932880543, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11564 + }, + { + "epoch": 4.620455453455853, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 11565 + }, + { + "epoch": 4.620854974031163, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11566 + }, + { + "epoch": 4.621254494606472, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11567 + }, + { + "epoch": 4.6216540151817815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11568 + }, + { + "epoch": 4.622053535757091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11569 + }, + { + "epoch": 4.622453056332401, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11570 + }, + { + "epoch": 4.622852576907711, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11571 + }, + { + "epoch": 4.623252097483021, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 11572 + }, + { + "epoch": 4.6236516180583305, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11573 + }, + { + "epoch": 4.624051138633639, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11574 + }, + { + "epoch": 4.624450659208949, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11575 + }, + { + "epoch": 4.624850179784259, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11576 + }, + { + "epoch": 4.625249700359569, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11577 + }, + { + "epoch": 4.6256492209348785, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11578 + }, + { + "epoch": 4.626048741510187, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11579 + }, + { + "epoch": 4.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11580 + }, + { + "epoch": 4.626847782660807, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11581 + }, + { + "epoch": 4.627247303236117, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11582 + }, + { + "epoch": 4.627646823811427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 11583 + }, + { + "epoch": 4.6280463443867355, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11584 + }, + { + "epoch": 4.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11585 + }, + { + "epoch": 4.628845385537355, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11586 + }, + { + "epoch": 4.629244906112665, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11587 + }, + { + "epoch": 4.629644426687975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11588 + }, + { + "epoch": 4.630043947263284, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11589 + }, + { + "epoch": 4.630443467838594, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11590 + }, + { + "epoch": 4.630842988413903, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 11591 + }, + { + "epoch": 4.631242508989213, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11592 + }, + { + "epoch": 4.631642029564523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11593 + }, + { + "epoch": 4.632041550139832, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11594 + }, + { + "epoch": 4.632441070715142, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 11595 + }, + { + "epoch": 4.632840591290451, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11596 + }, + { + "epoch": 4.633240111865761, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11597 + }, + { + "epoch": 4.633639632441071, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11598 + }, + { + "epoch": 4.6340391530163805, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11599 + }, + { + "epoch": 4.63443867359169, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11600 + }, + { + "epoch": 4.634838194166999, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11601 + }, + { + "epoch": 4.635237714742309, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11602 + }, + { + "epoch": 4.635637235317619, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11603 + }, + { + "epoch": 4.6360367558929285, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11604 + }, + { + "epoch": 4.636436276468238, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11605 + }, + { + "epoch": 4.636835797043548, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11606 + }, + { + "epoch": 4.637235317618857, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11607 + }, + { + "epoch": 4.637634838194167, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11608 + }, + { + "epoch": 4.638034358769477, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11609 + }, + { + "epoch": 4.638433879344786, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 11610 + }, + { + "epoch": 4.638833399920096, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11611 + }, + { + "epoch": 4.639232920495406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11612 + }, + { + "epoch": 4.639632441070715, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11613 + }, + { + "epoch": 4.640031961646025, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11614 + }, + { + "epoch": 4.640431482221334, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11615 + }, + { + "epoch": 4.640831002796644, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11616 + }, + { + "epoch": 4.641230523371954, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11617 + }, + { + "epoch": 4.641630043947263, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11618 + }, + { + "epoch": 4.642029564522573, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11619 + }, + { + "epoch": 4.642429085097882, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11620 + }, + { + "epoch": 4.642828605673192, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11621 + }, + { + "epoch": 4.643228126248502, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11622 + }, + { + "epoch": 4.643627646823812, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11623 + }, + { + "epoch": 4.644027167399121, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11624 + }, + { + "epoch": 4.6444266879744305, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11625 + }, + { + "epoch": 4.64482620854974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11626 + }, + { + "epoch": 4.64522572912505, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11627 + }, + { + "epoch": 4.64562524970036, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11628 + }, + { + "epoch": 4.64602477027567, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11629 + }, + { + "epoch": 4.6464242908509785, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11630 + }, + { + "epoch": 4.646823811426288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11631 + }, + { + "epoch": 4.647223332001598, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11632 + }, + { + "epoch": 4.647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11633 + }, + { + "epoch": 4.648022373152218, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11634 + }, + { + "epoch": 4.648421893727527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11635 + }, + { + "epoch": 4.648821414302836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11636 + }, + { + "epoch": 4.649220934878146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11637 + }, + { + "epoch": 4.649620455453456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11638 + }, + { + "epoch": 4.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11639 + }, + { + "epoch": 4.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11640 + }, + { + "epoch": 4.650819017179384, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11641 + }, + { + "epoch": 4.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 11642 + }, + { + "epoch": 4.651618058330004, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11643 + }, + { + "epoch": 4.652017578905314, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11644 + }, + { + "epoch": 4.6524170994806235, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11645 + }, + { + "epoch": 4.652816620055933, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11646 + }, + { + "epoch": 4.653216140631242, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11647 + }, + { + "epoch": 4.653615661206552, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11648 + }, + { + "epoch": 4.654015181781862, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11649 + }, + { + "epoch": 4.654414702357172, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11650 + }, + { + "epoch": 4.654814222932481, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11651 + }, + { + "epoch": 4.65521374350779, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11652 + }, + { + "epoch": 4.6556132640831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11653 + }, + { + "epoch": 4.65601278465841, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11654 + }, + { + "epoch": 4.65641230523372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11655 + }, + { + "epoch": 4.656811825809029, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11656 + }, + { + "epoch": 4.657211346384338, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11657 + }, + { + "epoch": 4.657610866959648, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11658 + }, + { + "epoch": 4.658010387534958, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11659 + }, + { + "epoch": 4.658409908110268, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11660 + }, + { + "epoch": 4.6588094286855775, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11661 + }, + { + "epoch": 4.659208949260887, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11662 + }, + { + "epoch": 4.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11663 + }, + { + "epoch": 4.660007990411506, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11664 + }, + { + "epoch": 4.660407510986816, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11665 + }, + { + "epoch": 4.6608070315621255, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11666 + }, + { + "epoch": 4.661206552137435, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11667 + }, + { + "epoch": 4.661606072712745, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11668 + }, + { + "epoch": 4.662005593288054, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11669 + }, + { + "epoch": 4.662405113863364, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11670 + }, + { + "epoch": 4.6628046344386735, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11671 + }, + { + "epoch": 4.663204155013983, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11672 + }, + { + "epoch": 4.663603675589293, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11673 + }, + { + "epoch": 4.664003196164602, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11674 + }, + { + "epoch": 4.664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11675 + }, + { + "epoch": 4.664802237315222, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11676 + }, + { + "epoch": 4.665201757890531, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11677 + }, + { + "epoch": 4.665601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11678 + }, + { + "epoch": 4.666000799041151, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11679 + }, + { + "epoch": 4.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11680 + }, + { + "epoch": 4.66679984019177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11681 + }, + { + "epoch": 4.667199360767079, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11682 + }, + { + "epoch": 4.667598881342389, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11683 + }, + { + "epoch": 4.667998401917699, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11684 + }, + { + "epoch": 4.668397922493009, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11685 + }, + { + "epoch": 4.668797443068318, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11686 + }, + { + "epoch": 4.6691969636436275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11687 + }, + { + "epoch": 4.669596484218937, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11688 + }, + { + "epoch": 4.669996004794247, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11689 + }, + { + "epoch": 4.670395525369557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11690 + }, + { + "epoch": 4.670795045944866, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11691 + }, + { + "epoch": 4.6711945665201755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11692 + }, + { + "epoch": 4.671594087095485, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11693 + }, + { + "epoch": 4.671993607670795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11694 + }, + { + "epoch": 4.672393128246105, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11695 + }, + { + "epoch": 4.672792648821415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11696 + }, + { + "epoch": 4.6731921693967235, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11697 + }, + { + "epoch": 4.673591689972033, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11698 + }, + { + "epoch": 4.673991210547343, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11699 + }, + { + "epoch": 4.674390731122653, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 11700 + }, + { + "epoch": 4.674790251697963, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11701 + }, + { + "epoch": 4.6751897722732725, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11702 + }, + { + "epoch": 4.675589292848581, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11703 + }, + { + "epoch": 4.675988813423891, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11704 + }, + { + "epoch": 4.676388333999201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11705 + }, + { + "epoch": 4.676787854574511, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 11706 + }, + { + "epoch": 4.6771873751498205, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11707 + }, + { + "epoch": 4.677586895725129, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11708 + }, + { + "epoch": 4.677986416300439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11709 + }, + { + "epoch": 4.678385936875749, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11710 + }, + { + "epoch": 4.678785457451059, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11711 + }, + { + "epoch": 4.679184978026369, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11712 + }, + { + "epoch": 4.679584498601678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11713 + }, + { + "epoch": 4.679984019176987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11714 + }, + { + "epoch": 4.680383539752297, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11715 + }, + { + "epoch": 4.680783060327607, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11716 + }, + { + "epoch": 4.681182580902917, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11717 + }, + { + "epoch": 4.681582101478226, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11718 + }, + { + "epoch": 4.681981622053536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11719 + }, + { + "epoch": 4.682381142628845, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11720 + }, + { + "epoch": 4.682780663204155, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11721 + }, + { + "epoch": 4.683180183779465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11722 + }, + { + "epoch": 4.683579704354774, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 11723 + }, + { + "epoch": 4.683979224930084, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11724 + }, + { + "epoch": 4.684378745505393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11725 + }, + { + "epoch": 4.684778266080703, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11726 + }, + { + "epoch": 4.685177786656013, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11727 + }, + { + "epoch": 4.6855773072313225, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11728 + }, + { + "epoch": 4.685976827806632, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11729 + }, + { + "epoch": 4.686376348381942, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11730 + }, + { + "epoch": 4.686775868957251, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11731 + }, + { + "epoch": 4.687175389532561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11732 + }, + { + "epoch": 4.6875749101078705, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11733 + }, + { + "epoch": 4.68797443068318, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11734 + }, + { + "epoch": 4.68837395125849, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11735 + }, + { + "epoch": 4.6887734718338, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11736 + }, + { + "epoch": 4.689172992409109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11737 + }, + { + "epoch": 4.689572512984419, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11738 + }, + { + "epoch": 4.689972033559728, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11739 + }, + { + "epoch": 4.690371554135038, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 11740 + }, + { + "epoch": 4.690771074710348, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11741 + }, + { + "epoch": 4.691170595285657, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 11742 + }, + { + "epoch": 4.691570115860967, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11743 + }, + { + "epoch": 4.691969636436276, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11744 + }, + { + "epoch": 4.692369157011586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11745 + }, + { + "epoch": 4.692768677586896, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11746 + }, + { + "epoch": 4.693168198162205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11747 + }, + { + "epoch": 4.693567718737515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11748 + }, + { + "epoch": 4.6939672393128244, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 11749 + }, + { + "epoch": 4.694366759888134, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11750 + }, + { + "epoch": 4.694766280463444, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11751 + }, + { + "epoch": 4.695165801038754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11752 + }, + { + "epoch": 4.695565321614064, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11753 + }, + { + "epoch": 4.6959648421893725, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11754 + }, + { + "epoch": 4.696364362764682, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11755 + }, + { + "epoch": 4.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11756 + }, + { + "epoch": 4.697163403915302, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11757 + }, + { + "epoch": 4.697562924490612, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11758 + }, + { + "epoch": 4.6979624450659205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11759 + }, + { + "epoch": 4.69836196564123, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11760 + }, + { + "epoch": 4.69876148621654, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11761 + }, + { + "epoch": 4.69916100679185, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 11762 + }, + { + "epoch": 4.69956052736716, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11763 + }, + { + "epoch": 4.699960047942469, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11764 + }, + { + "epoch": 4.700359568517778, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11765 + }, + { + "epoch": 4.700759089093088, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11766 + }, + { + "epoch": 4.701158609668398, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11767 + }, + { + "epoch": 4.701558130243708, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11768 + }, + { + "epoch": 4.7019576508190175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11769 + }, + { + "epoch": 4.702357171394327, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11770 + }, + { + "epoch": 4.702756691969636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11771 + }, + { + "epoch": 4.703156212544946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11772 + }, + { + "epoch": 4.703555733120256, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 11773 + }, + { + "epoch": 4.7039552536955656, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11774 + }, + { + "epoch": 4.704354774270875, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11775 + }, + { + "epoch": 4.704754294846184, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11776 + }, + { + "epoch": 4.705153815421494, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11777 + }, + { + "epoch": 4.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11778 + }, + { + "epoch": 4.705952856572114, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11779 + }, + { + "epoch": 4.706352377147423, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11780 + }, + { + "epoch": 4.706751897722732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11781 + }, + { + "epoch": 4.707151418298042, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 11782 + }, + { + "epoch": 4.707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11783 + }, + { + "epoch": 4.707950459448662, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11784 + }, + { + "epoch": 4.708349980023971, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11785 + }, + { + "epoch": 4.708749500599281, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11786 + }, + { + "epoch": 4.70914902117459, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11787 + }, + { + "epoch": 4.7095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11788 + }, + { + "epoch": 4.70994806232521, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11789 + }, + { + "epoch": 4.7103475829005195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11790 + }, + { + "epoch": 4.710747103475829, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11791 + }, + { + "epoch": 4.711146624051139, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11792 + }, + { + "epoch": 4.711546144626448, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11793 + }, + { + "epoch": 4.711945665201758, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11794 + }, + { + "epoch": 4.7123451857770675, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 11795 + }, + { + "epoch": 4.712744706352377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11796 + }, + { + "epoch": 4.713144226927687, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11797 + }, + { + "epoch": 4.713543747502996, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11798 + }, + { + "epoch": 4.713943268078306, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11799 + }, + { + "epoch": 4.714342788653616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11800 + }, + { + "epoch": 4.714742309228925, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 11801 + }, + { + "epoch": 4.715141829804235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11802 + }, + { + "epoch": 4.715541350379545, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11803 + }, + { + "epoch": 4.715940870954854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11804 + }, + { + "epoch": 4.716340391530164, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11805 + }, + { + "epoch": 4.716739912105473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11806 + }, + { + "epoch": 4.717139432680783, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11807 + }, + { + "epoch": 4.717538953256093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11808 + }, + { + "epoch": 4.717938473831403, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11809 + }, + { + "epoch": 4.718337994406712, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11810 + }, + { + "epoch": 4.718737514982021, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11811 + }, + { + "epoch": 4.719137035557331, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11812 + }, + { + "epoch": 4.719536556132641, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11813 + }, + { + "epoch": 4.719936076707951, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11814 + }, + { + "epoch": 4.72033559728326, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11815 + }, + { + "epoch": 4.7207351178585695, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11816 + }, + { + "epoch": 4.721134638433879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11817 + }, + { + "epoch": 4.721534159009189, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11818 + }, + { + "epoch": 4.721933679584499, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11819 + }, + { + "epoch": 4.722333200159808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11820 + }, + { + "epoch": 4.7227327207351175, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11821 + }, + { + "epoch": 4.723132241310427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 11822 + }, + { + "epoch": 4.723531761885737, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11823 + }, + { + "epoch": 4.723931282461047, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11824 + }, + { + "epoch": 4.724330803036357, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11825 + }, + { + "epoch": 4.7247303236116664, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11826 + }, + { + "epoch": 4.725129844186975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11827 + }, + { + "epoch": 4.725529364762285, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11828 + }, + { + "epoch": 4.725928885337595, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11829 + }, + { + "epoch": 4.726328405912905, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11830 + }, + { + "epoch": 4.7267279264882145, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11831 + }, + { + "epoch": 4.727127447063523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11832 + }, + { + "epoch": 4.727526967638833, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11833 + }, + { + "epoch": 4.727926488214143, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11834 + }, + { + "epoch": 4.728326008789453, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11835 + }, + { + "epoch": 4.7287255293647625, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 11836 + }, + { + "epoch": 4.729125049940071, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11837 + }, + { + "epoch": 4.729524570515381, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11838 + }, + { + "epoch": 4.729924091090691, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11839 + }, + { + "epoch": 4.730323611666001, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11840 + }, + { + "epoch": 4.730723132241311, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11841 + }, + { + "epoch": 4.73112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11842 + }, + { + "epoch": 4.73152217339193, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11843 + }, + { + "epoch": 4.731921693967239, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11844 + }, + { + "epoch": 4.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11845 + }, + { + "epoch": 4.732720735117859, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11846 + }, + { + "epoch": 4.733120255693168, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11847 + }, + { + "epoch": 4.733519776268478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11848 + }, + { + "epoch": 4.733919296843787, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11849 + }, + { + "epoch": 4.734318817419097, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11850 + }, + { + "epoch": 4.734718337994407, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 11851 + }, + { + "epoch": 4.7351178585697165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11852 + }, + { + "epoch": 4.735517379145026, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11853 + }, + { + "epoch": 4.735916899720335, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11854 + }, + { + "epoch": 4.736316420295645, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11855 + }, + { + "epoch": 4.736715940870955, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11856 + }, + { + "epoch": 4.7371154614462645, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11857 + }, + { + "epoch": 4.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11858 + }, + { + "epoch": 4.737914502596884, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 11859 + }, + { + "epoch": 4.738314023172194, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11860 + }, + { + "epoch": 4.738713543747503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11861 + }, + { + "epoch": 4.7391130643228125, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11862 + }, + { + "epoch": 4.739512584898122, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11863 + }, + { + "epoch": 4.739912105473432, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11864 + }, + { + "epoch": 4.740311626048742, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11865 + }, + { + "epoch": 4.740711146624051, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11866 + }, + { + "epoch": 4.741110667199361, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11867 + }, + { + "epoch": 4.74151018777467, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11868 + }, + { + "epoch": 4.74190970834998, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11869 + }, + { + "epoch": 4.74230922892529, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11870 + }, + { + "epoch": 4.742708749500599, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11871 + }, + { + "epoch": 4.743108270075909, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11872 + }, + { + "epoch": 4.743507790651218, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11873 + }, + { + "epoch": 4.743907311226528, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11874 + }, + { + "epoch": 4.744306831801838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11875 + }, + { + "epoch": 4.744706352377148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11876 + }, + { + "epoch": 4.745105872952457, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11877 + }, + { + "epoch": 4.7455053935277665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11878 + }, + { + "epoch": 4.745904914103076, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11879 + }, + { + "epoch": 4.746304434678386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11880 + }, + { + "epoch": 4.746703955253696, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11881 + }, + { + "epoch": 4.747103475829006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11882 + }, + { + "epoch": 4.7475029964043145, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11883 + }, + { + "epoch": 4.747902516979624, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11884 + }, + { + "epoch": 4.748302037554934, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11885 + }, + { + "epoch": 4.748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11886 + }, + { + "epoch": 4.749101078705554, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11887 + }, + { + "epoch": 4.7495005992808625, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11888 + }, + { + "epoch": 4.749900119856172, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11889 + }, + { + "epoch": 4.750299640431482, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11890 + }, + { + "epoch": 4.750699161006792, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11891 + }, + { + "epoch": 4.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11892 + }, + { + "epoch": 4.7514982021574115, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11893 + }, + { + "epoch": 4.75189772273272, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11894 + }, + { + "epoch": 4.75229724330803, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11895 + }, + { + "epoch": 4.75269676388334, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11896 + }, + { + "epoch": 4.75309628445865, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11897 + }, + { + "epoch": 4.7534958050339595, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11898 + }, + { + "epoch": 4.753895325609269, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11899 + }, + { + "epoch": 4.754294846184578, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11900 + }, + { + "epoch": 4.754694366759888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11901 + }, + { + "epoch": 4.755093887335198, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11902 + }, + { + "epoch": 4.755493407910508, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11903 + }, + { + "epoch": 4.755892928485817, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11904 + }, + { + "epoch": 4.756292449061126, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11905 + }, + { + "epoch": 4.756691969636436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11906 + }, + { + "epoch": 4.757091490211746, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11907 + }, + { + "epoch": 4.757491010787056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11908 + }, + { + "epoch": 4.757890531362365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11909 + }, + { + "epoch": 4.758290051937674, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11910 + }, + { + "epoch": 4.758689572512984, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11911 + }, + { + "epoch": 4.759089093088294, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11912 + }, + { + "epoch": 4.759488613663604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11913 + }, + { + "epoch": 4.759888134238913, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11914 + }, + { + "epoch": 4.760287654814223, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11915 + }, + { + "epoch": 4.760687175389533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11916 + }, + { + "epoch": 4.761086695964842, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11917 + }, + { + "epoch": 4.761486216540152, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11918 + }, + { + "epoch": 4.7618857371154615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11919 + }, + { + "epoch": 4.762285257690771, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 11920 + }, + { + "epoch": 4.762684778266081, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11921 + }, + { + "epoch": 4.76308429884139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11922 + }, + { + "epoch": 4.7634838194167, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 11923 + }, + { + "epoch": 4.7638833399920095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11924 + }, + { + "epoch": 4.764282860567319, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11925 + }, + { + "epoch": 4.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11926 + }, + { + "epoch": 4.765081901717938, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11927 + }, + { + "epoch": 4.765481422293248, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11928 + }, + { + "epoch": 4.765880942868558, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11929 + }, + { + "epoch": 4.766280463443867, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11930 + }, + { + "epoch": 4.766679984019177, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11931 + }, + { + "epoch": 4.767079504594487, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11932 + }, + { + "epoch": 4.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11933 + }, + { + "epoch": 4.767878545745106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11934 + }, + { + "epoch": 4.768278066320415, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11935 + }, + { + "epoch": 4.768677586895725, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11936 + }, + { + "epoch": 4.769077107471035, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11937 + }, + { + "epoch": 4.769476628046345, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11938 + }, + { + "epoch": 4.769876148621654, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11939 + }, + { + "epoch": 4.770275669196963, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11940 + }, + { + "epoch": 4.770675189772273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11941 + }, + { + "epoch": 4.771074710347583, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11942 + }, + { + "epoch": 4.771474230922893, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11943 + }, + { + "epoch": 4.771873751498202, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11944 + }, + { + "epoch": 4.7722732720735115, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11945 + }, + { + "epoch": 4.772672792648821, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11946 + }, + { + "epoch": 4.773072313224131, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11947 + }, + { + "epoch": 4.773471833799441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11948 + }, + { + "epoch": 4.773871354374751, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11949 + }, + { + "epoch": 4.77427087495006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11950 + }, + { + "epoch": 4.774670395525369, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11951 + }, + { + "epoch": 4.775069916100679, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 11952 + }, + { + "epoch": 4.775469436675989, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11953 + }, + { + "epoch": 4.775868957251299, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11954 + }, + { + "epoch": 4.7762684778266085, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11955 + }, + { + "epoch": 4.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11956 + }, + { + "epoch": 4.777067518977227, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11957 + }, + { + "epoch": 4.777467039552537, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11958 + }, + { + "epoch": 4.777866560127847, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11959 + }, + { + "epoch": 4.7782660807031565, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11960 + }, + { + "epoch": 4.778665601278465, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11961 + }, + { + "epoch": 4.779065121853775, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11962 + }, + { + "epoch": 4.779464642429085, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11963 + }, + { + "epoch": 4.779864163004395, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11964 + }, + { + "epoch": 4.7802636835797045, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11965 + }, + { + "epoch": 4.780663204155014, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11966 + }, + { + "epoch": 4.781062724730323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 11967 + }, + { + "epoch": 4.781462245305633, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11968 + }, + { + "epoch": 4.781861765880943, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11969 + }, + { + "epoch": 4.782261286456253, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11970 + }, + { + "epoch": 4.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11971 + }, + { + "epoch": 4.783060327606872, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11972 + }, + { + "epoch": 4.783459848182181, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11973 + }, + { + "epoch": 4.783859368757491, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11974 + }, + { + "epoch": 4.784258889332801, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 11975 + }, + { + "epoch": 4.78465840990811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11976 + }, + { + "epoch": 4.78505793048342, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11977 + }, + { + "epoch": 4.785457451058729, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11978 + }, + { + "epoch": 4.785856971634039, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11979 + }, + { + "epoch": 4.786256492209349, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11980 + }, + { + "epoch": 4.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11981 + }, + { + "epoch": 4.787055533359968, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11982 + }, + { + "epoch": 4.787455053935278, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11983 + }, + { + "epoch": 4.787854574510587, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11984 + }, + { + "epoch": 4.788254095085897, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11985 + }, + { + "epoch": 4.7886536156612065, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11986 + }, + { + "epoch": 4.789053136236516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11987 + }, + { + "epoch": 4.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11988 + }, + { + "epoch": 4.789852177387136, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11989 + }, + { + "epoch": 4.790251697962445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11990 + }, + { + "epoch": 4.7906512185377546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11991 + }, + { + "epoch": 4.791050739113064, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11992 + }, + { + "epoch": 4.791450259688374, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11993 + }, + { + "epoch": 4.791849780263684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11994 + }, + { + "epoch": 4.792249300838993, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11995 + }, + { + "epoch": 4.792648821414303, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11996 + }, + { + "epoch": 4.793048341989612, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11997 + }, + { + "epoch": 4.793447862564922, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11998 + }, + { + "epoch": 4.793847383140232, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 11999 + }, + { + "epoch": 4.794246903715541, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12000 + }, + { + "epoch": 4.794646424290851, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12001 + }, + { + "epoch": 4.79504594486616, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12002 + }, + { + "epoch": 4.79544546544147, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12003 + }, + { + "epoch": 4.79584498601678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12004 + }, + { + "epoch": 4.79624450659209, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12005 + }, + { + "epoch": 4.7966440271674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12006 + }, + { + "epoch": 4.7970435477427085, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12007 + }, + { + "epoch": 4.797443068318018, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12008 + }, + { + "epoch": 4.797842588893328, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12009 + }, + { + "epoch": 4.798242109468638, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12010 + }, + { + "epoch": 4.798641630043948, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12011 + }, + { + "epoch": 4.7990411506192565, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12012 + }, + { + "epoch": 4.799440671194566, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 12013 + }, + { + "epoch": 4.799840191769876, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12014 + }, + { + "epoch": 4.800239712345186, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12015 + }, + { + "epoch": 4.800639232920496, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12016 + }, + { + "epoch": 4.801038753495805, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12017 + }, + { + "epoch": 4.801438274071114, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12018 + }, + { + "epoch": 4.801837794646424, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12019 + }, + { + "epoch": 4.802237315221734, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12020 + }, + { + "epoch": 4.802636835797044, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12021 + }, + { + "epoch": 4.8030363563723535, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12022 + }, + { + "epoch": 4.803435876947663, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12023 + }, + { + "epoch": 4.803835397522972, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12024 + }, + { + "epoch": 4.804234918098282, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12025 + }, + { + "epoch": 4.804634438673592, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12026 + }, + { + "epoch": 4.8050339592489015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12027 + }, + { + "epoch": 4.805433479824211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12028 + }, + { + "epoch": 4.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12029 + }, + { + "epoch": 4.80623252097483, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12030 + }, + { + "epoch": 4.80663204155014, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12031 + }, + { + "epoch": 4.80703156212545, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12032 + }, + { + "epoch": 4.807431082700759, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12033 + }, + { + "epoch": 4.807830603276068, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12034 + }, + { + "epoch": 4.808230123851378, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12035 + }, + { + "epoch": 4.808629644426688, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12036 + }, + { + "epoch": 4.809029165001998, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12037 + }, + { + "epoch": 4.809428685577307, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12038 + }, + { + "epoch": 4.809828206152617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12039 + }, + { + "epoch": 4.810227726727927, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12040 + }, + { + "epoch": 4.810627247303236, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12041 + }, + { + "epoch": 4.811026767878546, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12042 + }, + { + "epoch": 4.8114262884538554, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12043 + }, + { + "epoch": 4.811825809029165, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12044 + }, + { + "epoch": 4.812225329604475, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 12045 + }, + { + "epoch": 4.812624850179784, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12046 + }, + { + "epoch": 4.813024370755094, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12047 + }, + { + "epoch": 4.8134238913304035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12048 + }, + { + "epoch": 4.813823411905713, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12049 + }, + { + "epoch": 4.814222932481023, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12050 + }, + { + "epoch": 4.814622453056332, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12051 + }, + { + "epoch": 4.815021973631642, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12052 + }, + { + "epoch": 4.8154214942069515, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12053 + }, + { + "epoch": 4.815821014782261, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12054 + }, + { + "epoch": 4.816220535357571, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12055 + }, + { + "epoch": 4.816620055932881, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12056 + }, + { + "epoch": 4.81701957650819, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12057 + }, + { + "epoch": 4.8174190970835, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12058 + }, + { + "epoch": 4.817818617658809, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12059 + }, + { + "epoch": 4.818218138234119, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12060 + }, + { + "epoch": 4.818617658809429, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12061 + }, + { + "epoch": 4.819017179384739, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12062 + }, + { + "epoch": 4.819416699960048, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12063 + }, + { + "epoch": 4.819816220535357, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12064 + }, + { + "epoch": 4.820215741110667, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12065 + }, + { + "epoch": 4.820615261685977, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12066 + }, + { + "epoch": 4.821014782261287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12067 + }, + { + "epoch": 4.821414302836596, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12068 + }, + { + "epoch": 4.8218138234119055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12069 + }, + { + "epoch": 4.822213343987215, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12070 + }, + { + "epoch": 4.822612864562525, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12071 + }, + { + "epoch": 4.823012385137835, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12072 + }, + { + "epoch": 4.823411905713145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12073 + }, + { + "epoch": 4.8238114262884535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12074 + }, + { + "epoch": 4.824210946863763, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12075 + }, + { + "epoch": 4.824610467439073, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12076 + }, + { + "epoch": 4.825009988014383, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12077 + }, + { + "epoch": 4.825409508589693, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 12078 + }, + { + "epoch": 4.825809029165002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12079 + }, + { + "epoch": 4.826208549740311, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12080 + }, + { + "epoch": 4.826608070315621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12081 + }, + { + "epoch": 4.827007590890931, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12082 + }, + { + "epoch": 4.827407111466241, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12083 + }, + { + "epoch": 4.8278066320415505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12084 + }, + { + "epoch": 4.828206152616859, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12085 + }, + { + "epoch": 4.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12086 + }, + { + "epoch": 4.829005193767479, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12087 + }, + { + "epoch": 4.829404714342789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12088 + }, + { + "epoch": 4.8298042349180985, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12089 + }, + { + "epoch": 4.830203755493407, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12090 + }, + { + "epoch": 4.830603276068717, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12091 + }, + { + "epoch": 4.831002796644027, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12092 + }, + { + "epoch": 4.831402317219337, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12093 + }, + { + "epoch": 4.831801837794647, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 12094 + }, + { + "epoch": 4.832201358369956, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12095 + }, + { + "epoch": 4.832600878945266, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12096 + }, + { + "epoch": 4.833000399520575, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12097 + }, + { + "epoch": 4.833399920095885, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12098 + }, + { + "epoch": 4.833799440671195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12099 + }, + { + "epoch": 4.834198961246504, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12100 + }, + { + "epoch": 4.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12101 + }, + { + "epoch": 4.834998002397123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12102 + }, + { + "epoch": 4.835397522972433, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12103 + }, + { + "epoch": 4.835797043547743, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12104 + }, + { + "epoch": 4.836196564123052, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12105 + }, + { + "epoch": 4.836596084698362, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12106 + }, + { + "epoch": 4.836995605273671, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12107 + }, + { + "epoch": 4.837395125848981, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12108 + }, + { + "epoch": 4.837794646424291, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12109 + }, + { + "epoch": 4.8381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12110 + }, + { + "epoch": 4.83859368757491, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12111 + }, + { + "epoch": 4.83899320815022, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12112 + }, + { + "epoch": 4.83939272872553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12113 + }, + { + "epoch": 4.839792249300839, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12114 + }, + { + "epoch": 4.8401917698761485, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12115 + }, + { + "epoch": 4.840591290451458, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12116 + }, + { + "epoch": 4.840990811026768, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12117 + }, + { + "epoch": 4.841390331602078, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12118 + }, + { + "epoch": 4.841789852177387, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12119 + }, + { + "epoch": 4.842189372752697, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12120 + }, + { + "epoch": 4.842588893328006, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12121 + }, + { + "epoch": 4.842988413903316, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12122 + }, + { + "epoch": 4.843387934478626, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 12123 + }, + { + "epoch": 4.843787455053935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12124 + }, + { + "epoch": 4.844186975629245, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12125 + }, + { + "epoch": 4.844586496204554, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12126 + }, + { + "epoch": 4.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12127 + }, + { + "epoch": 4.845385537355174, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12128 + }, + { + "epoch": 4.845785057930484, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12129 + }, + { + "epoch": 4.8461845785057935, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 12130 + }, + { + "epoch": 4.846584099081102, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12131 + }, + { + "epoch": 4.846983619656412, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12132 + }, + { + "epoch": 4.847383140231722, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12133 + }, + { + "epoch": 4.847782660807032, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12134 + }, + { + "epoch": 4.848182181382342, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12135 + }, + { + "epoch": 4.8485817019576505, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12136 + }, + { + "epoch": 4.84898122253296, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 12137 + }, + { + "epoch": 4.84938074310827, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12138 + }, + { + "epoch": 4.84978026368358, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12139 + }, + { + "epoch": 4.85017978425889, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12140 + }, + { + "epoch": 4.8505793048341985, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12141 + }, + { + "epoch": 4.850978825409508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12142 + }, + { + "epoch": 4.851378345984818, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12143 + }, + { + "epoch": 4.851777866560128, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12144 + }, + { + "epoch": 4.852177387135438, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12145 + }, + { + "epoch": 4.8525769077107475, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12146 + }, + { + "epoch": 4.852976428286056, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12147 + }, + { + "epoch": 4.853375948861366, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12148 + }, + { + "epoch": 4.853775469436676, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12149 + }, + { + "epoch": 4.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12150 + }, + { + "epoch": 4.8545745105872955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12151 + }, + { + "epoch": 4.854974031162605, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 12152 + }, + { + "epoch": 4.855373551737914, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12153 + }, + { + "epoch": 4.855773072313224, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12154 + }, + { + "epoch": 4.856172592888534, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12155 + }, + { + "epoch": 4.8565721134638435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12156 + }, + { + "epoch": 4.856971634039153, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12157 + }, + { + "epoch": 4.857371154614462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12158 + }, + { + "epoch": 4.857770675189772, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12159 + }, + { + "epoch": 4.858170195765082, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12160 + }, + { + "epoch": 4.858569716340392, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12161 + }, + { + "epoch": 4.858969236915701, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12162 + }, + { + "epoch": 4.859368757491011, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12163 + }, + { + "epoch": 4.85976827806632, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12164 + }, + { + "epoch": 4.86016779864163, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12165 + }, + { + "epoch": 4.86056731921694, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12166 + }, + { + "epoch": 4.860966839792249, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 12167 + }, + { + "epoch": 4.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12168 + }, + { + "epoch": 4.861765880942869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12169 + }, + { + "epoch": 4.862165401518178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12170 + }, + { + "epoch": 4.862564922093488, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12171 + }, + { + "epoch": 4.8629644426687975, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12172 + }, + { + "epoch": 4.863363963244107, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12173 + }, + { + "epoch": 4.863763483819417, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12174 + }, + { + "epoch": 4.864163004394726, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12175 + }, + { + "epoch": 4.864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12176 + }, + { + "epoch": 4.8649620455453455, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 12177 + }, + { + "epoch": 4.865361566120655, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12178 + }, + { + "epoch": 4.865761086695965, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12179 + }, + { + "epoch": 4.866160607271274, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12180 + }, + { + "epoch": 4.866560127846584, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12181 + }, + { + "epoch": 4.8669596484218935, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12182 + }, + { + "epoch": 4.867359168997203, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12183 + }, + { + "epoch": 4.867758689572513, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12184 + }, + { + "epoch": 4.868158210147823, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12185 + }, + { + "epoch": 4.868557730723133, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12186 + }, + { + "epoch": 4.868957251298442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12187 + }, + { + "epoch": 4.869356771873751, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12188 + }, + { + "epoch": 4.869756292449061, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12189 + }, + { + "epoch": 4.870155813024371, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12190 + }, + { + "epoch": 4.870555333599681, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12191 + }, + { + "epoch": 4.87095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12192 + }, + { + "epoch": 4.871354374750299, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12193 + }, + { + "epoch": 4.871753895325609, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12194 + }, + { + "epoch": 4.872153415900919, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12195 + }, + { + "epoch": 4.872552936476229, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 12196 + }, + { + "epoch": 4.872952457051538, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12197 + }, + { + "epoch": 4.8733519776268475, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12198 + }, + { + "epoch": 4.873751498202157, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 12199 + }, + { + "epoch": 4.874151018777467, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12200 + }, + { + "epoch": 4.874550539352777, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 12201 + }, + { + "epoch": 4.874950059928087, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12202 + }, + { + "epoch": 4.875349580503396, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12203 + }, + { + "epoch": 4.875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12204 + }, + { + "epoch": 4.876148621654015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12205 + }, + { + "epoch": 4.876548142229325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12206 + }, + { + "epoch": 4.876947662804635, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 12207 + }, + { + "epoch": 4.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12208 + }, + { + "epoch": 4.877746703955253, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12209 + }, + { + "epoch": 4.878146224530563, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12210 + }, + { + "epoch": 4.878545745105873, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12211 + }, + { + "epoch": 4.878945265681183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12212 + }, + { + "epoch": 4.8793447862564925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12213 + }, + { + "epoch": 4.879744306831801, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12214 + }, + { + "epoch": 4.880143827407111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12215 + }, + { + "epoch": 4.880543347982421, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12216 + }, + { + "epoch": 4.880942868557731, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12217 + }, + { + "epoch": 4.8813423891330405, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 12218 + }, + { + "epoch": 4.88174190970835, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12219 + }, + { + "epoch": 4.88214143028366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12220 + }, + { + "epoch": 4.882540950858969, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12221 + }, + { + "epoch": 4.882940471434279, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12222 + }, + { + "epoch": 4.883339992009589, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12223 + }, + { + "epoch": 4.883739512584898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12224 + }, + { + "epoch": 4.884139033160208, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12225 + }, + { + "epoch": 4.884538553735517, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12226 + }, + { + "epoch": 4.884938074310827, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12227 + }, + { + "epoch": 4.885337594886137, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12228 + }, + { + "epoch": 4.885737115461446, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12229 + }, + { + "epoch": 4.886136636036756, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12230 + }, + { + "epoch": 4.886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12231 + }, + { + "epoch": 4.886935677187375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12232 + }, + { + "epoch": 4.887335197762685, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12233 + }, + { + "epoch": 4.887734718337994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12234 + }, + { + "epoch": 4.888134238913304, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12235 + }, + { + "epoch": 4.888533759488614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12236 + }, + { + "epoch": 4.888933280063923, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12237 + }, + { + "epoch": 4.889332800639233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12238 + }, + { + "epoch": 4.8897323212145425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12239 + }, + { + "epoch": 4.890131841789852, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12240 + }, + { + "epoch": 4.890531362365162, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12241 + }, + { + "epoch": 4.890930882940472, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12242 + }, + { + "epoch": 4.891330403515781, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12243 + }, + { + "epoch": 4.8917299240910905, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12244 + }, + { + "epoch": 4.8921294446664, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 12245 + }, + { + "epoch": 4.89252896524171, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12246 + }, + { + "epoch": 4.89292848581702, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12247 + }, + { + "epoch": 4.893328006392329, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12248 + }, + { + "epoch": 4.893727526967639, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12249 + }, + { + "epoch": 4.894127047542948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12250 + }, + { + "epoch": 4.894526568118258, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12251 + }, + { + "epoch": 4.894926088693568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12252 + }, + { + "epoch": 4.895325609268878, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12253 + }, + { + "epoch": 4.895725129844187, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12254 + }, + { + "epoch": 4.896124650419496, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12255 + }, + { + "epoch": 4.896524170994806, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12256 + }, + { + "epoch": 4.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12257 + }, + { + "epoch": 4.897323212145426, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12258 + }, + { + "epoch": 4.8977227327207356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12259 + }, + { + "epoch": 4.8981222532960444, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12260 + }, + { + "epoch": 4.898521773871354, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12261 + }, + { + "epoch": 4.898921294446664, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12262 + }, + { + "epoch": 4.899320815021974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12263 + }, + { + "epoch": 4.899720335597284, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12264 + }, + { + "epoch": 4.9001198561725925, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12265 + }, + { + "epoch": 4.900519376747902, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12266 + }, + { + "epoch": 4.900918897323212, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12267 + }, + { + "epoch": 4.901318417898522, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12268 + }, + { + "epoch": 4.901717938473832, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12269 + }, + { + "epoch": 4.9021174590491405, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12270 + }, + { + "epoch": 4.90251697962445, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 12271 + }, + { + "epoch": 4.90291650019976, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12272 + }, + { + "epoch": 4.90331602077507, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12273 + }, + { + "epoch": 4.90371554135038, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12274 + }, + { + "epoch": 4.9041150619256895, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 12275 + }, + { + "epoch": 4.904514582500999, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12276 + }, + { + "epoch": 4.904914103076308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 12277 + }, + { + "epoch": 4.905313623651618, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12278 + }, + { + "epoch": 4.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12279 + }, + { + "epoch": 4.9061126648022375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12280 + }, + { + "epoch": 4.906512185377547, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12281 + }, + { + "epoch": 4.906911705952856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12282 + }, + { + "epoch": 4.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 12283 + }, + { + "epoch": 4.907710747103476, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12284 + }, + { + "epoch": 4.9081102676787856, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12285 + }, + { + "epoch": 4.908509788254095, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12286 + }, + { + "epoch": 4.908909308829404, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12287 + }, + { + "epoch": 4.909308829404714, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12288 + }, + { + "epoch": 4.909708349980024, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12289 + }, + { + "epoch": 4.910107870555334, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12290 + }, + { + "epoch": 4.910507391130643, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12291 + }, + { + "epoch": 4.910906911705953, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12292 + }, + { + "epoch": 4.911306432281263, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12293 + }, + { + "epoch": 4.911705952856572, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 12294 + }, + { + "epoch": 4.912105473431882, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12295 + }, + { + "epoch": 4.912504994007191, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12296 + }, + { + "epoch": 4.912904514582501, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12297 + }, + { + "epoch": 4.913304035157811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12298 + }, + { + "epoch": 4.91370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12299 + }, + { + "epoch": 4.91410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12300 + }, + { + "epoch": 4.9145025968837395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12301 + }, + { + "epoch": 4.914902117459049, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12302 + }, + { + "epoch": 4.915301638034359, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12303 + }, + { + "epoch": 4.915701158609668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12304 + }, + { + "epoch": 4.916100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12305 + }, + { + "epoch": 4.9165001997602875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12306 + }, + { + "epoch": 4.916899720335597, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12307 + }, + { + "epoch": 4.917299240910907, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12308 + }, + { + "epoch": 4.917698761486217, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12309 + }, + { + "epoch": 4.918098282061527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12310 + }, + { + "epoch": 4.918497802636836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12311 + }, + { + "epoch": 4.918897323212145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12312 + }, + { + "epoch": 4.919296843787455, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12313 + }, + { + "epoch": 4.919696364362765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12314 + }, + { + "epoch": 4.920095884938075, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12315 + }, + { + "epoch": 4.920495405513384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12316 + }, + { + "epoch": 4.920894926088693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12317 + }, + { + "epoch": 4.921294446664003, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12318 + }, + { + "epoch": 4.921693967239313, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12319 + }, + { + "epoch": 4.922093487814623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12320 + }, + { + "epoch": 4.922493008389932, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12321 + }, + { + "epoch": 4.922892528965241, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12322 + }, + { + "epoch": 4.923292049540551, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12323 + }, + { + "epoch": 4.923691570115861, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12324 + }, + { + "epoch": 4.924091090691171, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12325 + }, + { + "epoch": 4.924490611266481, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12326 + }, + { + "epoch": 4.9248901318417895, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12327 + }, + { + "epoch": 4.925289652417099, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12328 + }, + { + "epoch": 4.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12329 + }, + { + "epoch": 4.926088693567719, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12330 + }, + { + "epoch": 4.926488214143029, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2973, + "step": 12331 + }, + { + "epoch": 4.926887734718338, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12332 + }, + { + "epoch": 4.927287255293647, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 12333 + }, + { + "epoch": 4.927686775868957, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12334 + }, + { + "epoch": 4.928086296444267, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12335 + }, + { + "epoch": 4.928485817019577, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12336 + }, + { + "epoch": 4.9288853375948865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12337 + }, + { + "epoch": 4.929284858170195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 12338 + }, + { + "epoch": 4.929684378745505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12339 + }, + { + "epoch": 4.930083899320815, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12340 + }, + { + "epoch": 4.930483419896125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12341 + }, + { + "epoch": 4.9308829404714345, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12342 + }, + { + "epoch": 4.931282461046744, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12343 + }, + { + "epoch": 4.931681981622053, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12344 + }, + { + "epoch": 4.932081502197363, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12345 + }, + { + "epoch": 4.932481022772673, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12346 + }, + { + "epoch": 4.9328805433479825, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12347 + }, + { + "epoch": 4.933280063923292, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12348 + }, + { + "epoch": 4.933679584498602, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12349 + }, + { + "epoch": 4.934079105073911, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12350 + }, + { + "epoch": 4.934478625649221, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12351 + }, + { + "epoch": 4.934878146224531, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12352 + }, + { + "epoch": 4.93527766679984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12353 + }, + { + "epoch": 4.93567718737515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12354 + }, + { + "epoch": 4.936076707950459, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12355 + }, + { + "epoch": 4.936476228525769, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12356 + }, + { + "epoch": 4.936875749101079, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12357 + }, + { + "epoch": 4.937275269676388, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12358 + }, + { + "epoch": 4.937674790251698, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12359 + }, + { + "epoch": 4.938074310827007, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12360 + }, + { + "epoch": 4.938473831402317, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12361 + }, + { + "epoch": 4.938873351977627, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12362 + }, + { + "epoch": 4.9392728725529365, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12363 + }, + { + "epoch": 4.939672393128246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12364 + }, + { + "epoch": 4.940071913703556, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12365 + }, + { + "epoch": 4.940471434278866, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12366 + }, + { + "epoch": 4.940870954854175, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12367 + }, + { + "epoch": 4.9412704754294845, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12368 + }, + { + "epoch": 4.941669996004794, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12369 + }, + { + "epoch": 4.942069516580104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12370 + }, + { + "epoch": 4.942469037155414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12371 + }, + { + "epoch": 4.942868557730723, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12372 + }, + { + "epoch": 4.9432680783060325, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12373 + }, + { + "epoch": 4.943667598881342, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12374 + }, + { + "epoch": 4.944067119456652, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 12375 + }, + { + "epoch": 4.944466640031962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12376 + }, + { + "epoch": 4.944866160607271, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12377 + }, + { + "epoch": 4.945265681182581, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12378 + }, + { + "epoch": 4.94566520175789, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12379 + }, + { + "epoch": 4.9460647223332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12380 + }, + { + "epoch": 4.94646424290851, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12381 + }, + { + "epoch": 4.94686376348382, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12382 + }, + { + "epoch": 4.9472632840591295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12383 + }, + { + "epoch": 4.947662804634438, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12384 + }, + { + "epoch": 4.948062325209748, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12385 + }, + { + "epoch": 4.948461845785058, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12386 + }, + { + "epoch": 4.948861366360368, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12387 + }, + { + "epoch": 4.949260886935678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12388 + }, + { + "epoch": 4.9496604075109865, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12389 + }, + { + "epoch": 4.950059928086296, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12390 + }, + { + "epoch": 4.950459448661606, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12391 + }, + { + "epoch": 4.950858969236916, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12392 + }, + { + "epoch": 4.951258489812226, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 12393 + }, + { + "epoch": 4.9516580103875345, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12394 + }, + { + "epoch": 4.952057530962844, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 12395 + }, + { + "epoch": 4.952457051538154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12396 + }, + { + "epoch": 4.952856572113464, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12397 + }, + { + "epoch": 4.953256092688774, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12398 + }, + { + "epoch": 4.953655613264083, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12399 + }, + { + "epoch": 4.954055133839393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12400 + }, + { + "epoch": 4.954454654414702, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12401 + }, + { + "epoch": 4.954854174990012, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12402 + }, + { + "epoch": 4.955253695565322, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12403 + }, + { + "epoch": 4.9556532161406315, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12404 + }, + { + "epoch": 4.956052736715941, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12405 + }, + { + "epoch": 4.95645225729125, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12406 + }, + { + "epoch": 4.95685177786656, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12407 + }, + { + "epoch": 4.95725129844187, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12408 + }, + { + "epoch": 4.9576508190171795, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 12409 + }, + { + "epoch": 4.958050339592489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12410 + }, + { + "epoch": 4.958449860167798, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12411 + }, + { + "epoch": 4.958849380743108, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 12412 + }, + { + "epoch": 4.959248901318418, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12413 + }, + { + "epoch": 4.959648421893728, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12414 + }, + { + "epoch": 4.960047942469037, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12415 + }, + { + "epoch": 4.960447463044347, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12416 + }, + { + "epoch": 4.960846983619656, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12417 + }, + { + "epoch": 4.961246504194966, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12418 + }, + { + "epoch": 4.961646024770276, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12419 + }, + { + "epoch": 4.962045545345585, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12420 + }, + { + "epoch": 4.962445065920895, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12421 + }, + { + "epoch": 4.962844586496205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12422 + }, + { + "epoch": 4.963244107071514, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12423 + }, + { + "epoch": 4.963643627646824, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12424 + }, + { + "epoch": 4.964043148222133, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12425 + }, + { + "epoch": 4.964442668797443, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12426 + }, + { + "epoch": 4.964842189372753, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12427 + }, + { + "epoch": 4.965241709948062, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12428 + }, + { + "epoch": 4.965641230523372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12429 + }, + { + "epoch": 4.9660407510986815, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 12430 + }, + { + "epoch": 4.966440271673991, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12431 + }, + { + "epoch": 4.966839792249301, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12432 + }, + { + "epoch": 4.967239312824611, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12433 + }, + { + "epoch": 4.96763883339992, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12434 + }, + { + "epoch": 4.9680383539752295, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12435 + }, + { + "epoch": 4.968437874550539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12436 + }, + { + "epoch": 4.968837395125849, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12437 + }, + { + "epoch": 4.969236915701159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12438 + }, + { + "epoch": 4.969636436276469, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12439 + }, + { + "epoch": 4.970035956851778, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12440 + }, + { + "epoch": 4.970435477427087, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12441 + }, + { + "epoch": 4.970834998002397, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12442 + }, + { + "epoch": 4.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12443 + }, + { + "epoch": 4.971634039153017, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12444 + }, + { + "epoch": 4.972033559728326, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12445 + }, + { + "epoch": 4.972433080303635, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12446 + }, + { + "epoch": 4.972832600878945, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12447 + }, + { + "epoch": 4.973232121454255, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12448 + }, + { + "epoch": 4.973631642029565, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12449 + }, + { + "epoch": 4.974031162604874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12450 + }, + { + "epoch": 4.974430683180183, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12451 + }, + { + "epoch": 4.974830203755493, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12452 + }, + { + "epoch": 4.975229724330803, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12453 + }, + { + "epoch": 4.975629244906113, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12454 + }, + { + "epoch": 4.976028765481423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12455 + }, + { + "epoch": 4.976428286056732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12456 + }, + { + "epoch": 4.976827806632041, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12457 + }, + { + "epoch": 4.977227327207351, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12458 + }, + { + "epoch": 4.977626847782661, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12459 + }, + { + "epoch": 4.978026368357971, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 12460 + }, + { + "epoch": 4.97842588893328, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12461 + }, + { + "epoch": 4.978825409508589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 12462 + }, + { + "epoch": 4.979224930083899, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12463 + }, + { + "epoch": 4.979624450659209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 12464 + }, + { + "epoch": 4.980023971234519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 12465 + }, + { + "epoch": 4.9804234918098285, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12466 + }, + { + "epoch": 4.980823012385137, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12467 + }, + { + "epoch": 4.981222532960447, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12468 + }, + { + "epoch": 4.981622053535757, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12469 + }, + { + "epoch": 4.982021574111067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12470 + }, + { + "epoch": 4.9824210946863765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12471 + }, + { + "epoch": 4.982820615261686, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12472 + }, + { + "epoch": 4.983220135836996, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12473 + }, + { + "epoch": 4.983619656412305, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12474 + }, + { + "epoch": 4.984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12475 + }, + { + "epoch": 4.9844186975629245, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12476 + }, + { + "epoch": 4.984818218138234, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12477 + }, + { + "epoch": 4.985217738713544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12478 + }, + { + "epoch": 4.985617259288853, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12479 + }, + { + "epoch": 4.986016779864163, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12480 + }, + { + "epoch": 4.986416300439473, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12481 + }, + { + "epoch": 4.986815821014782, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12482 + }, + { + "epoch": 4.987215341590092, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12483 + }, + { + "epoch": 4.987614862165401, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12484 + }, + { + "epoch": 4.988014382740711, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12485 + }, + { + "epoch": 4.988413903316021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12486 + }, + { + "epoch": 4.98881342389133, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12487 + }, + { + "epoch": 4.98921294446664, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12488 + }, + { + "epoch": 4.98961246504195, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12489 + }, + { + "epoch": 4.99001198561726, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12490 + }, + { + "epoch": 4.990411506192569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12491 + }, + { + "epoch": 4.9908110267678785, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12492 + }, + { + "epoch": 4.991210547343188, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12493 + }, + { + "epoch": 4.991610067918498, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12494 + }, + { + "epoch": 4.992009588493808, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12495 + }, + { + "epoch": 4.992409109069117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12496 + }, + { + "epoch": 4.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12497 + }, + { + "epoch": 4.993208150219736, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12498 + }, + { + "epoch": 4.993607670795046, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12499 + }, + { + "epoch": 4.994007191370356, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 12500 + }, + { + "epoch": 4.994406711945665, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12501 + }, + { + "epoch": 4.9948062325209746, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12502 + }, + { + "epoch": 4.995205753096284, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12503 + }, + { + "epoch": 4.995605273671594, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12504 + }, + { + "epoch": 4.996004794246904, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12505 + }, + { + "epoch": 4.996404314822214, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12506 + }, + { + "epoch": 4.996803835397523, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12507 + }, + { + "epoch": 4.997203355972832, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12508 + }, + { + "epoch": 4.997602876548142, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12509 + }, + { + "epoch": 4.998002397123452, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12510 + }, + { + "epoch": 4.998401917698762, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 12511 + }, + { + "epoch": 4.9988014382740715, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12512 + }, + { + "epoch": 4.99920095884938, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12513 + }, + { + "epoch": 4.99960047942469, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12514 + }, + { + "epoch": 5.0, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12515 + }, + { + "epoch": 5.00039952057531, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12516 + }, + { + "epoch": 5.00079904115062, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12517 + }, + { + "epoch": 5.0011985617259285, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12518 + }, + { + "epoch": 5.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12519 + }, + { + "epoch": 5.001997602876548, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12520 + }, + { + "epoch": 5.002397123451858, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12521 + }, + { + "epoch": 5.002796644027168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12522 + }, + { + "epoch": 5.003196164602477, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12523 + }, + { + "epoch": 5.003595685177786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12524 + }, + { + "epoch": 5.003995205753096, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12525 + }, + { + "epoch": 5.004394726328406, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12526 + }, + { + "epoch": 5.004794246903716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12527 + }, + { + "epoch": 5.0051937674790254, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12528 + }, + { + "epoch": 5.005593288054335, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12529 + }, + { + "epoch": 5.005992808629644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12530 + }, + { + "epoch": 5.006392329204954, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12531 + }, + { + "epoch": 5.006791849780264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12532 + }, + { + "epoch": 5.0071913703555735, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12533 + }, + { + "epoch": 5.007590890930883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12534 + }, + { + "epoch": 5.007990411506192, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12535 + }, + { + "epoch": 5.008389932081502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12536 + }, + { + "epoch": 5.008789452656812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12537 + }, + { + "epoch": 5.0091889732321215, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12538 + }, + { + "epoch": 5.009588493807431, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12539 + }, + { + "epoch": 5.009988014382741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12540 + }, + { + "epoch": 5.01038753495805, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12541 + }, + { + "epoch": 5.01078705553336, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12542 + }, + { + "epoch": 5.01118657610867, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12543 + }, + { + "epoch": 5.011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12544 + }, + { + "epoch": 5.011985617259289, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12545 + }, + { + "epoch": 5.012385137834598, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12546 + }, + { + "epoch": 5.012784658409908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12547 + }, + { + "epoch": 5.013184178985218, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12548 + }, + { + "epoch": 5.013583699560527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12549 + }, + { + "epoch": 5.013983220135837, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 12550 + }, + { + "epoch": 5.014382740711147, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12551 + }, + { + "epoch": 5.014782261286456, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12552 + }, + { + "epoch": 5.015181781861766, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 12553 + }, + { + "epoch": 5.0155813024370755, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12554 + }, + { + "epoch": 5.015980823012385, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12555 + }, + { + "epoch": 5.016380343587695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12556 + }, + { + "epoch": 5.016779864163005, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12557 + }, + { + "epoch": 5.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12558 + }, + { + "epoch": 5.0175789053136235, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12559 + }, + { + "epoch": 5.017978425888933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12560 + }, + { + "epoch": 5.018377946464243, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12561 + }, + { + "epoch": 5.018777467039553, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12562 + }, + { + "epoch": 5.019176987614862, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12563 + }, + { + "epoch": 5.0195765081901715, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12564 + }, + { + "epoch": 5.019976028765481, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12565 + }, + { + "epoch": 5.020375549340791, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12566 + }, + { + "epoch": 5.020775069916101, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12567 + }, + { + "epoch": 5.021174590491411, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12568 + }, + { + "epoch": 5.02157411106672, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12569 + }, + { + "epoch": 5.021973631642029, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 12570 + }, + { + "epoch": 5.022373152217339, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12571 + }, + { + "epoch": 5.022772672792649, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12572 + }, + { + "epoch": 5.023172193367959, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12573 + }, + { + "epoch": 5.0235717139432685, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12574 + }, + { + "epoch": 5.023971234518577, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12575 + }, + { + "epoch": 5.024370755093887, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12576 + }, + { + "epoch": 5.024770275669197, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12577 + }, + { + "epoch": 5.025169796244507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12578 + }, + { + "epoch": 5.025569316819817, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12579 + }, + { + "epoch": 5.0259688373951255, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12580 + }, + { + "epoch": 5.026368357970435, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12581 + }, + { + "epoch": 5.026767878545745, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12582 + }, + { + "epoch": 5.027167399121055, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12583 + }, + { + "epoch": 5.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12584 + }, + { + "epoch": 5.027966440271674, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12585 + }, + { + "epoch": 5.028365960846983, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12586 + }, + { + "epoch": 5.028765481422293, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12587 + }, + { + "epoch": 5.029165001997603, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12588 + }, + { + "epoch": 5.029564522572913, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12589 + }, + { + "epoch": 5.029964043148222, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12590 + }, + { + "epoch": 5.030363563723531, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12591 + }, + { + "epoch": 5.030763084298841, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12592 + }, + { + "epoch": 5.031162604874151, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12593 + }, + { + "epoch": 5.031562125449461, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12594 + }, + { + "epoch": 5.0319616460247705, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12595 + }, + { + "epoch": 5.03236116660008, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12596 + }, + { + "epoch": 5.032760687175389, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12597 + }, + { + "epoch": 5.033160207750699, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 12598 + }, + { + "epoch": 5.033559728326009, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12599 + }, + { + "epoch": 5.0339592489013185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12600 + }, + { + "epoch": 5.034358769476628, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12601 + }, + { + "epoch": 5.034758290051938, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12602 + }, + { + "epoch": 5.035157810627247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12603 + }, + { + "epoch": 5.035557331202557, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 12604 + }, + { + "epoch": 5.035956851777867, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12605 + }, + { + "epoch": 5.036356372353176, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12606 + }, + { + "epoch": 5.036755892928486, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12607 + }, + { + "epoch": 5.037155413503795, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12608 + }, + { + "epoch": 5.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12609 + }, + { + "epoch": 5.037954454654415, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12610 + }, + { + "epoch": 5.038353975229724, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12611 + }, + { + "epoch": 5.038753495805034, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12612 + }, + { + "epoch": 5.039153016380344, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12613 + }, + { + "epoch": 5.039552536955653, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12614 + }, + { + "epoch": 5.039952057530963, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12615 + }, + { + "epoch": 5.040351578106272, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 12616 + }, + { + "epoch": 5.040751098681582, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12617 + }, + { + "epoch": 5.041150619256892, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 12618 + }, + { + "epoch": 5.041550139832202, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12619 + }, + { + "epoch": 5.041949660407511, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12620 + }, + { + "epoch": 5.0423491809828205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12621 + }, + { + "epoch": 5.04274870155813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12622 + }, + { + "epoch": 5.04314822213344, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 12623 + }, + { + "epoch": 5.04354774270875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12624 + }, + { + "epoch": 5.043947263284059, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12625 + }, + { + "epoch": 5.0443467838593685, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12626 + }, + { + "epoch": 5.044746304434678, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12627 + }, + { + "epoch": 5.045145825009988, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12628 + }, + { + "epoch": 5.045545345585298, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12629 + }, + { + "epoch": 5.045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12630 + }, + { + "epoch": 5.046344386735917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12631 + }, + { + "epoch": 5.046743907311226, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12632 + }, + { + "epoch": 5.047143427886536, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12633 + }, + { + "epoch": 5.047542948461846, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12634 + }, + { + "epoch": 5.047942469037156, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12635 + }, + { + "epoch": 5.048341989612465, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12636 + }, + { + "epoch": 5.048741510187774, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12637 + }, + { + "epoch": 5.049141030763084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12638 + }, + { + "epoch": 5.049540551338394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12639 + }, + { + "epoch": 5.049940071913704, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12640 + }, + { + "epoch": 5.0503395924890135, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12641 + }, + { + "epoch": 5.050739113064322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12642 + }, + { + "epoch": 5.051138633639632, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12643 + }, + { + "epoch": 5.051538154214942, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12644 + }, + { + "epoch": 5.051937674790252, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 12645 + }, + { + "epoch": 5.052337195365562, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12646 + }, + { + "epoch": 5.052736715940871, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12647 + }, + { + "epoch": 5.05313623651618, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12648 + }, + { + "epoch": 5.05353575709149, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12649 + }, + { + "epoch": 5.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12650 + }, + { + "epoch": 5.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12651 + }, + { + "epoch": 5.054734318817419, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12652 + }, + { + "epoch": 5.055133839392728, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12653 + }, + { + "epoch": 5.055533359968038, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12654 + }, + { + "epoch": 5.055932880543348, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12655 + }, + { + "epoch": 5.056332401118658, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12656 + }, + { + "epoch": 5.0567319216939675, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12657 + }, + { + "epoch": 5.057131442269277, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12658 + }, + { + "epoch": 5.057530962844586, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12659 + }, + { + "epoch": 5.057930483419896, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12660 + }, + { + "epoch": 5.058330003995206, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12661 + }, + { + "epoch": 5.0587295245705155, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12662 + }, + { + "epoch": 5.059129045145825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12663 + }, + { + "epoch": 5.059528565721135, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12664 + }, + { + "epoch": 5.059928086296444, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12665 + }, + { + "epoch": 5.060327606871754, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12666 + }, + { + "epoch": 5.0607271274470635, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12667 + }, + { + "epoch": 5.061126648022373, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12668 + }, + { + "epoch": 5.061526168597683, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12669 + }, + { + "epoch": 5.061925689172992, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 12670 + }, + { + "epoch": 5.062325209748302, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12671 + }, + { + "epoch": 5.062724730323612, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12672 + }, + { + "epoch": 5.063124250898921, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12673 + }, + { + "epoch": 5.063523771474231, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12674 + }, + { + "epoch": 5.063923292049541, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12675 + }, + { + "epoch": 5.06432281262485, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12676 + }, + { + "epoch": 5.06472233320016, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12677 + }, + { + "epoch": 5.065121853775469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12678 + }, + { + "epoch": 5.065521374350779, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12679 + }, + { + "epoch": 5.065920894926089, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12680 + }, + { + "epoch": 5.066320415501398, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12681 + }, + { + "epoch": 5.066719936076708, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12682 + }, + { + "epoch": 5.0671194566520175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12683 + }, + { + "epoch": 5.067518977227327, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12684 + }, + { + "epoch": 5.067918497802637, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12685 + }, + { + "epoch": 5.068318018377947, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12686 + }, + { + "epoch": 5.068717538953256, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12687 + }, + { + "epoch": 5.0691170595285655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12688 + }, + { + "epoch": 5.069516580103875, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12689 + }, + { + "epoch": 5.069916100679185, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12690 + }, + { + "epoch": 5.070315621254495, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12691 + }, + { + "epoch": 5.070715141829805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12692 + }, + { + "epoch": 5.0711146624051135, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12693 + }, + { + "epoch": 5.071514182980423, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12694 + }, + { + "epoch": 5.071913703555733, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12695 + }, + { + "epoch": 5.072313224131043, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 12696 + }, + { + "epoch": 5.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12697 + }, + { + "epoch": 5.073112265281662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12698 + }, + { + "epoch": 5.073511785856971, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12699 + }, + { + "epoch": 5.073911306432281, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12700 + }, + { + "epoch": 5.074310827007591, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12701 + }, + { + "epoch": 5.074710347582901, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 12702 + }, + { + "epoch": 5.0751098681582105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12703 + }, + { + "epoch": 5.075509388733519, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12704 + }, + { + "epoch": 5.075908909308829, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12705 + }, + { + "epoch": 5.076308429884139, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12706 + }, + { + "epoch": 5.076707950459449, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12707 + }, + { + "epoch": 5.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12708 + }, + { + "epoch": 5.077506991610068, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12709 + }, + { + "epoch": 5.077906512185377, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 12710 + }, + { + "epoch": 5.078306032760687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12711 + }, + { + "epoch": 5.078705553335997, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12712 + }, + { + "epoch": 5.079105073911307, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12713 + }, + { + "epoch": 5.079504594486616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12714 + }, + { + "epoch": 5.079904115061925, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12715 + }, + { + "epoch": 5.080303635637235, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12716 + }, + { + "epoch": 5.080703156212545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12717 + }, + { + "epoch": 5.081102676787855, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12718 + }, + { + "epoch": 5.081502197363164, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12719 + }, + { + "epoch": 5.081901717938474, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12720 + }, + { + "epoch": 5.082301238513783, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12721 + }, + { + "epoch": 5.082700759089093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12722 + }, + { + "epoch": 5.083100279664403, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12723 + }, + { + "epoch": 5.0834998002397125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12724 + }, + { + "epoch": 5.083899320815022, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12725 + }, + { + "epoch": 5.084298841390331, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 12726 + }, + { + "epoch": 5.084698361965641, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12727 + }, + { + "epoch": 5.085097882540951, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12728 + }, + { + "epoch": 5.0854974031162605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12729 + }, + { + "epoch": 5.08589692369157, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12730 + }, + { + "epoch": 5.08629644426688, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12731 + }, + { + "epoch": 5.086695964842189, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12732 + }, + { + "epoch": 5.087095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12733 + }, + { + "epoch": 5.087495005992809, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 12734 + }, + { + "epoch": 5.087894526568118, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12735 + }, + { + "epoch": 5.088294047143428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12736 + }, + { + "epoch": 5.088693567718738, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12737 + }, + { + "epoch": 5.089093088294047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 12738 + }, + { + "epoch": 5.089492608869357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12739 + }, + { + "epoch": 5.089892129444666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12740 + }, + { + "epoch": 5.090291650019976, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12741 + }, + { + "epoch": 5.090691170595286, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12742 + }, + { + "epoch": 5.091090691170595, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12743 + }, + { + "epoch": 5.091490211745905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12744 + }, + { + "epoch": 5.0918897323212144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12745 + }, + { + "epoch": 5.092289252896524, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12746 + }, + { + "epoch": 5.092688773471834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12747 + }, + { + "epoch": 5.093088294047144, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 12748 + }, + { + "epoch": 5.093487814622453, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 12749 + }, + { + "epoch": 5.0938873351977625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12750 + }, + { + "epoch": 5.094286855773072, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12751 + }, + { + "epoch": 5.094686376348382, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12752 + }, + { + "epoch": 5.095085896923692, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12753 + }, + { + "epoch": 5.095485417499002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12754 + }, + { + "epoch": 5.0958849380743105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12755 + }, + { + "epoch": 5.09628445864962, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12756 + }, + { + "epoch": 5.09668397922493, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12757 + }, + { + "epoch": 5.09708349980024, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12758 + }, + { + "epoch": 5.09748302037555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12759 + }, + { + "epoch": 5.097882540950859, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12760 + }, + { + "epoch": 5.098282061526168, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 12761 + }, + { + "epoch": 5.098681582101478, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12762 + }, + { + "epoch": 5.099081102676788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12763 + }, + { + "epoch": 5.099480623252098, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12764 + }, + { + "epoch": 5.0998801438274075, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12765 + }, + { + "epoch": 5.100279664402716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 12766 + }, + { + "epoch": 5.100679184978026, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12767 + }, + { + "epoch": 5.101078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 12768 + }, + { + "epoch": 5.101478226128646, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12769 + }, + { + "epoch": 5.1018777467039556, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12770 + }, + { + "epoch": 5.1022772672792644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12771 + }, + { + "epoch": 5.102676787854574, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12772 + }, + { + "epoch": 5.103076308429884, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12773 + }, + { + "epoch": 5.103475829005194, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 12774 + }, + { + "epoch": 5.103875349580504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12775 + }, + { + "epoch": 5.104274870155813, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12776 + }, + { + "epoch": 5.104674390731122, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12777 + }, + { + "epoch": 5.105073911306432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12778 + }, + { + "epoch": 5.105473431881742, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12779 + }, + { + "epoch": 5.105872952457052, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12780 + }, + { + "epoch": 5.106272473032361, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12781 + }, + { + "epoch": 5.106671993607671, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12782 + }, + { + "epoch": 5.10707151418298, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12783 + }, + { + "epoch": 5.10747103475829, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12784 + }, + { + "epoch": 5.1078705553336, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12785 + }, + { + "epoch": 5.1082700759089095, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 12786 + }, + { + "epoch": 5.108669596484219, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12787 + }, + { + "epoch": 5.109069117059528, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12788 + }, + { + "epoch": 5.109468637634838, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12789 + }, + { + "epoch": 5.109868158210148, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12790 + }, + { + "epoch": 5.1102676787854575, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12791 + }, + { + "epoch": 5.110667199360767, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12792 + }, + { + "epoch": 5.111066719936077, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12793 + }, + { + "epoch": 5.111466240511386, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12794 + }, + { + "epoch": 5.111865761086696, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12795 + }, + { + "epoch": 5.112265281662006, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12796 + }, + { + "epoch": 5.112664802237315, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12797 + }, + { + "epoch": 5.113064322812625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12798 + }, + { + "epoch": 5.113463843387935, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12799 + }, + { + "epoch": 5.113863363963244, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12800 + }, + { + "epoch": 5.114262884538554, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12801 + }, + { + "epoch": 5.114662405113863, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12802 + }, + { + "epoch": 5.115061925689173, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12803 + }, + { + "epoch": 5.115461446264483, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12804 + }, + { + "epoch": 5.115860966839792, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12805 + }, + { + "epoch": 5.116260487415102, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12806 + }, + { + "epoch": 5.116660007990411, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12807 + }, + { + "epoch": 5.117059528565721, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12808 + }, + { + "epoch": 5.117459049141031, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12809 + }, + { + "epoch": 5.117858569716341, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12810 + }, + { + "epoch": 5.11825809029165, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12811 + }, + { + "epoch": 5.1186576108669595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12812 + }, + { + "epoch": 5.119057131442269, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12813 + }, + { + "epoch": 5.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12814 + }, + { + "epoch": 5.119856172592889, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12815 + }, + { + "epoch": 5.120255693168198, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12816 + }, + { + "epoch": 5.1206552137435075, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12817 + }, + { + "epoch": 5.121054734318817, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12818 + }, + { + "epoch": 5.121454254894127, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12819 + }, + { + "epoch": 5.121853775469437, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12820 + }, + { + "epoch": 5.122253296044747, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3011, + "step": 12821 + }, + { + "epoch": 5.122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12822 + }, + { + "epoch": 5.123052337195365, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12823 + }, + { + "epoch": 5.123451857770675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12824 + }, + { + "epoch": 5.123851378345985, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12825 + }, + { + "epoch": 5.124250898921295, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 12826 + }, + { + "epoch": 5.1246504194966045, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12827 + }, + { + "epoch": 5.125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12828 + }, + { + "epoch": 5.125449460647223, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 12829 + }, + { + "epoch": 5.125848981222533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12830 + }, + { + "epoch": 5.126248501797843, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12831 + }, + { + "epoch": 5.1266480223731525, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12832 + }, + { + "epoch": 5.127047542948461, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12833 + }, + { + "epoch": 5.127447063523771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12834 + }, + { + "epoch": 5.127846584099081, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12835 + }, + { + "epoch": 5.128246104674391, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12836 + }, + { + "epoch": 5.128645625249701, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12837 + }, + { + "epoch": 5.12904514582501, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12838 + }, + { + "epoch": 5.129444666400319, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12839 + }, + { + "epoch": 5.129844186975629, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12840 + }, + { + "epoch": 5.130243707550939, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12841 + }, + { + "epoch": 5.130643228126249, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12842 + }, + { + "epoch": 5.131042748701558, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 12843 + }, + { + "epoch": 5.131442269276867, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12844 + }, + { + "epoch": 5.131841789852177, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12845 + }, + { + "epoch": 5.132241310427487, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12846 + }, + { + "epoch": 5.132640831002797, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12847 + }, + { + "epoch": 5.1330403515781065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12848 + }, + { + "epoch": 5.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12849 + }, + { + "epoch": 5.133839392728725, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12850 + }, + { + "epoch": 5.134238913304035, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 12851 + }, + { + "epoch": 5.134638433879345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12852 + }, + { + "epoch": 5.1350379544546545, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12853 + }, + { + "epoch": 5.135437475029964, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12854 + }, + { + "epoch": 5.135836995605274, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12855 + }, + { + "epoch": 5.136236516180583, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12856 + }, + { + "epoch": 5.136636036755893, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12857 + }, + { + "epoch": 5.1370355573312025, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12858 + }, + { + "epoch": 5.137435077906512, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12859 + }, + { + "epoch": 5.137834598481822, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12860 + }, + { + "epoch": 5.138234119057131, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12861 + }, + { + "epoch": 5.138633639632441, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12862 + }, + { + "epoch": 5.139033160207751, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12863 + }, + { + "epoch": 5.13943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12864 + }, + { + "epoch": 5.13983220135837, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12865 + }, + { + "epoch": 5.14023172193368, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12866 + }, + { + "epoch": 5.140631242508989, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12867 + }, + { + "epoch": 5.141030763084299, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12868 + }, + { + "epoch": 5.141430283659608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12869 + }, + { + "epoch": 5.141829804234918, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12870 + }, + { + "epoch": 5.142229324810228, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12871 + }, + { + "epoch": 5.142628845385538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12872 + }, + { + "epoch": 5.143028365960847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12873 + }, + { + "epoch": 5.1434278865361565, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12874 + }, + { + "epoch": 5.143827407111466, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12875 + }, + { + "epoch": 5.144226927686776, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12876 + }, + { + "epoch": 5.144626448262086, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12877 + }, + { + "epoch": 5.145025968837395, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12878 + }, + { + "epoch": 5.1454254894127045, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12879 + }, + { + "epoch": 5.145825009988014, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12880 + }, + { + "epoch": 5.146224530563324, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12881 + }, + { + "epoch": 5.146624051138634, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12882 + }, + { + "epoch": 5.147023571713944, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12883 + }, + { + "epoch": 5.1474230922892525, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12884 + }, + { + "epoch": 5.147822612864562, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12885 + }, + { + "epoch": 5.148222133439872, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12886 + }, + { + "epoch": 5.148621654015182, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12887 + }, + { + "epoch": 5.149021174590492, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12888 + }, + { + "epoch": 5.1494206951658015, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12889 + }, + { + "epoch": 5.14982021574111, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 12890 + }, + { + "epoch": 5.15021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12891 + }, + { + "epoch": 5.15061925689173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12892 + }, + { + "epoch": 5.15101877746704, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12893 + }, + { + "epoch": 5.1514182980423495, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12894 + }, + { + "epoch": 5.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12895 + }, + { + "epoch": 5.152217339192968, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12896 + }, + { + "epoch": 5.152616859768278, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12897 + }, + { + "epoch": 5.153016380343588, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12898 + }, + { + "epoch": 5.153415900918898, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12899 + }, + { + "epoch": 5.153815421494207, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12900 + }, + { + "epoch": 5.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12901 + }, + { + "epoch": 5.154614462644826, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12902 + }, + { + "epoch": 5.155013983220136, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12903 + }, + { + "epoch": 5.155413503795446, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12904 + }, + { + "epoch": 5.155813024370755, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12905 + }, + { + "epoch": 5.156212544946065, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12906 + }, + { + "epoch": 5.156612065521374, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12907 + }, + { + "epoch": 5.157011586096684, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 12908 + }, + { + "epoch": 5.157411106671994, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12909 + }, + { + "epoch": 5.157810627247303, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12910 + }, + { + "epoch": 5.158210147822613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12911 + }, + { + "epoch": 5.158609668397922, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12912 + }, + { + "epoch": 5.159009188973232, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12913 + }, + { + "epoch": 5.159408709548542, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12914 + }, + { + "epoch": 5.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12915 + }, + { + "epoch": 5.160207750699161, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12916 + }, + { + "epoch": 5.160607271274471, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12917 + }, + { + "epoch": 5.16100679184978, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12918 + }, + { + "epoch": 5.16140631242509, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12919 + }, + { + "epoch": 5.1618058330003995, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12920 + }, + { + "epoch": 5.162205353575709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12921 + }, + { + "epoch": 5.162604874151019, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12922 + }, + { + "epoch": 5.163004394726328, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12923 + }, + { + "epoch": 5.163403915301638, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12924 + }, + { + "epoch": 5.163803435876948, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12925 + }, + { + "epoch": 5.164202956452257, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12926 + }, + { + "epoch": 5.164602477027567, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12927 + }, + { + "epoch": 5.165001997602877, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12928 + }, + { + "epoch": 5.165401518178186, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12929 + }, + { + "epoch": 5.165801038753496, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12930 + }, + { + "epoch": 5.166200559328805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12931 + }, + { + "epoch": 5.166600079904115, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12932 + }, + { + "epoch": 5.166999600479425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12933 + }, + { + "epoch": 5.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12934 + }, + { + "epoch": 5.167798641630044, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12935 + }, + { + "epoch": 5.168198162205353, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12936 + }, + { + "epoch": 5.168597682780663, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12937 + }, + { + "epoch": 5.168997203355973, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12938 + }, + { + "epoch": 5.169396723931283, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12939 + }, + { + "epoch": 5.169796244506592, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12940 + }, + { + "epoch": 5.1701957650819015, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12941 + }, + { + "epoch": 5.170595285657211, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12942 + }, + { + "epoch": 5.170994806232521, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12943 + }, + { + "epoch": 5.171394326807831, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 12944 + }, + { + "epoch": 5.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12945 + }, + { + "epoch": 5.1721933679584495, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12946 + }, + { + "epoch": 5.172592888533759, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12947 + }, + { + "epoch": 5.172992409109069, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12948 + }, + { + "epoch": 5.173391929684379, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 12949 + }, + { + "epoch": 5.173791450259689, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12950 + }, + { + "epoch": 5.174190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12951 + }, + { + "epoch": 5.174590491410307, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12952 + }, + { + "epoch": 5.174990011985617, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12953 + }, + { + "epoch": 5.175389532560927, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12954 + }, + { + "epoch": 5.175789053136237, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12955 + }, + { + "epoch": 5.1761885737115465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12956 + }, + { + "epoch": 5.176588094286855, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12957 + }, + { + "epoch": 5.176987614862165, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12958 + }, + { + "epoch": 5.177387135437475, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12959 + }, + { + "epoch": 5.177786656012785, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12960 + }, + { + "epoch": 5.1781861765880945, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12961 + }, + { + "epoch": 5.178585697163404, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12962 + }, + { + "epoch": 5.178985217738713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12963 + }, + { + "epoch": 5.179384738314023, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12964 + }, + { + "epoch": 5.179784258889333, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12965 + }, + { + "epoch": 5.180183779464643, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12966 + }, + { + "epoch": 5.180583300039952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12967 + }, + { + "epoch": 5.180982820615261, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12968 + }, + { + "epoch": 5.181382341190571, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12969 + }, + { + "epoch": 5.181781861765881, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12970 + }, + { + "epoch": 5.182181382341191, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12971 + }, + { + "epoch": 5.1825809029165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12972 + }, + { + "epoch": 5.18298042349181, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12973 + }, + { + "epoch": 5.183379944067119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12974 + }, + { + "epoch": 5.183779464642429, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12975 + }, + { + "epoch": 5.184178985217739, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12976 + }, + { + "epoch": 5.1845785057930485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12977 + }, + { + "epoch": 5.184978026368358, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12978 + }, + { + "epoch": 5.185377546943668, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 12979 + }, + { + "epoch": 5.185777067518977, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12980 + }, + { + "epoch": 5.186176588094287, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12981 + }, + { + "epoch": 5.1865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12982 + }, + { + "epoch": 5.186975629244906, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12983 + }, + { + "epoch": 5.187375149820216, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12984 + }, + { + "epoch": 5.187774670395525, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12985 + }, + { + "epoch": 5.188174190970835, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12986 + }, + { + "epoch": 5.1885737115461446, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12987 + }, + { + "epoch": 5.188973232121454, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12988 + }, + { + "epoch": 5.189372752696764, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12989 + }, + { + "epoch": 5.189772273272074, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12990 + }, + { + "epoch": 5.190171793847383, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12991 + }, + { + "epoch": 5.190571314422693, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12992 + }, + { + "epoch": 5.190970834998002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12993 + }, + { + "epoch": 5.191370355573312, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 12994 + }, + { + "epoch": 5.191769876148622, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12995 + }, + { + "epoch": 5.192169396723932, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12996 + }, + { + "epoch": 5.192568917299241, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12997 + }, + { + "epoch": 5.19296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12998 + }, + { + "epoch": 5.19336795844986, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12999 + }, + { + "epoch": 5.19376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13000 + }, + { + "epoch": 5.19416699960048, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13001 + }, + { + "epoch": 5.194566520175789, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13002 + }, + { + "epoch": 5.1949660407510985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13003 + }, + { + "epoch": 5.195365561326408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13004 + }, + { + "epoch": 5.195765081901718, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13005 + }, + { + "epoch": 5.196164602477028, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13006 + }, + { + "epoch": 5.196564123052338, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13007 + }, + { + "epoch": 5.1969636436276465, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13008 + }, + { + "epoch": 5.197363164202956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13009 + }, + { + "epoch": 5.197762684778266, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13010 + }, + { + "epoch": 5.198162205353576, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13011 + }, + { + "epoch": 5.198561725928886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13012 + }, + { + "epoch": 5.1989612465041946, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13013 + }, + { + "epoch": 5.199360767079504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13014 + }, + { + "epoch": 5.199760287654814, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13015 + }, + { + "epoch": 5.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13016 + }, + { + "epoch": 5.200559328805434, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13017 + }, + { + "epoch": 5.2009588493807435, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13018 + }, + { + "epoch": 5.201358369956052, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13019 + }, + { + "epoch": 5.201757890531362, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13020 + }, + { + "epoch": 5.202157411106672, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13021 + }, + { + "epoch": 5.202556931681982, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13022 + }, + { + "epoch": 5.2029564522572915, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 13023 + }, + { + "epoch": 5.2033559728326, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13024 + }, + { + "epoch": 5.20375549340791, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13025 + }, + { + "epoch": 5.20415501398322, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13026 + }, + { + "epoch": 5.20455453455853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 13027 + }, + { + "epoch": 5.20495405513384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13028 + }, + { + "epoch": 5.205353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13029 + }, + { + "epoch": 5.205753096284458, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13030 + }, + { + "epoch": 5.206152616859768, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13031 + }, + { + "epoch": 5.206552137435078, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13032 + }, + { + "epoch": 5.206951658010388, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13033 + }, + { + "epoch": 5.207351178585697, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13034 + }, + { + "epoch": 5.207750699161007, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13035 + }, + { + "epoch": 5.208150219736316, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13036 + }, + { + "epoch": 5.208549740311626, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13037 + }, + { + "epoch": 5.208949260886936, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 13038 + }, + { + "epoch": 5.2093487814622454, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13039 + }, + { + "epoch": 5.209748302037555, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13040 + }, + { + "epoch": 5.210147822612864, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13041 + }, + { + "epoch": 5.210547343188174, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13042 + }, + { + "epoch": 5.210946863763484, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13043 + }, + { + "epoch": 5.2113463843387935, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13044 + }, + { + "epoch": 5.211745904914103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13045 + }, + { + "epoch": 5.212145425489413, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13046 + }, + { + "epoch": 5.212544946064722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13047 + }, + { + "epoch": 5.212944466640032, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13048 + }, + { + "epoch": 5.2133439872153415, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 13049 + }, + { + "epoch": 5.213743507790651, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13050 + }, + { + "epoch": 5.214143028365961, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13051 + }, + { + "epoch": 5.214542548941271, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13052 + }, + { + "epoch": 5.21494206951658, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13053 + }, + { + "epoch": 5.21534159009189, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13054 + }, + { + "epoch": 5.215741110667199, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13055 + }, + { + "epoch": 5.216140631242509, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 13056 + }, + { + "epoch": 5.216540151817819, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13057 + }, + { + "epoch": 5.216939672393128, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13058 + }, + { + "epoch": 5.217339192968438, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13059 + }, + { + "epoch": 5.217738713543747, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13060 + }, + { + "epoch": 5.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13061 + }, + { + "epoch": 5.218537754694367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 13062 + }, + { + "epoch": 5.218937275269677, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13063 + }, + { + "epoch": 5.219336795844986, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13064 + }, + { + "epoch": 5.2197363164202955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13065 + }, + { + "epoch": 5.220135836995605, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13066 + }, + { + "epoch": 5.220535357570915, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13067 + }, + { + "epoch": 5.220934878146225, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13068 + }, + { + "epoch": 5.221334398721535, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13069 + }, + { + "epoch": 5.2217339192968435, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13070 + }, + { + "epoch": 5.222133439872153, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13071 + }, + { + "epoch": 5.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13072 + }, + { + "epoch": 5.222932481022773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13073 + }, + { + "epoch": 5.223332001598083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13074 + }, + { + "epoch": 5.2237315221733915, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13075 + }, + { + "epoch": 5.224131042748701, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13076 + }, + { + "epoch": 5.224530563324011, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13077 + }, + { + "epoch": 5.224930083899321, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13078 + }, + { + "epoch": 5.225329604474631, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13079 + }, + { + "epoch": 5.2257291250499405, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13080 + }, + { + "epoch": 5.226128645625249, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13081 + }, + { + "epoch": 5.226528166200559, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13082 + }, + { + "epoch": 5.226927686775869, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13083 + }, + { + "epoch": 5.227327207351179, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13084 + }, + { + "epoch": 5.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13085 + }, + { + "epoch": 5.228126248501797, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13086 + }, + { + "epoch": 5.228525769077107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13087 + }, + { + "epoch": 5.228925289652417, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13088 + }, + { + "epoch": 5.229324810227727, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13089 + }, + { + "epoch": 5.229724330803037, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13090 + }, + { + "epoch": 5.230123851378346, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13091 + }, + { + "epoch": 5.230523371953655, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13092 + }, + { + "epoch": 5.230922892528965, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13093 + }, + { + "epoch": 5.231322413104275, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13094 + }, + { + "epoch": 5.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13095 + }, + { + "epoch": 5.232121454254894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13096 + }, + { + "epoch": 5.232520974830204, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13097 + }, + { + "epoch": 5.232920495405513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13098 + }, + { + "epoch": 5.233320015980823, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13099 + }, + { + "epoch": 5.233719536556133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13100 + }, + { + "epoch": 5.234119057131442, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13101 + }, + { + "epoch": 5.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13102 + }, + { + "epoch": 5.234918098282061, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13103 + }, + { + "epoch": 5.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13104 + }, + { + "epoch": 5.235717139432681, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13105 + }, + { + "epoch": 5.2361166600079905, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13106 + }, + { + "epoch": 5.2365161805833, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13107 + }, + { + "epoch": 5.23691570115861, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13108 + }, + { + "epoch": 5.237315221733919, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13109 + }, + { + "epoch": 5.237714742309229, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13110 + }, + { + "epoch": 5.2381142628845385, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13111 + }, + { + "epoch": 5.238513783459848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13112 + }, + { + "epoch": 5.238913304035158, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13113 + }, + { + "epoch": 5.239312824610467, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13114 + }, + { + "epoch": 5.239712345185777, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13115 + }, + { + "epoch": 5.240111865761087, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13116 + }, + { + "epoch": 5.240511386336396, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13117 + }, + { + "epoch": 5.240910906911706, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13118 + }, + { + "epoch": 5.241310427487016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13119 + }, + { + "epoch": 5.241709948062325, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13120 + }, + { + "epoch": 5.242109468637635, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13121 + }, + { + "epoch": 5.242508989212944, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13122 + }, + { + "epoch": 5.242908509788254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13123 + }, + { + "epoch": 5.243308030363564, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13124 + }, + { + "epoch": 5.243707550938874, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13125 + }, + { + "epoch": 5.244107071514183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13126 + }, + { + "epoch": 5.244506592089492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13127 + }, + { + "epoch": 5.244906112664802, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13128 + }, + { + "epoch": 5.245305633240112, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 13129 + }, + { + "epoch": 5.245705153815422, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13130 + }, + { + "epoch": 5.246104674390731, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13131 + }, + { + "epoch": 5.2465041949660405, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13132 + }, + { + "epoch": 5.24690371554135, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13133 + }, + { + "epoch": 5.24730323611666, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13134 + }, + { + "epoch": 5.24770275669197, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13135 + }, + { + "epoch": 5.24810227726728, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13136 + }, + { + "epoch": 5.2485017978425885, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13137 + }, + { + "epoch": 5.248901318417898, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13138 + }, + { + "epoch": 5.249300838993208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13139 + }, + { + "epoch": 5.249700359568518, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13140 + }, + { + "epoch": 5.250099880143828, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13141 + }, + { + "epoch": 5.2504994007191375, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13142 + }, + { + "epoch": 5.250898921294446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13143 + }, + { + "epoch": 5.251298441869756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13144 + }, + { + "epoch": 5.251697962445066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13145 + }, + { + "epoch": 5.252097483020376, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13146 + }, + { + "epoch": 5.2524970035956855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13147 + }, + { + "epoch": 5.252896524170994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13148 + }, + { + "epoch": 5.253296044746304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13149 + }, + { + "epoch": 5.253695565321614, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13150 + }, + { + "epoch": 5.254095085896924, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13151 + }, + { + "epoch": 5.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13152 + }, + { + "epoch": 5.254894127047543, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13153 + }, + { + "epoch": 5.255293647622852, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13154 + }, + { + "epoch": 5.255693168198162, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13155 + }, + { + "epoch": 5.256092688773472, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13156 + }, + { + "epoch": 5.256492209348782, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13157 + }, + { + "epoch": 5.256891729924091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13158 + }, + { + "epoch": 5.257291250499401, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13159 + }, + { + "epoch": 5.25769077107471, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13160 + }, + { + "epoch": 5.25809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13161 + }, + { + "epoch": 5.25848981222533, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13162 + }, + { + "epoch": 5.258889332800639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13163 + }, + { + "epoch": 5.259288853375949, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13164 + }, + { + "epoch": 5.259688373951258, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13165 + }, + { + "epoch": 5.260087894526568, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13166 + }, + { + "epoch": 5.260487415101878, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13167 + }, + { + "epoch": 5.2608869356771875, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13168 + }, + { + "epoch": 5.261286456252497, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13169 + }, + { + "epoch": 5.261685976827807, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13170 + }, + { + "epoch": 5.262085497403116, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13171 + }, + { + "epoch": 5.262485017978426, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13172 + }, + { + "epoch": 5.2628845385537355, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13173 + }, + { + "epoch": 5.263284059129045, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13174 + }, + { + "epoch": 5.263683579704355, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13175 + }, + { + "epoch": 5.264083100279665, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13176 + }, + { + "epoch": 5.264482620854974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13177 + }, + { + "epoch": 5.2648821414302835, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13178 + }, + { + "epoch": 5.265281662005593, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13179 + }, + { + "epoch": 5.265681182580903, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13180 + }, + { + "epoch": 5.266080703156213, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13181 + }, + { + "epoch": 5.266480223731522, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13182 + }, + { + "epoch": 5.266879744306832, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13183 + }, + { + "epoch": 5.267279264882141, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13184 + }, + { + "epoch": 5.267678785457451, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13185 + }, + { + "epoch": 5.268078306032761, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13186 + }, + { + "epoch": 5.26847782660807, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13187 + }, + { + "epoch": 5.26887734718338, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13188 + }, + { + "epoch": 5.269276867758689, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13189 + }, + { + "epoch": 5.269676388333999, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13190 + }, + { + "epoch": 5.270075908909309, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13191 + }, + { + "epoch": 5.270475429484619, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13192 + }, + { + "epoch": 5.270874950059928, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13193 + }, + { + "epoch": 5.2712744706352375, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13194 + }, + { + "epoch": 5.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13195 + }, + { + "epoch": 5.272073511785857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13196 + }, + { + "epoch": 5.272473032361167, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13197 + }, + { + "epoch": 5.272872552936477, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13198 + }, + { + "epoch": 5.2732720735117855, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13199 + }, + { + "epoch": 5.273671594087095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13200 + }, + { + "epoch": 5.274071114662405, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13201 + }, + { + "epoch": 5.274470635237715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13202 + }, + { + "epoch": 5.274870155813025, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13203 + }, + { + "epoch": 5.2752696763883336, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 13204 + }, + { + "epoch": 5.275669196963643, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13205 + }, + { + "epoch": 5.276068717538953, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13206 + }, + { + "epoch": 5.276468238114263, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13207 + }, + { + "epoch": 5.276867758689573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13208 + }, + { + "epoch": 5.2772672792648825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13209 + }, + { + "epoch": 5.277666799840191, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13210 + }, + { + "epoch": 5.278066320415501, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13211 + }, + { + "epoch": 5.278465840990811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 13212 + }, + { + "epoch": 5.278865361566121, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13213 + }, + { + "epoch": 5.2792648821414305, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13214 + }, + { + "epoch": 5.27966440271674, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13215 + }, + { + "epoch": 5.280063923292049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13216 + }, + { + "epoch": 5.280463443867359, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13217 + }, + { + "epoch": 5.280862964442669, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13218 + }, + { + "epoch": 5.281262485017979, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13219 + }, + { + "epoch": 5.281662005593288, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13220 + }, + { + "epoch": 5.282061526168597, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13221 + }, + { + "epoch": 5.282461046743907, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13222 + }, + { + "epoch": 5.282860567319217, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13223 + }, + { + "epoch": 5.283260087894527, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13224 + }, + { + "epoch": 5.283659608469836, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13225 + }, + { + "epoch": 5.284059129045146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13226 + }, + { + "epoch": 5.284458649620455, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13227 + }, + { + "epoch": 5.284858170195765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 13228 + }, + { + "epoch": 5.285257690771075, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 13229 + }, + { + "epoch": 5.285657211346384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13230 + }, + { + "epoch": 5.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13231 + }, + { + "epoch": 5.286456252497004, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13232 + }, + { + "epoch": 5.286855773072313, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 13233 + }, + { + "epoch": 5.287255293647623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13234 + }, + { + "epoch": 5.2876548142229325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13235 + }, + { + "epoch": 5.288054334798242, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13236 + }, + { + "epoch": 5.288453855373552, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13237 + }, + { + "epoch": 5.288853375948861, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13238 + }, + { + "epoch": 5.289252896524171, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13239 + }, + { + "epoch": 5.2896524170994805, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13240 + }, + { + "epoch": 5.29005193767479, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13241 + }, + { + "epoch": 5.2904514582501, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13242 + }, + { + "epoch": 5.29085097882541, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13243 + }, + { + "epoch": 5.291250499400719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 13244 + }, + { + "epoch": 5.291650019976029, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13245 + }, + { + "epoch": 5.292049540551338, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13246 + }, + { + "epoch": 5.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13247 + }, + { + "epoch": 5.292848581701958, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13248 + }, + { + "epoch": 5.293248102277268, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13249 + }, + { + "epoch": 5.293647622852577, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13250 + }, + { + "epoch": 5.294047143427886, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13251 + }, + { + "epoch": 5.294446664003196, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13252 + }, + { + "epoch": 5.294846184578506, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13253 + }, + { + "epoch": 5.295245705153816, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13254 + }, + { + "epoch": 5.295645225729125, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13255 + }, + { + "epoch": 5.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13256 + }, + { + "epoch": 5.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.2965, + "step": 13257 + }, + { + "epoch": 5.296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13258 + }, + { + "epoch": 5.297243308030364, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13259 + }, + { + "epoch": 5.297642828605674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13260 + }, + { + "epoch": 5.2980423491809825, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13261 + }, + { + "epoch": 5.298441869756292, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13262 + }, + { + "epoch": 5.298841390331602, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13263 + }, + { + "epoch": 5.299240910906912, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13264 + }, + { + "epoch": 5.299640431482222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13265 + }, + { + "epoch": 5.300039952057531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13266 + }, + { + "epoch": 5.30043947263284, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13267 + }, + { + "epoch": 5.30083899320815, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13268 + }, + { + "epoch": 5.30123851378346, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13269 + }, + { + "epoch": 5.30163803435877, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13270 + }, + { + "epoch": 5.3020375549340795, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13271 + }, + { + "epoch": 5.302437075509388, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13272 + }, + { + "epoch": 5.302836596084698, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13273 + }, + { + "epoch": 5.303236116660008, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13274 + }, + { + "epoch": 5.303635637235318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13275 + }, + { + "epoch": 5.3040351578106275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13276 + }, + { + "epoch": 5.304434678385936, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13277 + }, + { + "epoch": 5.304834198961246, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13278 + }, + { + "epoch": 5.305233719536556, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13279 + }, + { + "epoch": 5.305633240111866, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13280 + }, + { + "epoch": 5.3060327606871756, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13281 + }, + { + "epoch": 5.306432281262485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13282 + }, + { + "epoch": 5.306831801837794, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13283 + }, + { + "epoch": 5.307231322413104, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13284 + }, + { + "epoch": 5.307630842988414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13285 + }, + { + "epoch": 5.308030363563724, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13286 + }, + { + "epoch": 5.308429884139033, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13287 + }, + { + "epoch": 5.308829404714343, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13288 + }, + { + "epoch": 5.309228925289652, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13289 + }, + { + "epoch": 5.309628445864962, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13290 + }, + { + "epoch": 5.310027966440272, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13291 + }, + { + "epoch": 5.310427487015581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 13292 + }, + { + "epoch": 5.310827007590891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13293 + }, + { + "epoch": 5.3112265281662, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13294 + }, + { + "epoch": 5.31162604874151, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13295 + }, + { + "epoch": 5.31202556931682, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13296 + }, + { + "epoch": 5.3124250898921295, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13297 + }, + { + "epoch": 5.312824610467439, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13298 + }, + { + "epoch": 5.313224131042749, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13299 + }, + { + "epoch": 5.313623651618058, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13300 + }, + { + "epoch": 5.314023172193368, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13301 + }, + { + "epoch": 5.3144226927686775, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13302 + }, + { + "epoch": 5.314822213343987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13303 + }, + { + "epoch": 5.315221733919297, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13304 + }, + { + "epoch": 5.315621254494607, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13305 + }, + { + "epoch": 5.316020775069916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13306 + }, + { + "epoch": 5.316420295645226, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13307 + }, + { + "epoch": 5.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13308 + }, + { + "epoch": 5.317219336795845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13309 + }, + { + "epoch": 5.317618857371155, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13310 + }, + { + "epoch": 5.318018377946464, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13311 + }, + { + "epoch": 5.318417898521774, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13312 + }, + { + "epoch": 5.318817419097083, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13313 + }, + { + "epoch": 5.319216939672393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 13314 + }, + { + "epoch": 5.319616460247703, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13315 + }, + { + "epoch": 5.320015980823013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 13316 + }, + { + "epoch": 5.320415501398322, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13317 + }, + { + "epoch": 5.320815021973631, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13318 + }, + { + "epoch": 5.321214542548941, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13319 + }, + { + "epoch": 5.321614063124251, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13320 + }, + { + "epoch": 5.322013583699561, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13321 + }, + { + "epoch": 5.322413104274871, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13322 + }, + { + "epoch": 5.3228126248501795, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13323 + }, + { + "epoch": 5.323212145425489, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13324 + }, + { + "epoch": 5.323611666000799, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13325 + }, + { + "epoch": 5.324011186576109, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13326 + }, + { + "epoch": 5.324410707151419, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13327 + }, + { + "epoch": 5.3248102277267275, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13328 + }, + { + "epoch": 5.325209748302037, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13329 + }, + { + "epoch": 5.325609268877347, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13330 + }, + { + "epoch": 5.326008789452657, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13331 + }, + { + "epoch": 5.326408310027967, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13332 + }, + { + "epoch": 5.3268078306032765, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13333 + }, + { + "epoch": 5.327207351178585, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13334 + }, + { + "epoch": 5.327606871753895, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13335 + }, + { + "epoch": 5.328006392329205, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 13336 + }, + { + "epoch": 5.328405912904515, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13337 + }, + { + "epoch": 5.3288054334798245, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13338 + }, + { + "epoch": 5.329204954055134, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13339 + }, + { + "epoch": 5.329604474630443, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13340 + }, + { + "epoch": 5.330003995205753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13341 + }, + { + "epoch": 5.330403515781063, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13342 + }, + { + "epoch": 5.3308030363563725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13343 + }, + { + "epoch": 5.331202556931682, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13344 + }, + { + "epoch": 5.331602077506991, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13345 + }, + { + "epoch": 5.332001598082301, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13346 + }, + { + "epoch": 5.332401118657611, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13347 + }, + { + "epoch": 5.332800639232921, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13348 + }, + { + "epoch": 5.33320015980823, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13349 + }, + { + "epoch": 5.33359968038354, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13350 + }, + { + "epoch": 5.333999200958849, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13351 + }, + { + "epoch": 5.334398721534159, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13352 + }, + { + "epoch": 5.334798242109469, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13353 + }, + { + "epoch": 5.335197762684778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13354 + }, + { + "epoch": 5.335597283260088, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13355 + }, + { + "epoch": 5.335996803835398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13356 + }, + { + "epoch": 5.336396324410707, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13357 + }, + { + "epoch": 5.336795844986017, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13358 + }, + { + "epoch": 5.3371953655613265, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13359 + }, + { + "epoch": 5.337594886136636, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13360 + }, + { + "epoch": 5.337994406711946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13361 + }, + { + "epoch": 5.338393927287255, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13362 + }, + { + "epoch": 5.338793447862565, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13363 + }, + { + "epoch": 5.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13364 + }, + { + "epoch": 5.339592489013184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13365 + }, + { + "epoch": 5.339992009588494, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13366 + }, + { + "epoch": 5.340391530163803, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13367 + }, + { + "epoch": 5.340791050739113, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13368 + }, + { + "epoch": 5.3411905713144225, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13369 + }, + { + "epoch": 5.341590091889732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13370 + }, + { + "epoch": 5.341989612465042, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13371 + }, + { + "epoch": 5.342389133040352, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13372 + }, + { + "epoch": 5.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13373 + }, + { + "epoch": 5.343188174190971, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 13374 + }, + { + "epoch": 5.34358769476628, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13375 + }, + { + "epoch": 5.34398721534159, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13376 + }, + { + "epoch": 5.3443867359169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13377 + }, + { + "epoch": 5.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13378 + }, + { + "epoch": 5.345185777067519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13379 + }, + { + "epoch": 5.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13380 + }, + { + "epoch": 5.345984818218138, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13381 + }, + { + "epoch": 5.346384338793448, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13382 + }, + { + "epoch": 5.346783859368758, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13383 + }, + { + "epoch": 5.347183379944067, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13384 + }, + { + "epoch": 5.3475829005193765, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13385 + }, + { + "epoch": 5.347982421094686, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13386 + }, + { + "epoch": 5.348381941669996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13387 + }, + { + "epoch": 5.348781462245306, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13388 + }, + { + "epoch": 5.349180982820616, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13389 + }, + { + "epoch": 5.3495805033959245, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13390 + }, + { + "epoch": 5.349980023971234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13391 + }, + { + "epoch": 5.350379544546544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13392 + }, + { + "epoch": 5.350779065121854, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13393 + }, + { + "epoch": 5.351178585697164, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13394 + }, + { + "epoch": 5.351578106272473, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13395 + }, + { + "epoch": 5.351977626847782, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13396 + }, + { + "epoch": 5.352377147423092, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13397 + }, + { + "epoch": 5.352776667998402, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13398 + }, + { + "epoch": 5.353176188573712, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13399 + }, + { + "epoch": 5.3535757091490215, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13400 + }, + { + "epoch": 5.35397522972433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13401 + }, + { + "epoch": 5.35437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13402 + }, + { + "epoch": 5.35477427087495, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13403 + }, + { + "epoch": 5.35517379145026, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13404 + }, + { + "epoch": 5.3555733120255695, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13405 + }, + { + "epoch": 5.355972832600879, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13406 + }, + { + "epoch": 5.356372353176188, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13407 + }, + { + "epoch": 5.356771873751498, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 13408 + }, + { + "epoch": 5.357171394326808, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13409 + }, + { + "epoch": 5.357570914902118, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13410 + }, + { + "epoch": 5.357970435477427, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13411 + }, + { + "epoch": 5.358369956052737, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13412 + }, + { + "epoch": 5.358769476628046, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13413 + }, + { + "epoch": 5.359168997203356, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13414 + }, + { + "epoch": 5.359568517778666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13415 + }, + { + "epoch": 5.359968038353975, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 13416 + }, + { + "epoch": 5.360367558929285, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13417 + }, + { + "epoch": 5.360767079504594, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13418 + }, + { + "epoch": 5.361166600079904, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13419 + }, + { + "epoch": 5.361566120655214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13420 + }, + { + "epoch": 5.361965641230523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13421 + }, + { + "epoch": 5.362365161805833, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13422 + }, + { + "epoch": 5.362764682381143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13423 + }, + { + "epoch": 5.363164202956452, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13424 + }, + { + "epoch": 5.363563723531762, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 13425 + }, + { + "epoch": 5.3639632441070715, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13426 + }, + { + "epoch": 5.364362764682381, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13427 + }, + { + "epoch": 5.364762285257691, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13428 + }, + { + "epoch": 5.365161805833001, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13429 + }, + { + "epoch": 5.36556132640831, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 13430 + }, + { + "epoch": 5.3659608469836195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13431 + }, + { + "epoch": 5.366360367558929, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13432 + }, + { + "epoch": 5.366759888134239, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13433 + }, + { + "epoch": 5.367159408709549, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13434 + }, + { + "epoch": 5.367558929284858, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 13435 + }, + { + "epoch": 5.367958449860168, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13436 + }, + { + "epoch": 5.368357970435477, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 13437 + }, + { + "epoch": 5.368757491010787, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13438 + }, + { + "epoch": 5.369157011586097, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13439 + }, + { + "epoch": 5.369556532161407, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13440 + }, + { + "epoch": 5.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13441 + }, + { + "epoch": 5.370355573312025, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13442 + }, + { + "epoch": 5.370755093887335, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 13443 + }, + { + "epoch": 5.371154614462645, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13444 + }, + { + "epoch": 5.371554135037955, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13445 + }, + { + "epoch": 5.3719536556132645, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13446 + }, + { + "epoch": 5.372353176188573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13447 + }, + { + "epoch": 5.372752696763883, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13448 + }, + { + "epoch": 5.373152217339193, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13449 + }, + { + "epoch": 5.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13450 + }, + { + "epoch": 5.373951258489813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13451 + }, + { + "epoch": 5.3743507790651215, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13452 + }, + { + "epoch": 5.374750299640431, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13453 + }, + { + "epoch": 5.375149820215741, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13454 + }, + { + "epoch": 5.375549340791051, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13455 + }, + { + "epoch": 5.375948861366361, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13456 + }, + { + "epoch": 5.3763483819416695, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13457 + }, + { + "epoch": 5.376747902516979, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13458 + }, + { + "epoch": 5.377147423092289, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13459 + }, + { + "epoch": 5.377546943667599, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13460 + }, + { + "epoch": 5.377946464242909, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13461 + }, + { + "epoch": 5.3783459848182185, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13462 + }, + { + "epoch": 5.378745505393527, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13463 + }, + { + "epoch": 5.379145025968837, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13464 + }, + { + "epoch": 5.379544546544147, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13465 + }, + { + "epoch": 5.379944067119457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13466 + }, + { + "epoch": 5.3803435876947665, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13467 + }, + { + "epoch": 5.380743108270076, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13468 + }, + { + "epoch": 5.381142628845385, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13469 + }, + { + "epoch": 5.381542149420695, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13470 + }, + { + "epoch": 5.381941669996005, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13471 + }, + { + "epoch": 5.3823411905713145, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13472 + }, + { + "epoch": 5.382740711146624, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13473 + }, + { + "epoch": 5.383140231721933, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13474 + }, + { + "epoch": 5.383539752297243, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13475 + }, + { + "epoch": 5.383939272872553, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13476 + }, + { + "epoch": 5.384338793447863, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13477 + }, + { + "epoch": 5.384738314023172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13478 + }, + { + "epoch": 5.385137834598482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13479 + }, + { + "epoch": 5.385537355173791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13480 + }, + { + "epoch": 5.385936875749101, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13481 + }, + { + "epoch": 5.386336396324411, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13482 + }, + { + "epoch": 5.38673591689972, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13483 + }, + { + "epoch": 5.38713543747503, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13484 + }, + { + "epoch": 5.38753495805034, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13485 + }, + { + "epoch": 5.387934478625649, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13486 + }, + { + "epoch": 5.388333999200959, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 13487 + }, + { + "epoch": 5.3887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13488 + }, + { + "epoch": 5.389133040351578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13489 + }, + { + "epoch": 5.389532560926888, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13490 + }, + { + "epoch": 5.389932081502197, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13491 + }, + { + "epoch": 5.390331602077507, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13492 + }, + { + "epoch": 5.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13493 + }, + { + "epoch": 5.391130643228126, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13494 + }, + { + "epoch": 5.391530163803436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13495 + }, + { + "epoch": 5.391929684378746, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13496 + }, + { + "epoch": 5.392329204954055, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13497 + }, + { + "epoch": 5.3927287255293646, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 13498 + }, + { + "epoch": 5.393128246104674, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13499 + }, + { + "epoch": 5.393527766679984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13500 + }, + { + "epoch": 5.393927287255294, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13501 + }, + { + "epoch": 5.394326807830604, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13502 + }, + { + "epoch": 5.394726328405913, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13503 + }, + { + "epoch": 5.395125848981222, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13504 + }, + { + "epoch": 5.395525369556532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13505 + }, + { + "epoch": 5.395924890131842, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13506 + }, + { + "epoch": 5.396324410707152, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13507 + }, + { + "epoch": 5.396723931282461, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13508 + }, + { + "epoch": 5.39712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13509 + }, + { + "epoch": 5.39752297243308, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13510 + }, + { + "epoch": 5.39792249300839, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13511 + }, + { + "epoch": 5.3983220135837, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13512 + }, + { + "epoch": 5.39872153415901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13513 + }, + { + "epoch": 5.3991210547343185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13514 + }, + { + "epoch": 5.399520575309628, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13515 + }, + { + "epoch": 5.399920095884938, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13516 + }, + { + "epoch": 5.400319616460248, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13517 + }, + { + "epoch": 5.400719137035558, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13518 + }, + { + "epoch": 5.401118657610867, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13519 + }, + { + "epoch": 5.401518178186176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13520 + }, + { + "epoch": 5.401917698761486, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13521 + }, + { + "epoch": 5.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13522 + }, + { + "epoch": 5.402716739912106, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13523 + }, + { + "epoch": 5.4031162604874154, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13524 + }, + { + "epoch": 5.403515781062724, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13525 + }, + { + "epoch": 5.403915301638034, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13526 + }, + { + "epoch": 5.404314822213344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13527 + }, + { + "epoch": 5.404714342788654, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13528 + }, + { + "epoch": 5.4051138633639635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13529 + }, + { + "epoch": 5.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.2988, + "step": 13530 + }, + { + "epoch": 5.405912904514582, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13531 + }, + { + "epoch": 5.406312425089892, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13532 + }, + { + "epoch": 5.406711945665202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13533 + }, + { + "epoch": 5.4071114662405115, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13534 + }, + { + "epoch": 5.407510986815821, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13535 + }, + { + "epoch": 5.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13536 + }, + { + "epoch": 5.40831002796644, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 13537 + }, + { + "epoch": 5.40870954854175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13538 + }, + { + "epoch": 5.40910906911706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13539 + }, + { + "epoch": 5.409508589692369, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13540 + }, + { + "epoch": 5.409908110267679, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13541 + }, + { + "epoch": 5.410307630842988, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13542 + }, + { + "epoch": 5.410707151418298, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13543 + }, + { + "epoch": 5.411106671993608, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13544 + }, + { + "epoch": 5.411506192568917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13545 + }, + { + "epoch": 5.411905713144227, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13546 + }, + { + "epoch": 5.412305233719536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13547 + }, + { + "epoch": 5.412704754294846, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13548 + }, + { + "epoch": 5.413104274870156, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13549 + }, + { + "epoch": 5.4135037954454654, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13550 + }, + { + "epoch": 5.413903316020775, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13551 + }, + { + "epoch": 5.414302836596085, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13552 + }, + { + "epoch": 5.414702357171394, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 13553 + }, + { + "epoch": 5.415101877746704, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13554 + }, + { + "epoch": 5.4155013983220135, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13555 + }, + { + "epoch": 5.415900918897323, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 13556 + }, + { + "epoch": 5.416300439472633, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13557 + }, + { + "epoch": 5.416699960047943, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13558 + }, + { + "epoch": 5.417099480623252, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13559 + }, + { + "epoch": 5.4174990011985615, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 13560 + }, + { + "epoch": 5.417898521773871, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13561 + }, + { + "epoch": 5.418298042349181, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13562 + }, + { + "epoch": 5.418697562924491, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13563 + }, + { + "epoch": 5.4190970834998, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13564 + }, + { + "epoch": 5.41949660407511, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13565 + }, + { + "epoch": 5.419896124650419, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13566 + }, + { + "epoch": 5.420295645225729, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13567 + }, + { + "epoch": 5.420695165801039, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13568 + }, + { + "epoch": 5.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13569 + }, + { + "epoch": 5.421494206951658, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13570 + }, + { + "epoch": 5.421893727526967, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13571 + }, + { + "epoch": 5.422293248102277, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13572 + }, + { + "epoch": 5.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13573 + }, + { + "epoch": 5.423092289252897, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13574 + }, + { + "epoch": 5.4234918098282066, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13575 + }, + { + "epoch": 5.4238913304035155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13576 + }, + { + "epoch": 5.424290850978825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13577 + }, + { + "epoch": 5.424690371554135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13578 + }, + { + "epoch": 5.425089892129445, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13579 + }, + { + "epoch": 5.425489412704755, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13580 + }, + { + "epoch": 5.4258889332800635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13581 + }, + { + "epoch": 5.426288453855373, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13582 + }, + { + "epoch": 5.426687974430683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13583 + }, + { + "epoch": 5.427087495005993, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13584 + }, + { + "epoch": 5.427487015581303, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13585 + }, + { + "epoch": 5.427886536156612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13586 + }, + { + "epoch": 5.428286056731921, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13587 + }, + { + "epoch": 5.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 13588 + }, + { + "epoch": 5.429085097882541, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13589 + }, + { + "epoch": 5.429484618457851, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13590 + }, + { + "epoch": 5.4298841390331605, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13591 + }, + { + "epoch": 5.43028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13592 + }, + { + "epoch": 5.430683180183779, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13593 + }, + { + "epoch": 5.431082700759089, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13594 + }, + { + "epoch": 5.431482221334399, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13595 + }, + { + "epoch": 5.4318817419097085, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13596 + }, + { + "epoch": 5.432281262485018, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13597 + }, + { + "epoch": 5.432680783060327, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13598 + }, + { + "epoch": 5.433080303635637, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13599 + }, + { + "epoch": 5.433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13600 + }, + { + "epoch": 5.433879344786257, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13601 + }, + { + "epoch": 5.434278865361566, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13602 + }, + { + "epoch": 5.434678385936876, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 13603 + }, + { + "epoch": 5.435077906512185, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13604 + }, + { + "epoch": 5.435477427087495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13605 + }, + { + "epoch": 5.435876947662805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13606 + }, + { + "epoch": 5.436276468238114, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13607 + }, + { + "epoch": 5.436675988813424, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13608 + }, + { + "epoch": 5.437075509388734, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13609 + }, + { + "epoch": 5.437475029964043, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13610 + }, + { + "epoch": 5.437874550539353, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13611 + }, + { + "epoch": 5.438274071114662, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13612 + }, + { + "epoch": 5.438673591689972, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13613 + }, + { + "epoch": 5.439073112265282, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13614 + }, + { + "epoch": 5.439472632840591, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13615 + }, + { + "epoch": 5.439872153415901, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13616 + }, + { + "epoch": 5.4402716739912105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 13617 + }, + { + "epoch": 5.44067119456652, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13618 + }, + { + "epoch": 5.44107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 13619 + }, + { + "epoch": 5.44147023571714, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13620 + }, + { + "epoch": 5.441869756292449, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13621 + }, + { + "epoch": 5.4422692768677585, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13622 + }, + { + "epoch": 5.442668797443068, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13623 + }, + { + "epoch": 5.443068318018378, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13624 + }, + { + "epoch": 5.443467838593688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13625 + }, + { + "epoch": 5.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13626 + }, + { + "epoch": 5.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13627 + }, + { + "epoch": 5.444666400319616, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13628 + }, + { + "epoch": 5.445065920894926, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13629 + }, + { + "epoch": 5.445465441470236, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13630 + }, + { + "epoch": 5.445864962045546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13631 + }, + { + "epoch": 5.446264482620855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13632 + }, + { + "epoch": 5.446664003196164, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13633 + }, + { + "epoch": 5.447063523771474, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13634 + }, + { + "epoch": 5.447463044346784, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13635 + }, + { + "epoch": 5.447862564922094, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13636 + }, + { + "epoch": 5.448262085497403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13637 + }, + { + "epoch": 5.448661606072712, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13638 + }, + { + "epoch": 5.449061126648022, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13639 + }, + { + "epoch": 5.449460647223332, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13640 + }, + { + "epoch": 5.449860167798642, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13641 + }, + { + "epoch": 5.450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13642 + }, + { + "epoch": 5.4506592089492605, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13643 + }, + { + "epoch": 5.45105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13644 + }, + { + "epoch": 5.45145825009988, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 13645 + }, + { + "epoch": 5.45185777067519, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13646 + }, + { + "epoch": 5.4522572912505, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13647 + }, + { + "epoch": 5.452656811825809, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13648 + }, + { + "epoch": 5.453056332401118, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13649 + }, + { + "epoch": 5.453455852976428, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13650 + }, + { + "epoch": 5.453855373551738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13651 + }, + { + "epoch": 5.454254894127048, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 13652 + }, + { + "epoch": 5.4546544147023575, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13653 + }, + { + "epoch": 5.455053935277666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13654 + }, + { + "epoch": 5.455453455852976, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13655 + }, + { + "epoch": 5.455852976428286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13656 + }, + { + "epoch": 5.456252497003596, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13657 + }, + { + "epoch": 5.4566520175789055, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13658 + }, + { + "epoch": 5.457051538154215, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13659 + }, + { + "epoch": 5.457451058729524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13660 + }, + { + "epoch": 5.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13661 + }, + { + "epoch": 5.458250099880144, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13662 + }, + { + "epoch": 5.4586496204554535, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13663 + }, + { + "epoch": 5.459049141030763, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 13664 + }, + { + "epoch": 5.459448661606073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13665 + }, + { + "epoch": 5.459848182181382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13666 + }, + { + "epoch": 5.460247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13667 + }, + { + "epoch": 5.460647223332002, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 13668 + }, + { + "epoch": 5.461046743907311, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13669 + }, + { + "epoch": 5.461446264482621, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13670 + }, + { + "epoch": 5.46184578505793, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13671 + }, + { + "epoch": 5.46224530563324, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13672 + }, + { + "epoch": 5.46264482620855, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13673 + }, + { + "epoch": 5.463044346783859, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13674 + }, + { + "epoch": 5.463443867359169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13675 + }, + { + "epoch": 5.463843387934479, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13676 + }, + { + "epoch": 5.464242908509788, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13677 + }, + { + "epoch": 5.464642429085098, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13678 + }, + { + "epoch": 5.4650419496604075, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13679 + }, + { + "epoch": 5.465441470235717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13680 + }, + { + "epoch": 5.465840990811027, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13681 + }, + { + "epoch": 5.466240511386337, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13682 + }, + { + "epoch": 5.466640031961646, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13683 + }, + { + "epoch": 5.4670395525369555, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13684 + }, + { + "epoch": 5.467439073112265, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13685 + }, + { + "epoch": 5.467838593687575, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13686 + }, + { + "epoch": 5.468238114262885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13687 + }, + { + "epoch": 5.468637634838194, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13688 + }, + { + "epoch": 5.4690371554135035, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13689 + }, + { + "epoch": 5.469436675988813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13690 + }, + { + "epoch": 5.469836196564123, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13691 + }, + { + "epoch": 5.470235717139433, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13692 + }, + { + "epoch": 5.470635237714743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13693 + }, + { + "epoch": 5.471034758290052, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13694 + }, + { + "epoch": 5.471434278865361, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13695 + }, + { + "epoch": 5.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13696 + }, + { + "epoch": 5.472233320015981, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13697 + }, + { + "epoch": 5.472632840591291, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13698 + }, + { + "epoch": 5.4730323611666005, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13699 + }, + { + "epoch": 5.473431881741909, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13700 + }, + { + "epoch": 5.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13701 + }, + { + "epoch": 5.474230922892529, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13702 + }, + { + "epoch": 5.474630443467839, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13703 + }, + { + "epoch": 5.475029964043149, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 13704 + }, + { + "epoch": 5.4754294846184575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13705 + }, + { + "epoch": 5.475829005193767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13706 + }, + { + "epoch": 5.476228525769077, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13707 + }, + { + "epoch": 5.476628046344387, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13708 + }, + { + "epoch": 5.477027566919697, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13709 + }, + { + "epoch": 5.477427087495006, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13710 + }, + { + "epoch": 5.477826608070315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13711 + }, + { + "epoch": 5.478226128645625, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13712 + }, + { + "epoch": 5.478625649220935, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13713 + }, + { + "epoch": 5.479025169796245, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 13714 + }, + { + "epoch": 5.479424690371554, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13715 + }, + { + "epoch": 5.479824210946864, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13716 + }, + { + "epoch": 5.480223731522173, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 13717 + }, + { + "epoch": 5.480623252097483, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13718 + }, + { + "epoch": 5.481022772672793, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13719 + }, + { + "epoch": 5.4814222932481025, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13720 + }, + { + "epoch": 5.481821813823412, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13721 + }, + { + "epoch": 5.482221334398721, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13722 + }, + { + "epoch": 5.482620854974031, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13723 + }, + { + "epoch": 5.483020375549341, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13724 + }, + { + "epoch": 5.4834198961246505, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13725 + }, + { + "epoch": 5.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 13726 + }, + { + "epoch": 5.484218937275269, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13727 + }, + { + "epoch": 5.484618457850579, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13728 + }, + { + "epoch": 5.485017978425889, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13729 + }, + { + "epoch": 5.485417499001199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13730 + }, + { + "epoch": 5.485817019576508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13731 + }, + { + "epoch": 5.486216540151818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13732 + }, + { + "epoch": 5.486616060727127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13733 + }, + { + "epoch": 5.487015581302437, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13734 + }, + { + "epoch": 5.487415101877747, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13735 + }, + { + "epoch": 5.487814622453056, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13736 + }, + { + "epoch": 5.488214143028366, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13737 + }, + { + "epoch": 5.488613663603676, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13738 + }, + { + "epoch": 5.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13739 + }, + { + "epoch": 5.489412704754295, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13740 + }, + { + "epoch": 5.4898122253296044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13741 + }, + { + "epoch": 5.490211745904914, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13742 + }, + { + "epoch": 5.490611266480224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13743 + }, + { + "epoch": 5.491010787055533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 13744 + }, + { + "epoch": 5.491410307630843, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13745 + }, + { + "epoch": 5.4918098282061525, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13746 + }, + { + "epoch": 5.492209348781462, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13747 + }, + { + "epoch": 5.492608869356772, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13748 + }, + { + "epoch": 5.493008389932082, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13749 + }, + { + "epoch": 5.493407910507391, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13750 + }, + { + "epoch": 5.4938074310827005, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13751 + }, + { + "epoch": 5.49420695165801, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13752 + }, + { + "epoch": 5.49460647223332, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13753 + }, + { + "epoch": 5.49500599280863, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13754 + }, + { + "epoch": 5.49540551338394, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13755 + }, + { + "epoch": 5.495805033959249, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13756 + }, + { + "epoch": 5.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13757 + }, + { + "epoch": 5.496604075109868, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13758 + }, + { + "epoch": 5.497003595685178, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13759 + }, + { + "epoch": 5.497403116260488, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13760 + }, + { + "epoch": 5.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13761 + }, + { + "epoch": 5.498202157411106, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13762 + }, + { + "epoch": 5.498601677986416, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13763 + }, + { + "epoch": 5.499001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13764 + }, + { + "epoch": 5.499400719137036, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13765 + }, + { + "epoch": 5.4998002397123456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13766 + }, + { + "epoch": 5.5001997602876544, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13767 + }, + { + "epoch": 5.500599280862964, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13768 + }, + { + "epoch": 5.500998801438274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13769 + }, + { + "epoch": 5.501398322013584, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13770 + }, + { + "epoch": 5.501797842588894, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13771 + }, + { + "epoch": 5.502197363164203, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13772 + }, + { + "epoch": 5.502596883739512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13773 + }, + { + "epoch": 5.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13774 + }, + { + "epoch": 5.503395924890132, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13775 + }, + { + "epoch": 5.503795445465442, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13776 + }, + { + "epoch": 5.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13777 + }, + { + "epoch": 5.50459448661606, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13778 + }, + { + "epoch": 5.50499400719137, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13779 + }, + { + "epoch": 5.50539352776668, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13780 + }, + { + "epoch": 5.50579304834199, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13781 + }, + { + "epoch": 5.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13782 + }, + { + "epoch": 5.506592089492609, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 13783 + }, + { + "epoch": 5.506991610067918, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13784 + }, + { + "epoch": 5.507391130643228, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13785 + }, + { + "epoch": 5.507790651218538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13786 + }, + { + "epoch": 5.5081901717938475, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13787 + }, + { + "epoch": 5.508589692369157, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13788 + }, + { + "epoch": 5.508989212944467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13789 + }, + { + "epoch": 5.509388733519776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13790 + }, + { + "epoch": 5.509788254095086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13791 + }, + { + "epoch": 5.5101877746703956, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13792 + }, + { + "epoch": 5.510587295245705, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 13793 + }, + { + "epoch": 5.510986815821015, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13794 + }, + { + "epoch": 5.511386336396324, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13795 + }, + { + "epoch": 5.511785856971634, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13796 + }, + { + "epoch": 5.512185377546944, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13797 + }, + { + "epoch": 5.512584898122253, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13798 + }, + { + "epoch": 5.512984418697563, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13799 + }, + { + "epoch": 5.513383939272872, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13800 + }, + { + "epoch": 5.513783459848182, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13801 + }, + { + "epoch": 5.514182980423492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13802 + }, + { + "epoch": 5.514582500998801, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13803 + }, + { + "epoch": 5.514982021574111, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13804 + }, + { + "epoch": 5.515381542149421, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13805 + }, + { + "epoch": 5.515781062724731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13806 + }, + { + "epoch": 5.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13807 + }, + { + "epoch": 5.5165801038753495, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13808 + }, + { + "epoch": 5.516979624450659, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13809 + }, + { + "epoch": 5.517379145025969, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13810 + }, + { + "epoch": 5.517778665601279, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13811 + }, + { + "epoch": 5.518178186176588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13812 + }, + { + "epoch": 5.5185777067518975, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13813 + }, + { + "epoch": 5.518977227327207, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13814 + }, + { + "epoch": 5.519376747902517, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13815 + }, + { + "epoch": 5.519776268477827, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13816 + }, + { + "epoch": 5.520175789053136, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13817 + }, + { + "epoch": 5.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13818 + }, + { + "epoch": 5.520974830203755, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13819 + }, + { + "epoch": 5.521374350779065, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13820 + }, + { + "epoch": 5.521773871354375, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13821 + }, + { + "epoch": 5.522173391929685, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13822 + }, + { + "epoch": 5.5225729125049945, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13823 + }, + { + "epoch": 5.522972433080303, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13824 + }, + { + "epoch": 5.523371953655613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13825 + }, + { + "epoch": 5.523771474230923, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13826 + }, + { + "epoch": 5.524170994806233, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13827 + }, + { + "epoch": 5.5245705153815425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13828 + }, + { + "epoch": 5.524970035956851, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13829 + }, + { + "epoch": 5.525369556532161, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13830 + }, + { + "epoch": 5.525769077107471, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13831 + }, + { + "epoch": 5.526168597682781, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13832 + }, + { + "epoch": 5.526568118258091, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13833 + }, + { + "epoch": 5.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13834 + }, + { + "epoch": 5.527367159408709, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13835 + }, + { + "epoch": 5.527766679984019, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13836 + }, + { + "epoch": 5.528166200559329, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13837 + }, + { + "epoch": 5.528565721134639, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13838 + }, + { + "epoch": 5.528965241709948, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13839 + }, + { + "epoch": 5.529364762285257, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13840 + }, + { + "epoch": 5.529764282860567, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13841 + }, + { + "epoch": 5.530163803435877, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13842 + }, + { + "epoch": 5.530563324011187, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13843 + }, + { + "epoch": 5.5309628445864965, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13844 + }, + { + "epoch": 5.531362365161806, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13845 + }, + { + "epoch": 5.531761885737115, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13846 + }, + { + "epoch": 5.532161406312425, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13847 + }, + { + "epoch": 5.532560926887735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13848 + }, + { + "epoch": 5.5329604474630445, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13849 + }, + { + "epoch": 5.533359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13850 + }, + { + "epoch": 5.533759488613663, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13851 + }, + { + "epoch": 5.534159009188973, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 13852 + }, + { + "epoch": 5.534558529764283, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13853 + }, + { + "epoch": 5.5349580503395925, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13854 + }, + { + "epoch": 5.535357570914902, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13855 + }, + { + "epoch": 5.535757091490212, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13856 + }, + { + "epoch": 5.536156612065521, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13857 + }, + { + "epoch": 5.536556132640831, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13858 + }, + { + "epoch": 5.536955653216141, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13859 + }, + { + "epoch": 5.53735517379145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 13860 + }, + { + "epoch": 5.53775469436676, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 13861 + }, + { + "epoch": 5.53815421494207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13862 + }, + { + "epoch": 5.538553735517379, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 13863 + }, + { + "epoch": 5.538953256092689, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13864 + }, + { + "epoch": 5.539352776667998, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13865 + }, + { + "epoch": 5.539752297243308, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13866 + }, + { + "epoch": 5.540151817818618, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 13867 + }, + { + "epoch": 5.540551338393927, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13868 + }, + { + "epoch": 5.540950858969237, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13869 + }, + { + "epoch": 5.5413503795445465, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13870 + }, + { + "epoch": 5.541749900119856, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13871 + }, + { + "epoch": 5.542149420695166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13872 + }, + { + "epoch": 5.542548941270476, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13873 + }, + { + "epoch": 5.542948461845785, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13874 + }, + { + "epoch": 5.5433479824210945, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13875 + }, + { + "epoch": 5.543747502996404, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 13876 + }, + { + "epoch": 5.544147023571714, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13877 + }, + { + "epoch": 5.544546544147024, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13878 + }, + { + "epoch": 5.544946064722334, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13879 + }, + { + "epoch": 5.5453455852976425, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13880 + }, + { + "epoch": 5.545745105872952, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13881 + }, + { + "epoch": 5.546144626448262, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13882 + }, + { + "epoch": 5.546544147023572, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13883 + }, + { + "epoch": 5.546943667598882, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13884 + }, + { + "epoch": 5.547343188174191, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13885 + }, + { + "epoch": 5.5477427087495, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13886 + }, + { + "epoch": 5.54814222932481, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 13887 + }, + { + "epoch": 5.54854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13888 + }, + { + "epoch": 5.54894127047543, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13889 + }, + { + "epoch": 5.549340791050739, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13890 + }, + { + "epoch": 5.549740311626048, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13891 + }, + { + "epoch": 5.550139832201358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13892 + }, + { + "epoch": 5.550539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13893 + }, + { + "epoch": 5.550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13894 + }, + { + "epoch": 5.551338393927288, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13895 + }, + { + "epoch": 5.551737914502597, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13896 + }, + { + "epoch": 5.552137435077906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13897 + }, + { + "epoch": 5.552536955653216, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13898 + }, + { + "epoch": 5.552936476228526, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13899 + }, + { + "epoch": 5.553335996803836, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13900 + }, + { + "epoch": 5.553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13901 + }, + { + "epoch": 5.554135037954454, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13902 + }, + { + "epoch": 5.554534558529764, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13903 + }, + { + "epoch": 5.554934079105074, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13904 + }, + { + "epoch": 5.555333599680384, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13905 + }, + { + "epoch": 5.555733120255693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13906 + }, + { + "epoch": 5.556132640831002, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13907 + }, + { + "epoch": 5.556532161406312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13908 + }, + { + "epoch": 5.556931681981622, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13909 + }, + { + "epoch": 5.557331202556932, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13910 + }, + { + "epoch": 5.5577307231322415, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13911 + }, + { + "epoch": 5.558130243707551, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13912 + }, + { + "epoch": 5.558529764282861, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13913 + }, + { + "epoch": 5.55892928485817, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13914 + }, + { + "epoch": 5.55932880543348, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13915 + }, + { + "epoch": 5.5597283260087895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13916 + }, + { + "epoch": 5.560127846584099, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13917 + }, + { + "epoch": 5.560527367159409, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13918 + }, + { + "epoch": 5.560926887734718, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13919 + }, + { + "epoch": 5.561326408310028, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13920 + }, + { + "epoch": 5.561725928885338, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13921 + }, + { + "epoch": 5.562125449460647, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13922 + }, + { + "epoch": 5.562524970035957, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13923 + }, + { + "epoch": 5.562924490611266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13924 + }, + { + "epoch": 5.563324011186576, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13925 + }, + { + "epoch": 5.563723531761886, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13926 + }, + { + "epoch": 5.564123052337195, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 13927 + }, + { + "epoch": 5.564522572912505, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13928 + }, + { + "epoch": 5.564922093487815, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13929 + }, + { + "epoch": 5.565321614063124, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13930 + }, + { + "epoch": 5.565721134638434, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13931 + }, + { + "epoch": 5.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13932 + }, + { + "epoch": 5.566520175789053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13933 + }, + { + "epoch": 5.566919696364363, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13934 + }, + { + "epoch": 5.567319216939673, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13935 + }, + { + "epoch": 5.567718737514982, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13936 + }, + { + "epoch": 5.5681182580902915, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13937 + }, + { + "epoch": 5.568517778665601, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13938 + }, + { + "epoch": 5.568917299240911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13939 + }, + { + "epoch": 5.569316819816221, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13940 + }, + { + "epoch": 5.56971634039153, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13941 + }, + { + "epoch": 5.5701158609668395, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13942 + }, + { + "epoch": 5.570515381542149, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13943 + }, + { + "epoch": 5.570914902117459, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13944 + }, + { + "epoch": 5.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13945 + }, + { + "epoch": 5.571713943268079, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13946 + }, + { + "epoch": 5.572113463843388, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13947 + }, + { + "epoch": 5.572512984418697, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13948 + }, + { + "epoch": 5.572912504994007, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13949 + }, + { + "epoch": 5.573312025569317, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13950 + }, + { + "epoch": 5.573711546144627, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13951 + }, + { + "epoch": 5.5741110667199365, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13952 + }, + { + "epoch": 5.574510587295245, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13953 + }, + { + "epoch": 5.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13954 + }, + { + "epoch": 5.575309628445865, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13955 + }, + { + "epoch": 5.575709149021175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13956 + }, + { + "epoch": 5.5761086695964845, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13957 + }, + { + "epoch": 5.5765081901717934, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13958 + }, + { + "epoch": 5.576907710747103, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 13959 + }, + { + "epoch": 5.577307231322413, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13960 + }, + { + "epoch": 5.577706751897723, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13961 + }, + { + "epoch": 5.578106272473033, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13962 + }, + { + "epoch": 5.578505793048342, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13963 + }, + { + "epoch": 5.578905313623651, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13964 + }, + { + "epoch": 5.579304834198961, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13965 + }, + { + "epoch": 5.579704354774271, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13966 + }, + { + "epoch": 5.580103875349581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13967 + }, + { + "epoch": 5.58050339592489, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13968 + }, + { + "epoch": 5.5809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 13969 + }, + { + "epoch": 5.581302437075509, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13970 + }, + { + "epoch": 5.581701957650819, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13971 + }, + { + "epoch": 5.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13972 + }, + { + "epoch": 5.5825009988014385, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13973 + }, + { + "epoch": 5.582900519376748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13974 + }, + { + "epoch": 5.583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13975 + }, + { + "epoch": 5.583699560527367, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13976 + }, + { + "epoch": 5.584099081102677, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13977 + }, + { + "epoch": 5.5844986016779865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13978 + }, + { + "epoch": 5.584898122253296, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13979 + }, + { + "epoch": 5.585297642828605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13980 + }, + { + "epoch": 5.585697163403915, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13981 + }, + { + "epoch": 5.586096683979225, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13982 + }, + { + "epoch": 5.5864962045545346, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13983 + }, + { + "epoch": 5.586895725129844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13984 + }, + { + "epoch": 5.587295245705154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13985 + }, + { + "epoch": 5.587694766280464, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13986 + }, + { + "epoch": 5.588094286855773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13987 + }, + { + "epoch": 5.588493807431083, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 13988 + }, + { + "epoch": 5.588893328006392, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13989 + }, + { + "epoch": 5.589292848581702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13990 + }, + { + "epoch": 5.589692369157012, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13991 + }, + { + "epoch": 5.590091889732321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13992 + }, + { + "epoch": 5.590491410307631, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13993 + }, + { + "epoch": 5.59089093088294, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 13994 + }, + { + "epoch": 5.59129045145825, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13995 + }, + { + "epoch": 5.59168997203356, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13996 + }, + { + "epoch": 5.592089492608869, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13997 + }, + { + "epoch": 5.592489013184179, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13998 + }, + { + "epoch": 5.5928885337594885, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13999 + }, + { + "epoch": 5.593288054334798, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14000 + }, + { + "epoch": 5.593687574910108, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14001 + }, + { + "epoch": 5.594087095485418, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14002 + }, + { + "epoch": 5.594486616060728, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 14003 + }, + { + "epoch": 5.5948861366360365, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14004 + }, + { + "epoch": 5.595285657211346, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14005 + }, + { + "epoch": 5.595685177786656, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14006 + }, + { + "epoch": 5.596084698361966, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14007 + }, + { + "epoch": 5.596484218937276, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14008 + }, + { + "epoch": 5.5968837395125846, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14009 + }, + { + "epoch": 5.597283260087894, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.2923, + "step": 14010 + }, + { + "epoch": 5.597682780663204, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14011 + }, + { + "epoch": 5.598082301238514, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14012 + }, + { + "epoch": 5.598481821813824, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14013 + }, + { + "epoch": 5.598881342389133, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14014 + }, + { + "epoch": 5.599280862964442, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14015 + }, + { + "epoch": 5.599680383539752, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14016 + }, + { + "epoch": 5.600079904115062, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 14017 + }, + { + "epoch": 5.600479424690372, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14018 + }, + { + "epoch": 5.6008789452656815, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14019 + }, + { + "epoch": 5.60127846584099, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14020 + }, + { + "epoch": 5.6016779864163, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14021 + }, + { + "epoch": 5.60207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14022 + }, + { + "epoch": 5.60247702756692, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14023 + }, + { + "epoch": 5.60287654814223, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 14024 + }, + { + "epoch": 5.603276068717539, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14025 + }, + { + "epoch": 5.603675589292848, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14026 + }, + { + "epoch": 5.604075109868158, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14027 + }, + { + "epoch": 5.604474630443468, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14028 + }, + { + "epoch": 5.604874151018778, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14029 + }, + { + "epoch": 5.605273671594087, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 14030 + }, + { + "epoch": 5.605673192169396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14031 + }, + { + "epoch": 5.606072712744706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14032 + }, + { + "epoch": 5.606472233320016, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14033 + }, + { + "epoch": 5.606871753895326, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14034 + }, + { + "epoch": 5.6072712744706354, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14035 + }, + { + "epoch": 5.607670795045945, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14036 + }, + { + "epoch": 5.608070315621254, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14037 + }, + { + "epoch": 5.608469836196564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14038 + }, + { + "epoch": 5.608869356771874, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14039 + }, + { + "epoch": 5.6092688773471835, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14040 + }, + { + "epoch": 5.609668397922493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14041 + }, + { + "epoch": 5.610067918497803, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14042 + }, + { + "epoch": 5.610467439073112, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14043 + }, + { + "epoch": 5.610866959648422, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14044 + }, + { + "epoch": 5.6112664802237315, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14045 + }, + { + "epoch": 5.611666000799041, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14046 + }, + { + "epoch": 5.612065521374351, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14047 + }, + { + "epoch": 5.61246504194966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14048 + }, + { + "epoch": 5.61286456252497, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14049 + }, + { + "epoch": 5.61326408310028, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14050 + }, + { + "epoch": 5.613663603675589, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14051 + }, + { + "epoch": 5.614063124250899, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14052 + }, + { + "epoch": 5.614462644826209, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14053 + }, + { + "epoch": 5.614862165401518, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14054 + }, + { + "epoch": 5.615261685976828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14055 + }, + { + "epoch": 5.615661206552137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14056 + }, + { + "epoch": 5.616060727127447, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14057 + }, + { + "epoch": 5.616460247702757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14058 + }, + { + "epoch": 5.616859768278067, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14059 + }, + { + "epoch": 5.617259288853376, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14060 + }, + { + "epoch": 5.6176588094286855, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14061 + }, + { + "epoch": 5.618058330003995, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14062 + }, + { + "epoch": 5.618457850579305, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14063 + }, + { + "epoch": 5.618857371154615, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14064 + }, + { + "epoch": 5.619256891729924, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14065 + }, + { + "epoch": 5.6196564123052335, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14066 + }, + { + "epoch": 5.620055932880543, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14067 + }, + { + "epoch": 5.620455453455853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14068 + }, + { + "epoch": 5.620854974031163, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14069 + }, + { + "epoch": 5.621254494606472, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14070 + }, + { + "epoch": 5.6216540151817815, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14071 + }, + { + "epoch": 5.622053535757091, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14072 + }, + { + "epoch": 5.622453056332401, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14073 + }, + { + "epoch": 5.622852576907711, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14074 + }, + { + "epoch": 5.623252097483021, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14075 + }, + { + "epoch": 5.6236516180583305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14076 + }, + { + "epoch": 5.624051138633639, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14077 + }, + { + "epoch": 5.624450659208949, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14078 + }, + { + "epoch": 5.624850179784259, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14079 + }, + { + "epoch": 5.625249700359569, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.2954, + "step": 14080 + }, + { + "epoch": 5.6256492209348785, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14081 + }, + { + "epoch": 5.626048741510187, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14082 + }, + { + "epoch": 5.626448262085497, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14083 + }, + { + "epoch": 5.626847782660807, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14084 + }, + { + "epoch": 5.627247303236117, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 14085 + }, + { + "epoch": 5.627646823811427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14086 + }, + { + "epoch": 5.6280463443867355, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14087 + }, + { + "epoch": 5.628445864962045, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14088 + }, + { + "epoch": 5.628845385537355, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14089 + }, + { + "epoch": 5.629244906112665, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14090 + }, + { + "epoch": 5.629644426687975, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14091 + }, + { + "epoch": 5.630043947263284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14092 + }, + { + "epoch": 5.630443467838594, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14093 + }, + { + "epoch": 5.630842988413903, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14094 + }, + { + "epoch": 5.631242508989213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14095 + }, + { + "epoch": 5.631642029564523, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14096 + }, + { + "epoch": 5.632041550139832, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14097 + }, + { + "epoch": 5.632441070715142, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14098 + }, + { + "epoch": 5.632840591290451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14099 + }, + { + "epoch": 5.633240111865761, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14100 + }, + { + "epoch": 5.633639632441071, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14101 + }, + { + "epoch": 5.6340391530163805, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14102 + }, + { + "epoch": 5.63443867359169, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14103 + }, + { + "epoch": 5.634838194166999, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14104 + }, + { + "epoch": 5.635237714742309, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14105 + }, + { + "epoch": 5.635637235317619, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14106 + }, + { + "epoch": 5.6360367558929285, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14107 + }, + { + "epoch": 5.636436276468238, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14108 + }, + { + "epoch": 5.636835797043548, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14109 + }, + { + "epoch": 5.637235317618857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 14110 + }, + { + "epoch": 5.637634838194167, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14111 + }, + { + "epoch": 5.638034358769477, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14112 + }, + { + "epoch": 5.638433879344786, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14113 + }, + { + "epoch": 5.638833399920096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14114 + }, + { + "epoch": 5.639232920495406, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14115 + }, + { + "epoch": 5.639632441070715, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14116 + }, + { + "epoch": 5.640031961646025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14117 + }, + { + "epoch": 5.640431482221334, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14118 + }, + { + "epoch": 5.640831002796644, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14119 + }, + { + "epoch": 5.641230523371954, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14120 + }, + { + "epoch": 5.641630043947263, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14121 + }, + { + "epoch": 5.642029564522573, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14122 + }, + { + "epoch": 5.642429085097882, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 14123 + }, + { + "epoch": 5.642828605673192, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14124 + }, + { + "epoch": 5.643228126248502, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14125 + }, + { + "epoch": 5.643627646823812, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14126 + }, + { + "epoch": 5.644027167399121, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14127 + }, + { + "epoch": 5.6444266879744305, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14128 + }, + { + "epoch": 5.64482620854974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14129 + }, + { + "epoch": 5.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14130 + }, + { + "epoch": 5.64562524970036, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14131 + }, + { + "epoch": 5.64602477027567, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14132 + }, + { + "epoch": 5.6464242908509785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14133 + }, + { + "epoch": 5.646823811426288, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14134 + }, + { + "epoch": 5.647223332001598, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14135 + }, + { + "epoch": 5.647622852576908, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14136 + }, + { + "epoch": 5.648022373152218, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14137 + }, + { + "epoch": 5.648421893727527, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14138 + }, + { + "epoch": 5.648821414302836, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14139 + }, + { + "epoch": 5.649220934878146, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14140 + }, + { + "epoch": 5.649620455453456, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14141 + }, + { + "epoch": 5.650019976028766, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14142 + }, + { + "epoch": 5.6504194966040755, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14143 + }, + { + "epoch": 5.650819017179384, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14144 + }, + { + "epoch": 5.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14145 + }, + { + "epoch": 5.651618058330004, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14146 + }, + { + "epoch": 5.652017578905314, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14147 + }, + { + "epoch": 5.6524170994806235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14148 + }, + { + "epoch": 5.652816620055933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14149 + }, + { + "epoch": 5.653216140631242, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 14150 + }, + { + "epoch": 5.653615661206552, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14151 + }, + { + "epoch": 5.654015181781862, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14152 + }, + { + "epoch": 5.654414702357172, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14153 + }, + { + "epoch": 5.654814222932481, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14154 + }, + { + "epoch": 5.65521374350779, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14155 + }, + { + "epoch": 5.6556132640831, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14156 + }, + { + "epoch": 5.65601278465841, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14157 + }, + { + "epoch": 5.65641230523372, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14158 + }, + { + "epoch": 5.656811825809029, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14159 + }, + { + "epoch": 5.657211346384338, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14160 + }, + { + "epoch": 5.657610866959648, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 14161 + }, + { + "epoch": 5.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14162 + }, + { + "epoch": 5.658409908110268, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14163 + }, + { + "epoch": 5.6588094286855775, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14164 + }, + { + "epoch": 5.659208949260887, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14165 + }, + { + "epoch": 5.659608469836197, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14166 + }, + { + "epoch": 5.660007990411506, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14167 + }, + { + "epoch": 5.660407510986816, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14168 + }, + { + "epoch": 5.6608070315621255, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14169 + }, + { + "epoch": 5.661206552137435, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14170 + }, + { + "epoch": 5.661606072712745, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14171 + }, + { + "epoch": 5.662005593288054, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14172 + }, + { + "epoch": 5.662405113863364, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14173 + }, + { + "epoch": 5.6628046344386735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14174 + }, + { + "epoch": 5.663204155013983, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14175 + }, + { + "epoch": 5.663603675589293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 14176 + }, + { + "epoch": 5.664003196164602, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 14177 + }, + { + "epoch": 5.664402716739912, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14178 + }, + { + "epoch": 5.664802237315222, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14179 + }, + { + "epoch": 5.665201757890531, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14180 + }, + { + "epoch": 5.665601278465841, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14181 + }, + { + "epoch": 5.666000799041151, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14182 + }, + { + "epoch": 5.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14183 + }, + { + "epoch": 5.66679984019177, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14184 + }, + { + "epoch": 5.667199360767079, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14185 + }, + { + "epoch": 5.667598881342389, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14186 + }, + { + "epoch": 5.667998401917699, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14187 + }, + { + "epoch": 5.668397922493009, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14188 + }, + { + "epoch": 5.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14189 + }, + { + "epoch": 5.6691969636436275, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14190 + }, + { + "epoch": 5.669596484218937, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14191 + }, + { + "epoch": 5.669996004794247, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14192 + }, + { + "epoch": 5.670395525369557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14193 + }, + { + "epoch": 5.670795045944866, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14194 + }, + { + "epoch": 5.6711945665201755, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14195 + }, + { + "epoch": 5.671594087095485, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14196 + }, + { + "epoch": 5.671993607670795, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14197 + }, + { + "epoch": 5.672393128246105, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14198 + }, + { + "epoch": 5.672792648821415, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14199 + }, + { + "epoch": 5.6731921693967235, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14200 + }, + { + "epoch": 5.673591689972033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14201 + }, + { + "epoch": 5.673991210547343, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14202 + }, + { + "epoch": 5.674390731122653, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14203 + }, + { + "epoch": 5.674790251697963, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14204 + }, + { + "epoch": 5.6751897722732725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 14205 + }, + { + "epoch": 5.675589292848581, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14206 + }, + { + "epoch": 5.675988813423891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14207 + }, + { + "epoch": 5.676388333999201, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14208 + }, + { + "epoch": 5.676787854574511, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14209 + }, + { + "epoch": 5.6771873751498205, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14210 + }, + { + "epoch": 5.677586895725129, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14211 + }, + { + "epoch": 5.677986416300439, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14212 + }, + { + "epoch": 5.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14213 + }, + { + "epoch": 5.678785457451059, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14214 + }, + { + "epoch": 5.679184978026369, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14215 + }, + { + "epoch": 5.679584498601678, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14216 + }, + { + "epoch": 5.679984019176987, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14217 + }, + { + "epoch": 5.680383539752297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14218 + }, + { + "epoch": 5.680783060327607, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14219 + }, + { + "epoch": 5.681182580902917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14220 + }, + { + "epoch": 5.681582101478226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14221 + }, + { + "epoch": 5.681981622053536, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14222 + }, + { + "epoch": 5.682381142628845, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14223 + }, + { + "epoch": 5.682780663204155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14224 + }, + { + "epoch": 5.683180183779465, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14225 + }, + { + "epoch": 5.683579704354774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 14226 + }, + { + "epoch": 5.683979224930084, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14227 + }, + { + "epoch": 5.684378745505393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14228 + }, + { + "epoch": 5.684778266080703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14229 + }, + { + "epoch": 5.685177786656013, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14230 + }, + { + "epoch": 5.6855773072313225, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14231 + }, + { + "epoch": 5.685976827806632, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14232 + }, + { + "epoch": 5.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14233 + }, + { + "epoch": 5.686775868957251, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14234 + }, + { + "epoch": 5.687175389532561, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14235 + }, + { + "epoch": 5.6875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14236 + }, + { + "epoch": 5.68797443068318, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14237 + }, + { + "epoch": 5.68837395125849, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14238 + }, + { + "epoch": 5.6887734718338, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14239 + }, + { + "epoch": 5.689172992409109, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14240 + }, + { + "epoch": 5.689572512984419, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14241 + }, + { + "epoch": 5.689972033559728, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 14242 + }, + { + "epoch": 5.690371554135038, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14243 + }, + { + "epoch": 5.690771074710348, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14244 + }, + { + "epoch": 5.691170595285657, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14245 + }, + { + "epoch": 5.691570115860967, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14246 + }, + { + "epoch": 5.691969636436276, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14247 + }, + { + "epoch": 5.692369157011586, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14248 + }, + { + "epoch": 5.692768677586896, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14249 + }, + { + "epoch": 5.693168198162205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14250 + }, + { + "epoch": 5.693567718737515, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 14251 + }, + { + "epoch": 5.6939672393128244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14252 + }, + { + "epoch": 5.694366759888134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14253 + }, + { + "epoch": 5.694766280463444, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14254 + }, + { + "epoch": 5.695165801038754, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14255 + }, + { + "epoch": 5.695565321614064, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14256 + }, + { + "epoch": 5.6959648421893725, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14257 + }, + { + "epoch": 5.696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14258 + }, + { + "epoch": 5.696763883339992, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14259 + }, + { + "epoch": 5.697163403915302, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14260 + }, + { + "epoch": 5.697562924490612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14261 + }, + { + "epoch": 5.6979624450659205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14262 + }, + { + "epoch": 5.69836196564123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14263 + }, + { + "epoch": 5.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14264 + }, + { + "epoch": 5.69916100679185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14265 + }, + { + "epoch": 5.69956052736716, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14266 + }, + { + "epoch": 5.699960047942469, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14267 + }, + { + "epoch": 5.700359568517778, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 14268 + }, + { + "epoch": 5.700759089093088, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14269 + }, + { + "epoch": 5.701158609668398, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14270 + }, + { + "epoch": 5.701558130243708, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14271 + }, + { + "epoch": 5.7019576508190175, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 14272 + }, + { + "epoch": 5.702357171394327, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14273 + }, + { + "epoch": 5.702756691969636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14274 + }, + { + "epoch": 5.703156212544946, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14275 + }, + { + "epoch": 5.703555733120256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14276 + }, + { + "epoch": 5.7039552536955656, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14277 + }, + { + "epoch": 5.704354774270875, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14278 + }, + { + "epoch": 5.704754294846184, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14279 + }, + { + "epoch": 5.705153815421494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14280 + }, + { + "epoch": 5.705553335996804, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14281 + }, + { + "epoch": 5.705952856572114, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14282 + }, + { + "epoch": 5.706352377147423, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14283 + }, + { + "epoch": 5.706751897722732, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14284 + }, + { + "epoch": 5.707151418298042, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14285 + }, + { + "epoch": 5.707550938873352, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14286 + }, + { + "epoch": 5.707950459448662, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14287 + }, + { + "epoch": 5.708349980023971, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14288 + }, + { + "epoch": 5.708749500599281, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14289 + }, + { + "epoch": 5.70914902117459, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 14290 + }, + { + "epoch": 5.7095485417499, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14291 + }, + { + "epoch": 5.70994806232521, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14292 + }, + { + "epoch": 5.7103475829005195, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14293 + }, + { + "epoch": 5.710747103475829, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14294 + }, + { + "epoch": 5.711146624051139, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14295 + }, + { + "epoch": 5.711546144626448, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14296 + }, + { + "epoch": 5.711945665201758, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 14297 + }, + { + "epoch": 5.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14298 + }, + { + "epoch": 5.712744706352377, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 14299 + }, + { + "epoch": 5.713144226927687, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14300 + }, + { + "epoch": 5.713543747502996, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14301 + }, + { + "epoch": 5.713943268078306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14302 + }, + { + "epoch": 5.714342788653616, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14303 + }, + { + "epoch": 5.714742309228925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14304 + }, + { + "epoch": 5.715141829804235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14305 + }, + { + "epoch": 5.715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14306 + }, + { + "epoch": 5.715940870954854, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14307 + }, + { + "epoch": 5.716340391530164, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14308 + }, + { + "epoch": 5.716739912105473, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14309 + }, + { + "epoch": 5.717139432680783, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14310 + }, + { + "epoch": 5.717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14311 + }, + { + "epoch": 5.717938473831403, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14312 + }, + { + "epoch": 5.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14313 + }, + { + "epoch": 5.718737514982021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14314 + }, + { + "epoch": 5.719137035557331, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14315 + }, + { + "epoch": 5.719536556132641, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14316 + }, + { + "epoch": 5.719936076707951, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 14317 + }, + { + "epoch": 5.72033559728326, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14318 + }, + { + "epoch": 5.7207351178585695, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14319 + }, + { + "epoch": 5.721134638433879, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14320 + }, + { + "epoch": 5.721534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14321 + }, + { + "epoch": 5.721933679584499, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14322 + }, + { + "epoch": 5.722333200159808, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14323 + }, + { + "epoch": 5.7227327207351175, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 14324 + }, + { + "epoch": 5.723132241310427, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14325 + }, + { + "epoch": 5.723531761885737, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14326 + }, + { + "epoch": 5.723931282461047, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14327 + }, + { + "epoch": 5.724330803036357, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14328 + }, + { + "epoch": 5.7247303236116664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14329 + }, + { + "epoch": 5.725129844186975, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14330 + }, + { + "epoch": 5.725529364762285, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14331 + }, + { + "epoch": 5.725928885337595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.301, + "step": 14332 + }, + { + "epoch": 5.726328405912905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14333 + }, + { + "epoch": 5.7267279264882145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14334 + }, + { + "epoch": 5.727127447063523, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14335 + }, + { + "epoch": 5.727526967638833, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14336 + }, + { + "epoch": 5.727926488214143, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14337 + }, + { + "epoch": 5.728326008789453, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14338 + }, + { + "epoch": 5.7287255293647625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14339 + }, + { + "epoch": 5.729125049940071, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14340 + }, + { + "epoch": 5.729524570515381, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14341 + }, + { + "epoch": 5.729924091090691, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14342 + }, + { + "epoch": 5.730323611666001, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14343 + }, + { + "epoch": 5.730723132241311, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14344 + }, + { + "epoch": 5.73112265281662, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14345 + }, + { + "epoch": 5.73152217339193, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14346 + }, + { + "epoch": 5.731921693967239, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14347 + }, + { + "epoch": 5.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14348 + }, + { + "epoch": 5.732720735117859, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14349 + }, + { + "epoch": 5.733120255693168, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14350 + }, + { + "epoch": 5.733519776268478, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14351 + }, + { + "epoch": 5.733919296843787, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 14352 + }, + { + "epoch": 5.734318817419097, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14353 + }, + { + "epoch": 5.734718337994407, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14354 + }, + { + "epoch": 5.7351178585697165, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14355 + }, + { + "epoch": 5.735517379145026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14356 + }, + { + "epoch": 5.735916899720335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14357 + }, + { + "epoch": 5.736316420295645, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14358 + }, + { + "epoch": 5.736715940870955, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14359 + }, + { + "epoch": 5.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14360 + }, + { + "epoch": 5.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14361 + }, + { + "epoch": 5.737914502596884, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14362 + }, + { + "epoch": 5.738314023172194, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14363 + }, + { + "epoch": 5.738713543747503, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14364 + }, + { + "epoch": 5.7391130643228125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14365 + }, + { + "epoch": 5.739512584898122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14366 + }, + { + "epoch": 5.739912105473432, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 14367 + }, + { + "epoch": 5.740311626048742, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14368 + }, + { + "epoch": 5.740711146624051, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14369 + }, + { + "epoch": 5.741110667199361, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14370 + }, + { + "epoch": 5.74151018777467, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14371 + }, + { + "epoch": 5.74190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14372 + }, + { + "epoch": 5.74230922892529, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14373 + }, + { + "epoch": 5.742708749500599, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14374 + }, + { + "epoch": 5.743108270075909, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 14375 + }, + { + "epoch": 5.743507790651218, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14376 + }, + { + "epoch": 5.743907311226528, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14377 + }, + { + "epoch": 5.744306831801838, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14378 + }, + { + "epoch": 5.744706352377148, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14379 + }, + { + "epoch": 5.745105872952457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14380 + }, + { + "epoch": 5.7455053935277665, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14381 + }, + { + "epoch": 5.745904914103076, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14382 + }, + { + "epoch": 5.746304434678386, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 14383 + }, + { + "epoch": 5.746703955253696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14384 + }, + { + "epoch": 5.747103475829006, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14385 + }, + { + "epoch": 5.7475029964043145, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14386 + }, + { + "epoch": 5.747902516979624, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14387 + }, + { + "epoch": 5.748302037554934, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14388 + }, + { + "epoch": 5.748701558130244, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14389 + }, + { + "epoch": 5.749101078705554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14390 + }, + { + "epoch": 5.7495005992808625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14391 + }, + { + "epoch": 5.749900119856172, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14392 + }, + { + "epoch": 5.750299640431482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14393 + }, + { + "epoch": 5.750699161006792, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14394 + }, + { + "epoch": 5.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14395 + }, + { + "epoch": 5.7514982021574115, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14396 + }, + { + "epoch": 5.75189772273272, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14397 + }, + { + "epoch": 5.75229724330803, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14398 + }, + { + "epoch": 5.75269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14399 + }, + { + "epoch": 5.75309628445865, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14400 + }, + { + "epoch": 5.7534958050339595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14401 + }, + { + "epoch": 5.753895325609269, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14402 + }, + { + "epoch": 5.754294846184578, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14403 + }, + { + "epoch": 5.754694366759888, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14404 + }, + { + "epoch": 5.755093887335198, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14405 + }, + { + "epoch": 5.755493407910508, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14406 + }, + { + "epoch": 5.755892928485817, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14407 + }, + { + "epoch": 5.756292449061126, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14408 + }, + { + "epoch": 5.756691969636436, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14409 + }, + { + "epoch": 5.757091490211746, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14410 + }, + { + "epoch": 5.757491010787056, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14411 + }, + { + "epoch": 5.757890531362365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14412 + }, + { + "epoch": 5.758290051937674, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14413 + }, + { + "epoch": 5.758689572512984, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14414 + }, + { + "epoch": 5.759089093088294, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14415 + }, + { + "epoch": 5.759488613663604, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14416 + }, + { + "epoch": 5.759888134238913, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14417 + }, + { + "epoch": 5.760287654814223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14418 + }, + { + "epoch": 5.760687175389533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14419 + }, + { + "epoch": 5.761086695964842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 14420 + }, + { + "epoch": 5.761486216540152, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14421 + }, + { + "epoch": 5.7618857371154615, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14422 + }, + { + "epoch": 5.762285257690771, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14423 + }, + { + "epoch": 5.762684778266081, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14424 + }, + { + "epoch": 5.76308429884139, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14425 + }, + { + "epoch": 5.7634838194167, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14426 + }, + { + "epoch": 5.7638833399920095, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14427 + }, + { + "epoch": 5.764282860567319, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14428 + }, + { + "epoch": 5.764682381142629, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14429 + }, + { + "epoch": 5.765081901717938, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14430 + }, + { + "epoch": 5.765481422293248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14431 + }, + { + "epoch": 5.765880942868558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14432 + }, + { + "epoch": 5.766280463443867, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14433 + }, + { + "epoch": 5.766679984019177, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14434 + }, + { + "epoch": 5.767079504594487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14435 + }, + { + "epoch": 5.767479025169797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14436 + }, + { + "epoch": 5.767878545745106, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14437 + }, + { + "epoch": 5.768278066320415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14438 + }, + { + "epoch": 5.768677586895725, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14439 + }, + { + "epoch": 5.769077107471035, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14440 + }, + { + "epoch": 5.769476628046345, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14441 + }, + { + "epoch": 5.769876148621654, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14442 + }, + { + "epoch": 5.770275669196963, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14443 + }, + { + "epoch": 5.770675189772273, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14444 + }, + { + "epoch": 5.771074710347583, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14445 + }, + { + "epoch": 5.771474230922893, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14446 + }, + { + "epoch": 5.771873751498202, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14447 + }, + { + "epoch": 5.7722732720735115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14448 + }, + { + "epoch": 5.772672792648821, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14449 + }, + { + "epoch": 5.773072313224131, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14450 + }, + { + "epoch": 5.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14451 + }, + { + "epoch": 5.773871354374751, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14452 + }, + { + "epoch": 5.77427087495006, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14453 + }, + { + "epoch": 5.774670395525369, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14454 + }, + { + "epoch": 5.775069916100679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14455 + }, + { + "epoch": 5.775469436675989, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14456 + }, + { + "epoch": 5.775868957251299, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14457 + }, + { + "epoch": 5.7762684778266085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14458 + }, + { + "epoch": 5.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14459 + }, + { + "epoch": 5.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14460 + }, + { + "epoch": 5.777467039552537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 14461 + }, + { + "epoch": 5.777866560127847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14462 + }, + { + "epoch": 5.7782660807031565, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14463 + }, + { + "epoch": 5.778665601278465, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14464 + }, + { + "epoch": 5.779065121853775, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 14465 + }, + { + "epoch": 5.779464642429085, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14466 + }, + { + "epoch": 5.779864163004395, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14467 + }, + { + "epoch": 5.7802636835797045, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14468 + }, + { + "epoch": 5.780663204155014, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14469 + }, + { + "epoch": 5.781062724730323, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14470 + }, + { + "epoch": 5.781462245305633, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14471 + }, + { + "epoch": 5.781861765880943, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14472 + }, + { + "epoch": 5.782261286456253, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14473 + }, + { + "epoch": 5.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14474 + }, + { + "epoch": 5.783060327606872, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14475 + }, + { + "epoch": 5.783459848182181, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14476 + }, + { + "epoch": 5.783859368757491, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14477 + }, + { + "epoch": 5.784258889332801, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14478 + }, + { + "epoch": 5.78465840990811, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14479 + }, + { + "epoch": 5.78505793048342, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14480 + }, + { + "epoch": 5.785457451058729, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14481 + }, + { + "epoch": 5.785856971634039, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 14482 + }, + { + "epoch": 5.786256492209349, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14483 + }, + { + "epoch": 5.7866560127846585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14484 + }, + { + "epoch": 5.787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 14485 + }, + { + "epoch": 5.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 14486 + }, + { + "epoch": 5.787854574510587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14487 + }, + { + "epoch": 5.788254095085897, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14488 + }, + { + "epoch": 5.7886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14489 + }, + { + "epoch": 5.789053136236516, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14490 + }, + { + "epoch": 5.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14491 + }, + { + "epoch": 5.789852177387136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14492 + }, + { + "epoch": 5.790251697962445, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14493 + }, + { + "epoch": 5.7906512185377546, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14494 + }, + { + "epoch": 5.791050739113064, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14495 + }, + { + "epoch": 5.791450259688374, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 14496 + }, + { + "epoch": 5.791849780263684, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14497 + }, + { + "epoch": 5.792249300838993, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14498 + }, + { + "epoch": 5.792648821414303, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14499 + }, + { + "epoch": 5.793048341989612, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 14500 + }, + { + "epoch": 5.793447862564922, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14501 + }, + { + "epoch": 5.793847383140232, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14502 + }, + { + "epoch": 5.794246903715541, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14503 + }, + { + "epoch": 5.794646424290851, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.2985, + "step": 14504 + }, + { + "epoch": 5.79504594486616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14505 + }, + { + "epoch": 5.79544546544147, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14506 + }, + { + "epoch": 5.79584498601678, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14507 + }, + { + "epoch": 5.79624450659209, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14508 + }, + { + "epoch": 5.7966440271674, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14509 + }, + { + "epoch": 5.7970435477427085, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14510 + }, + { + "epoch": 5.797443068318018, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14511 + }, + { + "epoch": 5.797842588893328, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14512 + }, + { + "epoch": 5.798242109468638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14513 + }, + { + "epoch": 5.798641630043948, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14514 + }, + { + "epoch": 5.7990411506192565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14515 + }, + { + "epoch": 5.799440671194566, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14516 + }, + { + "epoch": 5.799840191769876, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14517 + }, + { + "epoch": 5.800239712345186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14518 + }, + { + "epoch": 5.800639232920496, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14519 + }, + { + "epoch": 5.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14520 + }, + { + "epoch": 5.801438274071114, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14521 + }, + { + "epoch": 5.801837794646424, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14522 + }, + { + "epoch": 5.802237315221734, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14523 + }, + { + "epoch": 5.802636835797044, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14524 + }, + { + "epoch": 5.8030363563723535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14525 + }, + { + "epoch": 5.803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14526 + }, + { + "epoch": 5.803835397522972, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14527 + }, + { + "epoch": 5.804234918098282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14528 + }, + { + "epoch": 5.804634438673592, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14529 + }, + { + "epoch": 5.8050339592489015, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14530 + }, + { + "epoch": 5.805433479824211, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14531 + }, + { + "epoch": 5.80583300039952, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 14532 + }, + { + "epoch": 5.80623252097483, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14533 + }, + { + "epoch": 5.80663204155014, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14534 + }, + { + "epoch": 5.80703156212545, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14535 + }, + { + "epoch": 5.807431082700759, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 14536 + }, + { + "epoch": 5.807830603276068, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14537 + }, + { + "epoch": 5.808230123851378, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14538 + }, + { + "epoch": 5.808629644426688, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14539 + }, + { + "epoch": 5.809029165001998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14540 + }, + { + "epoch": 5.809428685577307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14541 + }, + { + "epoch": 5.809828206152617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14542 + }, + { + "epoch": 5.810227726727927, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14543 + }, + { + "epoch": 5.810627247303236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14544 + }, + { + "epoch": 5.811026767878546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14545 + }, + { + "epoch": 5.8114262884538554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14546 + }, + { + "epoch": 5.811825809029165, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14547 + }, + { + "epoch": 5.812225329604475, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14548 + }, + { + "epoch": 5.812624850179784, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14549 + }, + { + "epoch": 5.813024370755094, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14550 + }, + { + "epoch": 5.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 14551 + }, + { + "epoch": 5.813823411905713, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14552 + }, + { + "epoch": 5.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14553 + }, + { + "epoch": 5.814622453056332, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14554 + }, + { + "epoch": 5.815021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14555 + }, + { + "epoch": 5.8154214942069515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14556 + }, + { + "epoch": 5.815821014782261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14557 + }, + { + "epoch": 5.816220535357571, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14558 + }, + { + "epoch": 5.816620055932881, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14559 + }, + { + "epoch": 5.81701957650819, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14560 + }, + { + "epoch": 5.8174190970835, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14561 + }, + { + "epoch": 5.817818617658809, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14562 + }, + { + "epoch": 5.818218138234119, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14563 + }, + { + "epoch": 5.818617658809429, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 14564 + }, + { + "epoch": 5.819017179384739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14565 + }, + { + "epoch": 5.819416699960048, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14566 + }, + { + "epoch": 5.819816220535357, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14567 + }, + { + "epoch": 5.820215741110667, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14568 + }, + { + "epoch": 5.820615261685977, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14569 + }, + { + "epoch": 5.821014782261287, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14570 + }, + { + "epoch": 5.821414302836596, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14571 + }, + { + "epoch": 5.8218138234119055, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14572 + }, + { + "epoch": 5.822213343987215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14573 + }, + { + "epoch": 5.822612864562525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14574 + }, + { + "epoch": 5.823012385137835, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14575 + }, + { + "epoch": 5.823411905713145, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14576 + }, + { + "epoch": 5.8238114262884535, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14577 + }, + { + "epoch": 5.824210946863763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14578 + }, + { + "epoch": 5.824610467439073, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14579 + }, + { + "epoch": 5.825009988014383, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14580 + }, + { + "epoch": 5.825409508589693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14581 + }, + { + "epoch": 5.825809029165002, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14582 + }, + { + "epoch": 5.826208549740311, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14583 + }, + { + "epoch": 5.826608070315621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14584 + }, + { + "epoch": 5.827007590890931, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14585 + }, + { + "epoch": 5.827407111466241, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14586 + }, + { + "epoch": 5.8278066320415505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14587 + }, + { + "epoch": 5.828206152616859, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 14588 + }, + { + "epoch": 5.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14589 + }, + { + "epoch": 5.829005193767479, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14590 + }, + { + "epoch": 5.829404714342789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14591 + }, + { + "epoch": 5.8298042349180985, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14592 + }, + { + "epoch": 5.830203755493407, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14593 + }, + { + "epoch": 5.830603276068717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14594 + }, + { + "epoch": 5.831002796644027, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14595 + }, + { + "epoch": 5.831402317219337, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14596 + }, + { + "epoch": 5.831801837794647, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 14597 + }, + { + "epoch": 5.832201358369956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14598 + }, + { + "epoch": 5.832600878945266, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14599 + }, + { + "epoch": 5.833000399520575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 14600 + }, + { + "epoch": 5.833399920095885, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14601 + }, + { + "epoch": 5.833799440671195, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14602 + }, + { + "epoch": 5.834198961246504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14603 + }, + { + "epoch": 5.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14604 + }, + { + "epoch": 5.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14605 + }, + { + "epoch": 5.835397522972433, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 14606 + }, + { + "epoch": 5.835797043547743, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14607 + }, + { + "epoch": 5.836196564123052, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14608 + }, + { + "epoch": 5.836596084698362, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14609 + }, + { + "epoch": 5.836995605273671, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14610 + }, + { + "epoch": 5.837395125848981, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14611 + }, + { + "epoch": 5.837794646424291, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14612 + }, + { + "epoch": 5.8381941669996005, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14613 + }, + { + "epoch": 5.83859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14614 + }, + { + "epoch": 5.83899320815022, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14615 + }, + { + "epoch": 5.83939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14616 + }, + { + "epoch": 5.839792249300839, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14617 + }, + { + "epoch": 5.8401917698761485, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 14618 + }, + { + "epoch": 5.840591290451458, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14619 + }, + { + "epoch": 5.840990811026768, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14620 + }, + { + "epoch": 5.841390331602078, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14621 + }, + { + "epoch": 5.841789852177387, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14622 + }, + { + "epoch": 5.842189372752697, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14623 + }, + { + "epoch": 5.842588893328006, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14624 + }, + { + "epoch": 5.842988413903316, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14625 + }, + { + "epoch": 5.843387934478626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14626 + }, + { + "epoch": 5.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14627 + }, + { + "epoch": 5.844186975629245, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14628 + }, + { + "epoch": 5.844586496204554, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14629 + }, + { + "epoch": 5.844986016779864, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14630 + }, + { + "epoch": 5.845385537355174, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14631 + }, + { + "epoch": 5.845785057930484, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14632 + }, + { + "epoch": 5.8461845785057935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14633 + }, + { + "epoch": 5.846584099081102, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14634 + }, + { + "epoch": 5.846983619656412, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14635 + }, + { + "epoch": 5.847383140231722, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14636 + }, + { + "epoch": 5.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14637 + }, + { + "epoch": 5.848182181382342, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14638 + }, + { + "epoch": 5.8485817019576505, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14639 + }, + { + "epoch": 5.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14640 + }, + { + "epoch": 5.84938074310827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14641 + }, + { + "epoch": 5.84978026368358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 14642 + }, + { + "epoch": 5.85017978425889, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14643 + }, + { + "epoch": 5.8505793048341985, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14644 + }, + { + "epoch": 5.850978825409508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14645 + }, + { + "epoch": 5.851378345984818, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14646 + }, + { + "epoch": 5.851777866560128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14647 + }, + { + "epoch": 5.852177387135438, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14648 + }, + { + "epoch": 5.8525769077107475, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 14649 + }, + { + "epoch": 5.852976428286056, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14650 + }, + { + "epoch": 5.853375948861366, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 14651 + }, + { + "epoch": 5.853775469436676, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14652 + }, + { + "epoch": 5.854174990011986, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14653 + }, + { + "epoch": 5.8545745105872955, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14654 + }, + { + "epoch": 5.854974031162605, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14655 + }, + { + "epoch": 5.855373551737914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14656 + }, + { + "epoch": 5.855773072313224, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 14657 + }, + { + "epoch": 5.856172592888534, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14658 + }, + { + "epoch": 5.8565721134638435, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14659 + }, + { + "epoch": 5.856971634039153, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14660 + }, + { + "epoch": 5.857371154614462, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14661 + }, + { + "epoch": 5.857770675189772, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14662 + }, + { + "epoch": 5.858170195765082, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14663 + }, + { + "epoch": 5.858569716340392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14664 + }, + { + "epoch": 5.858969236915701, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14665 + }, + { + "epoch": 5.859368757491011, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14666 + }, + { + "epoch": 5.85976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14667 + }, + { + "epoch": 5.86016779864163, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14668 + }, + { + "epoch": 5.86056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14669 + }, + { + "epoch": 5.860966839792249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14670 + }, + { + "epoch": 5.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14671 + }, + { + "epoch": 5.861765880942869, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14672 + }, + { + "epoch": 5.862165401518178, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14673 + }, + { + "epoch": 5.862564922093488, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14674 + }, + { + "epoch": 5.8629644426687975, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14675 + }, + { + "epoch": 5.863363963244107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14676 + }, + { + "epoch": 5.863763483819417, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 14677 + }, + { + "epoch": 5.864163004394726, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14678 + }, + { + "epoch": 5.864562524970036, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14679 + }, + { + "epoch": 5.8649620455453455, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 14680 + }, + { + "epoch": 5.865361566120655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14681 + }, + { + "epoch": 5.865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14682 + }, + { + "epoch": 5.866160607271274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14683 + }, + { + "epoch": 5.866560127846584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14684 + }, + { + "epoch": 5.8669596484218935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14685 + }, + { + "epoch": 5.867359168997203, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14686 + }, + { + "epoch": 5.867758689572513, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14687 + }, + { + "epoch": 5.868158210147823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14688 + }, + { + "epoch": 5.868557730723133, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14689 + }, + { + "epoch": 5.868957251298442, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 14690 + }, + { + "epoch": 5.869356771873751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14691 + }, + { + "epoch": 5.869756292449061, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14692 + }, + { + "epoch": 5.870155813024371, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14693 + }, + { + "epoch": 5.870555333599681, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14694 + }, + { + "epoch": 5.87095485417499, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14695 + }, + { + "epoch": 5.871354374750299, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14696 + }, + { + "epoch": 5.871753895325609, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14697 + }, + { + "epoch": 5.872153415900919, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14698 + }, + { + "epoch": 5.872552936476229, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14699 + }, + { + "epoch": 5.872952457051538, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14700 + }, + { + "epoch": 5.8733519776268475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14701 + }, + { + "epoch": 5.873751498202157, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 14702 + }, + { + "epoch": 5.874151018777467, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 14703 + }, + { + "epoch": 5.874550539352777, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14704 + }, + { + "epoch": 5.874950059928087, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14705 + }, + { + "epoch": 5.875349580503396, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14706 + }, + { + "epoch": 5.875749101078705, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14707 + }, + { + "epoch": 5.876148621654015, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14708 + }, + { + "epoch": 5.876548142229325, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 14709 + }, + { + "epoch": 5.876947662804635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 14710 + }, + { + "epoch": 5.877347183379944, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14711 + }, + { + "epoch": 5.877746703955253, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14712 + }, + { + "epoch": 5.878146224530563, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14713 + }, + { + "epoch": 5.878545745105873, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14714 + }, + { + "epoch": 5.878945265681183, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14715 + }, + { + "epoch": 5.8793447862564925, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14716 + }, + { + "epoch": 5.879744306831801, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14717 + }, + { + "epoch": 5.880143827407111, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14718 + }, + { + "epoch": 5.880543347982421, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14719 + }, + { + "epoch": 5.880942868557731, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14720 + }, + { + "epoch": 5.8813423891330405, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14721 + }, + { + "epoch": 5.88174190970835, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14722 + }, + { + "epoch": 5.88214143028366, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 14723 + }, + { + "epoch": 5.882540950858969, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14724 + }, + { + "epoch": 5.882940471434279, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14725 + }, + { + "epoch": 5.883339992009589, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 14726 + }, + { + "epoch": 5.883739512584898, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14727 + }, + { + "epoch": 5.884139033160208, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14728 + }, + { + "epoch": 5.884538553735517, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14729 + }, + { + "epoch": 5.884938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14730 + }, + { + "epoch": 5.885337594886137, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14731 + }, + { + "epoch": 5.885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14732 + }, + { + "epoch": 5.886136636036756, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14733 + }, + { + "epoch": 5.886536156612065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 14734 + }, + { + "epoch": 5.886935677187375, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14735 + }, + { + "epoch": 5.887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14736 + }, + { + "epoch": 5.887734718337994, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14737 + }, + { + "epoch": 5.888134238913304, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14738 + }, + { + "epoch": 5.888533759488614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 14739 + }, + { + "epoch": 5.888933280063923, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14740 + }, + { + "epoch": 5.889332800639233, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 14741 + }, + { + "epoch": 5.8897323212145425, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14742 + }, + { + "epoch": 5.890131841789852, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 14743 + }, + { + "epoch": 5.890531362365162, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14744 + }, + { + "epoch": 5.890930882940472, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14745 + }, + { + "epoch": 5.891330403515781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14746 + }, + { + "epoch": 5.8917299240910905, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14747 + }, + { + "epoch": 5.8921294446664, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14748 + }, + { + "epoch": 5.89252896524171, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 14749 + }, + { + "epoch": 5.89292848581702, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14750 + }, + { + "epoch": 5.893328006392329, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14751 + }, + { + "epoch": 5.893727526967639, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14752 + }, + { + "epoch": 5.894127047542948, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14753 + }, + { + "epoch": 5.894526568118258, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14754 + }, + { + "epoch": 5.894926088693568, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14755 + }, + { + "epoch": 5.895325609268878, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14756 + }, + { + "epoch": 5.895725129844187, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14757 + }, + { + "epoch": 5.896124650419496, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 14758 + }, + { + "epoch": 5.896524170994806, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14759 + }, + { + "epoch": 5.896923691570116, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14760 + }, + { + "epoch": 5.897323212145426, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14761 + }, + { + "epoch": 5.8977227327207356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14762 + }, + { + "epoch": 5.8981222532960444, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14763 + }, + { + "epoch": 5.898521773871354, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14764 + }, + { + "epoch": 5.898921294446664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14765 + }, + { + "epoch": 5.899320815021974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14766 + }, + { + "epoch": 5.899720335597284, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14767 + }, + { + "epoch": 5.9001198561725925, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14768 + }, + { + "epoch": 5.900519376747902, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14769 + }, + { + "epoch": 5.900918897323212, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14770 + }, + { + "epoch": 5.901318417898522, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 14771 + }, + { + "epoch": 5.901717938473832, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14772 + }, + { + "epoch": 5.9021174590491405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14773 + }, + { + "epoch": 5.90251697962445, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14774 + }, + { + "epoch": 5.90291650019976, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14775 + }, + { + "epoch": 5.90331602077507, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 14776 + }, + { + "epoch": 5.90371554135038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14777 + }, + { + "epoch": 5.9041150619256895, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14778 + }, + { + "epoch": 5.904514582500999, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14779 + }, + { + "epoch": 5.904914103076308, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14780 + }, + { + "epoch": 5.905313623651618, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14781 + }, + { + "epoch": 5.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 14782 + }, + { + "epoch": 5.9061126648022375, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14783 + }, + { + "epoch": 5.906512185377547, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14784 + }, + { + "epoch": 5.906911705952856, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14785 + }, + { + "epoch": 5.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14786 + }, + { + "epoch": 5.907710747103476, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14787 + }, + { + "epoch": 5.9081102676787856, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14788 + }, + { + "epoch": 5.908509788254095, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14789 + }, + { + "epoch": 5.908909308829404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14790 + }, + { + "epoch": 5.909308829404714, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14791 + }, + { + "epoch": 5.909708349980024, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14792 + }, + { + "epoch": 5.910107870555334, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14793 + }, + { + "epoch": 5.910507391130643, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14794 + }, + { + "epoch": 5.910906911705953, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14795 + }, + { + "epoch": 5.911306432281263, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 14796 + }, + { + "epoch": 5.911705952856572, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14797 + }, + { + "epoch": 5.912105473431882, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14798 + }, + { + "epoch": 5.912504994007191, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14799 + }, + { + "epoch": 5.912904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 14800 + }, + { + "epoch": 5.913304035157811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14801 + }, + { + "epoch": 5.91370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14802 + }, + { + "epoch": 5.91410307630843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14803 + }, + { + "epoch": 5.9145025968837395, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14804 + }, + { + "epoch": 5.914902117459049, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 14805 + }, + { + "epoch": 5.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14806 + }, + { + "epoch": 5.915701158609668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14807 + }, + { + "epoch": 5.916100679184978, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14808 + }, + { + "epoch": 5.9165001997602875, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14809 + }, + { + "epoch": 5.916899720335597, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14810 + }, + { + "epoch": 5.917299240910907, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14811 + }, + { + "epoch": 5.917698761486217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14812 + }, + { + "epoch": 5.918098282061527, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14813 + }, + { + "epoch": 5.918497802636836, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14814 + }, + { + "epoch": 5.918897323212145, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14815 + }, + { + "epoch": 5.919296843787455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14816 + }, + { + "epoch": 5.919696364362765, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14817 + }, + { + "epoch": 5.920095884938075, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14818 + }, + { + "epoch": 5.920495405513384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14819 + }, + { + "epoch": 5.920894926088693, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14820 + }, + { + "epoch": 5.921294446664003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14821 + }, + { + "epoch": 5.921693967239313, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14822 + }, + { + "epoch": 5.922093487814623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14823 + }, + { + "epoch": 5.922493008389932, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14824 + }, + { + "epoch": 5.922892528965241, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14825 + }, + { + "epoch": 5.923292049540551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14826 + }, + { + "epoch": 5.923691570115861, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14827 + }, + { + "epoch": 5.924091090691171, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14828 + }, + { + "epoch": 5.924490611266481, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14829 + }, + { + "epoch": 5.9248901318417895, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14830 + }, + { + "epoch": 5.925289652417099, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14831 + }, + { + "epoch": 5.925689172992409, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14832 + }, + { + "epoch": 5.926088693567719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14833 + }, + { + "epoch": 5.926488214143029, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14834 + }, + { + "epoch": 5.926887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14835 + }, + { + "epoch": 5.927287255293647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14836 + }, + { + "epoch": 5.927686775868957, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 14837 + }, + { + "epoch": 5.928086296444267, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14838 + }, + { + "epoch": 5.928485817019577, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14839 + }, + { + "epoch": 5.9288853375948865, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14840 + }, + { + "epoch": 5.929284858170195, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 14841 + }, + { + "epoch": 5.929684378745505, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14842 + }, + { + "epoch": 5.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14843 + }, + { + "epoch": 5.930483419896125, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 14844 + }, + { + "epoch": 5.9308829404714345, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 14845 + }, + { + "epoch": 5.931282461046744, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14846 + }, + { + "epoch": 5.931681981622053, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14847 + }, + { + "epoch": 5.932081502197363, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14848 + }, + { + "epoch": 5.932481022772673, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14849 + }, + { + "epoch": 5.9328805433479825, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14850 + }, + { + "epoch": 5.933280063923292, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14851 + }, + { + "epoch": 5.933679584498602, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14852 + }, + { + "epoch": 5.934079105073911, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14853 + }, + { + "epoch": 5.934478625649221, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14854 + }, + { + "epoch": 5.934878146224531, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14855 + }, + { + "epoch": 5.93527766679984, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14856 + }, + { + "epoch": 5.93567718737515, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14857 + }, + { + "epoch": 5.936076707950459, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14858 + }, + { + "epoch": 5.936476228525769, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14859 + }, + { + "epoch": 5.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 14860 + }, + { + "epoch": 5.937275269676388, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 14861 + }, + { + "epoch": 5.937674790251698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14862 + }, + { + "epoch": 5.938074310827007, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14863 + }, + { + "epoch": 5.938473831402317, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14864 + }, + { + "epoch": 5.938873351977627, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14865 + }, + { + "epoch": 5.9392728725529365, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 14866 + }, + { + "epoch": 5.939672393128246, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14867 + }, + { + "epoch": 5.940071913703556, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14868 + }, + { + "epoch": 5.940471434278866, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14869 + }, + { + "epoch": 5.940870954854175, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14870 + }, + { + "epoch": 5.9412704754294845, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14871 + }, + { + "epoch": 5.941669996004794, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14872 + }, + { + "epoch": 5.942069516580104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14873 + }, + { + "epoch": 5.942469037155414, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14874 + }, + { + "epoch": 5.942868557730723, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14875 + }, + { + "epoch": 5.9432680783060325, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14876 + }, + { + "epoch": 5.943667598881342, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14877 + }, + { + "epoch": 5.944067119456652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14878 + }, + { + "epoch": 5.944466640031962, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14879 + }, + { + "epoch": 5.944866160607271, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 14880 + }, + { + "epoch": 5.945265681182581, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14881 + }, + { + "epoch": 5.94566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14882 + }, + { + "epoch": 5.9460647223332, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14883 + }, + { + "epoch": 5.94646424290851, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14884 + }, + { + "epoch": 5.94686376348382, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14885 + }, + { + "epoch": 5.9472632840591295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14886 + }, + { + "epoch": 5.947662804634438, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 14887 + }, + { + "epoch": 5.948062325209748, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14888 + }, + { + "epoch": 5.948461845785058, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14889 + }, + { + "epoch": 5.948861366360368, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14890 + }, + { + "epoch": 5.949260886935678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14891 + }, + { + "epoch": 5.9496604075109865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14892 + }, + { + "epoch": 5.950059928086296, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14893 + }, + { + "epoch": 5.950459448661606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14894 + }, + { + "epoch": 5.950858969236916, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14895 + }, + { + "epoch": 5.951258489812226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 14896 + }, + { + "epoch": 5.9516580103875345, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 14897 + }, + { + "epoch": 5.952057530962844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14898 + }, + { + "epoch": 5.952457051538154, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14899 + }, + { + "epoch": 5.952856572113464, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14900 + }, + { + "epoch": 5.953256092688774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14901 + }, + { + "epoch": 5.953655613264083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14902 + }, + { + "epoch": 5.954055133839393, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14903 + }, + { + "epoch": 5.954454654414702, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14904 + }, + { + "epoch": 5.954854174990012, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14905 + }, + { + "epoch": 5.955253695565322, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14906 + }, + { + "epoch": 5.9556532161406315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14907 + }, + { + "epoch": 5.956052736715941, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14908 + }, + { + "epoch": 5.95645225729125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14909 + }, + { + "epoch": 5.95685177786656, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14910 + }, + { + "epoch": 5.95725129844187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14911 + }, + { + "epoch": 5.9576508190171795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14912 + }, + { + "epoch": 5.958050339592489, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14913 + }, + { + "epoch": 5.958449860167798, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14914 + }, + { + "epoch": 5.958849380743108, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14915 + }, + { + "epoch": 5.959248901318418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14916 + }, + { + "epoch": 5.959648421893728, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14917 + }, + { + "epoch": 5.960047942469037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14918 + }, + { + "epoch": 5.960447463044347, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14919 + }, + { + "epoch": 5.960846983619656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14920 + }, + { + "epoch": 5.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14921 + }, + { + "epoch": 5.961646024770276, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14922 + }, + { + "epoch": 5.962045545345585, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14923 + }, + { + "epoch": 5.962445065920895, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14924 + }, + { + "epoch": 5.962844586496205, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14925 + }, + { + "epoch": 5.963244107071514, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14926 + }, + { + "epoch": 5.963643627646824, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14927 + }, + { + "epoch": 5.964043148222133, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14928 + }, + { + "epoch": 5.964442668797443, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14929 + }, + { + "epoch": 5.964842189372753, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14930 + }, + { + "epoch": 5.965241709948062, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14931 + }, + { + "epoch": 5.965641230523372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14932 + }, + { + "epoch": 5.9660407510986815, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14933 + }, + { + "epoch": 5.966440271673991, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14934 + }, + { + "epoch": 5.966839792249301, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14935 + }, + { + "epoch": 5.967239312824611, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14936 + }, + { + "epoch": 5.96763883339992, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14937 + }, + { + "epoch": 5.9680383539752295, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14938 + }, + { + "epoch": 5.968437874550539, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14939 + }, + { + "epoch": 5.968837395125849, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14940 + }, + { + "epoch": 5.969236915701159, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14941 + }, + { + "epoch": 5.969636436276469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14942 + }, + { + "epoch": 5.970035956851778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14943 + }, + { + "epoch": 5.970435477427087, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14944 + }, + { + "epoch": 5.970834998002397, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 14945 + }, + { + "epoch": 5.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14946 + }, + { + "epoch": 5.971634039153017, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14947 + }, + { + "epoch": 5.972033559728326, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14948 + }, + { + "epoch": 5.972433080303635, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14949 + }, + { + "epoch": 5.972832600878945, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14950 + }, + { + "epoch": 5.973232121454255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14951 + }, + { + "epoch": 5.973631642029565, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14952 + }, + { + "epoch": 5.974031162604874, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14953 + }, + { + "epoch": 5.974430683180183, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14954 + }, + { + "epoch": 5.974830203755493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14955 + }, + { + "epoch": 5.975229724330803, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14956 + }, + { + "epoch": 5.975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14957 + }, + { + "epoch": 5.976028765481423, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14958 + }, + { + "epoch": 5.976428286056732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14959 + }, + { + "epoch": 5.976827806632041, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 14960 + }, + { + "epoch": 5.977227327207351, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14961 + }, + { + "epoch": 5.977626847782661, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14962 + }, + { + "epoch": 5.978026368357971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14963 + }, + { + "epoch": 5.97842588893328, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14964 + }, + { + "epoch": 5.978825409508589, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14965 + }, + { + "epoch": 5.979224930083899, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14966 + }, + { + "epoch": 5.979624450659209, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14967 + }, + { + "epoch": 5.980023971234519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14968 + }, + { + "epoch": 5.9804234918098285, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14969 + }, + { + "epoch": 5.980823012385137, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14970 + }, + { + "epoch": 5.981222532960447, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14971 + }, + { + "epoch": 5.981622053535757, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 14972 + }, + { + "epoch": 5.982021574111067, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14973 + }, + { + "epoch": 5.9824210946863765, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14974 + }, + { + "epoch": 5.982820615261686, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14975 + }, + { + "epoch": 5.983220135836996, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14976 + }, + { + "epoch": 5.983619656412305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14977 + }, + { + "epoch": 5.984019176987615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14978 + }, + { + "epoch": 5.9844186975629245, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14979 + }, + { + "epoch": 5.984818218138234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14980 + }, + { + "epoch": 5.985217738713544, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14981 + }, + { + "epoch": 5.985617259288853, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14982 + }, + { + "epoch": 5.986016779864163, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14983 + }, + { + "epoch": 5.986416300439473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14984 + }, + { + "epoch": 5.986815821014782, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14985 + }, + { + "epoch": 5.987215341590092, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14986 + }, + { + "epoch": 5.987614862165401, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14987 + }, + { + "epoch": 5.988014382740711, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14988 + }, + { + "epoch": 5.988413903316021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14989 + }, + { + "epoch": 5.98881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14990 + }, + { + "epoch": 5.98921294446664, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14991 + }, + { + "epoch": 5.98961246504195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14992 + }, + { + "epoch": 5.99001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14993 + }, + { + "epoch": 5.990411506192569, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14994 + }, + { + "epoch": 5.9908110267678785, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14995 + }, + { + "epoch": 5.991210547343188, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14996 + }, + { + "epoch": 5.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14997 + }, + { + "epoch": 5.992009588493808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14998 + }, + { + "epoch": 5.992409109069117, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14999 + }, + { + "epoch": 5.9928086296444265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15000 + }, + { + "epoch": 5.993208150219736, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15001 + }, + { + "epoch": 5.993607670795046, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 15002 + }, + { + "epoch": 5.994007191370356, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15003 + }, + { + "epoch": 5.994406711945665, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15004 + }, + { + "epoch": 5.9948062325209746, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15005 + }, + { + "epoch": 5.995205753096284, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 15006 + }, + { + "epoch": 5.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15007 + }, + { + "epoch": 5.996004794246904, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 15008 + }, + { + "epoch": 5.996404314822214, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15009 + }, + { + "epoch": 5.996803835397523, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15010 + }, + { + "epoch": 5.997203355972832, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15011 + }, + { + "epoch": 5.997602876548142, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15012 + }, + { + "epoch": 5.998002397123452, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15013 + }, + { + "epoch": 5.998401917698762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15014 + }, + { + "epoch": 5.9988014382740715, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15015 + }, + { + "epoch": 5.99920095884938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15016 + }, + { + "epoch": 5.99960047942469, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 15017 + }, + { + "epoch": 6.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15018 + }, + { + "epoch": 6.00039952057531, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15019 + }, + { + "epoch": 6.00079904115062, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15020 + }, + { + "epoch": 6.0011985617259285, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15021 + }, + { + "epoch": 6.001598082301238, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15022 + }, + { + "epoch": 6.001997602876548, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15023 + }, + { + "epoch": 6.002397123451858, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15024 + }, + { + "epoch": 6.002796644027168, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 15025 + }, + { + "epoch": 6.003196164602477, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15026 + }, + { + "epoch": 6.003595685177786, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15027 + }, + { + "epoch": 6.003995205753096, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15028 + }, + { + "epoch": 6.004394726328406, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15029 + }, + { + "epoch": 6.004794246903716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15030 + }, + { + "epoch": 6.0051937674790254, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15031 + }, + { + "epoch": 6.005593288054335, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15032 + }, + { + "epoch": 6.005992808629644, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15033 + }, + { + "epoch": 6.006392329204954, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15034 + }, + { + "epoch": 6.006791849780264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15035 + }, + { + "epoch": 6.0071913703555735, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 15036 + }, + { + "epoch": 6.007590890930883, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15037 + }, + { + "epoch": 6.007990411506192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15038 + }, + { + "epoch": 6.008389932081502, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15039 + }, + { + "epoch": 6.008789452656812, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15040 + }, + { + "epoch": 6.0091889732321215, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15041 + }, + { + "epoch": 6.009588493807431, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15042 + }, + { + "epoch": 6.009988014382741, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15043 + }, + { + "epoch": 6.01038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15044 + }, + { + "epoch": 6.01078705553336, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15045 + }, + { + "epoch": 6.01118657610867, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15046 + }, + { + "epoch": 6.011586096683979, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15047 + }, + { + "epoch": 6.011985617259289, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15048 + }, + { + "epoch": 6.012385137834598, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15049 + }, + { + "epoch": 6.012784658409908, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15050 + }, + { + "epoch": 6.013184178985218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15051 + }, + { + "epoch": 6.013583699560527, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15052 + }, + { + "epoch": 6.013983220135837, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15053 + }, + { + "epoch": 6.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15054 + }, + { + "epoch": 6.014782261286456, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15055 + }, + { + "epoch": 6.015181781861766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15056 + }, + { + "epoch": 6.0155813024370755, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15057 + }, + { + "epoch": 6.015980823012385, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 15058 + }, + { + "epoch": 6.016380343587695, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15059 + }, + { + "epoch": 6.016779864163005, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 15060 + }, + { + "epoch": 6.017179384738314, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15061 + }, + { + "epoch": 6.0175789053136235, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15062 + }, + { + "epoch": 6.017978425888933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15063 + }, + { + "epoch": 6.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15064 + }, + { + "epoch": 6.018777467039553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15065 + }, + { + "epoch": 6.019176987614862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15066 + }, + { + "epoch": 6.0195765081901715, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15067 + }, + { + "epoch": 6.019976028765481, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15068 + }, + { + "epoch": 6.020375549340791, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15069 + }, + { + "epoch": 6.020775069916101, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 15070 + }, + { + "epoch": 6.021174590491411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15071 + }, + { + "epoch": 6.02157411106672, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15072 + }, + { + "epoch": 6.021973631642029, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15073 + }, + { + "epoch": 6.022373152217339, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15074 + }, + { + "epoch": 6.022772672792649, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 15075 + }, + { + "epoch": 6.023172193367959, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15076 + }, + { + "epoch": 6.0235717139432685, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15077 + }, + { + "epoch": 6.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15078 + }, + { + "epoch": 6.024370755093887, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15079 + }, + { + "epoch": 6.024770275669197, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15080 + }, + { + "epoch": 6.025169796244507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15081 + }, + { + "epoch": 6.025569316819817, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15082 + }, + { + "epoch": 6.0259688373951255, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15083 + }, + { + "epoch": 6.026368357970435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15084 + }, + { + "epoch": 6.026767878545745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15085 + }, + { + "epoch": 6.027167399121055, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15086 + }, + { + "epoch": 6.027566919696365, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15087 + }, + { + "epoch": 6.027966440271674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15088 + }, + { + "epoch": 6.028365960846983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15089 + }, + { + "epoch": 6.028765481422293, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15090 + }, + { + "epoch": 6.029165001997603, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15091 + }, + { + "epoch": 6.029564522572913, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15092 + }, + { + "epoch": 6.029964043148222, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15093 + }, + { + "epoch": 6.030363563723531, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15094 + }, + { + "epoch": 6.030763084298841, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15095 + }, + { + "epoch": 6.031162604874151, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15096 + }, + { + "epoch": 6.031562125449461, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15097 + }, + { + "epoch": 6.0319616460247705, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15098 + }, + { + "epoch": 6.03236116660008, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15099 + }, + { + "epoch": 6.032760687175389, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15100 + }, + { + "epoch": 6.033160207750699, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15101 + }, + { + "epoch": 6.033559728326009, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15102 + }, + { + "epoch": 6.0339592489013185, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15103 + }, + { + "epoch": 6.034358769476628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15104 + }, + { + "epoch": 6.034758290051938, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15105 + }, + { + "epoch": 6.035157810627247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15106 + }, + { + "epoch": 6.035557331202557, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15107 + }, + { + "epoch": 6.035956851777867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 15108 + }, + { + "epoch": 6.036356372353176, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15109 + }, + { + "epoch": 6.036755892928486, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15110 + }, + { + "epoch": 6.037155413503795, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15111 + }, + { + "epoch": 6.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15112 + }, + { + "epoch": 6.037954454654415, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15113 + }, + { + "epoch": 6.038353975229724, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15114 + }, + { + "epoch": 6.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15115 + }, + { + "epoch": 6.039153016380344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15116 + }, + { + "epoch": 6.039552536955653, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15117 + }, + { + "epoch": 6.039952057530963, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15118 + }, + { + "epoch": 6.040351578106272, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15119 + }, + { + "epoch": 6.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15120 + }, + { + "epoch": 6.041150619256892, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15121 + }, + { + "epoch": 6.041550139832202, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15122 + }, + { + "epoch": 6.041949660407511, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15123 + }, + { + "epoch": 6.0423491809828205, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 15124 + }, + { + "epoch": 6.04274870155813, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15125 + }, + { + "epoch": 6.04314822213344, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15126 + }, + { + "epoch": 6.04354774270875, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15127 + }, + { + "epoch": 6.043947263284059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15128 + }, + { + "epoch": 6.0443467838593685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15129 + }, + { + "epoch": 6.044746304434678, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15130 + }, + { + "epoch": 6.045145825009988, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15131 + }, + { + "epoch": 6.045545345585298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15132 + }, + { + "epoch": 6.045944866160608, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15133 + }, + { + "epoch": 6.046344386735917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15134 + }, + { + "epoch": 6.046743907311226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15135 + }, + { + "epoch": 6.047143427886536, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15136 + }, + { + "epoch": 6.047542948461846, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15137 + }, + { + "epoch": 6.047942469037156, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15138 + }, + { + "epoch": 6.048341989612465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15139 + }, + { + "epoch": 6.048741510187774, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15140 + }, + { + "epoch": 6.049141030763084, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15141 + }, + { + "epoch": 6.049540551338394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15142 + }, + { + "epoch": 6.049940071913704, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15143 + }, + { + "epoch": 6.0503395924890135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15144 + }, + { + "epoch": 6.050739113064322, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15145 + }, + { + "epoch": 6.051138633639632, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15146 + }, + { + "epoch": 6.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15147 + }, + { + "epoch": 6.051937674790252, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15148 + }, + { + "epoch": 6.052337195365562, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15149 + }, + { + "epoch": 6.052736715940871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15150 + }, + { + "epoch": 6.05313623651618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15151 + }, + { + "epoch": 6.05353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15152 + }, + { + "epoch": 6.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 15153 + }, + { + "epoch": 6.05433479824211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 15154 + }, + { + "epoch": 6.054734318817419, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15155 + }, + { + "epoch": 6.055133839392728, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15156 + }, + { + "epoch": 6.055533359968038, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15157 + }, + { + "epoch": 6.055932880543348, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15158 + }, + { + "epoch": 6.056332401118658, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15159 + }, + { + "epoch": 6.0567319216939675, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15160 + }, + { + "epoch": 6.057131442269277, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15161 + }, + { + "epoch": 6.057530962844586, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15162 + }, + { + "epoch": 6.057930483419896, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15163 + }, + { + "epoch": 6.058330003995206, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15164 + }, + { + "epoch": 6.0587295245705155, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15165 + }, + { + "epoch": 6.059129045145825, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15166 + }, + { + "epoch": 6.059528565721135, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15167 + }, + { + "epoch": 6.059928086296444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15168 + }, + { + "epoch": 6.060327606871754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15169 + }, + { + "epoch": 6.0607271274470635, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15170 + }, + { + "epoch": 6.061126648022373, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15171 + }, + { + "epoch": 6.061526168597683, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15172 + }, + { + "epoch": 6.061925689172992, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15173 + }, + { + "epoch": 6.062325209748302, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15174 + }, + { + "epoch": 6.062724730323612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 15175 + }, + { + "epoch": 6.063124250898921, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15176 + }, + { + "epoch": 6.063523771474231, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15177 + }, + { + "epoch": 6.063923292049541, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15178 + }, + { + "epoch": 6.06432281262485, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 15179 + }, + { + "epoch": 6.06472233320016, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15180 + }, + { + "epoch": 6.065121853775469, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15181 + }, + { + "epoch": 6.065521374350779, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15182 + }, + { + "epoch": 6.065920894926089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15183 + }, + { + "epoch": 6.066320415501398, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15184 + }, + { + "epoch": 6.066719936076708, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15185 + }, + { + "epoch": 6.0671194566520175, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15186 + }, + { + "epoch": 6.067518977227327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15187 + }, + { + "epoch": 6.067918497802637, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15188 + }, + { + "epoch": 6.068318018377947, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15189 + }, + { + "epoch": 6.068717538953256, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15190 + }, + { + "epoch": 6.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15191 + }, + { + "epoch": 6.069516580103875, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15192 + }, + { + "epoch": 6.069916100679185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15193 + }, + { + "epoch": 6.070315621254495, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15194 + }, + { + "epoch": 6.070715141829805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15195 + }, + { + "epoch": 6.0711146624051135, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15196 + }, + { + "epoch": 6.071514182980423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15197 + }, + { + "epoch": 6.071913703555733, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15198 + }, + { + "epoch": 6.072313224131043, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15199 + }, + { + "epoch": 6.072712744706353, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15200 + }, + { + "epoch": 6.073112265281662, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15201 + }, + { + "epoch": 6.073511785856971, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15202 + }, + { + "epoch": 6.073911306432281, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15203 + }, + { + "epoch": 6.074310827007591, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15204 + }, + { + "epoch": 6.074710347582901, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15205 + }, + { + "epoch": 6.0751098681582105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15206 + }, + { + "epoch": 6.075509388733519, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15207 + }, + { + "epoch": 6.075908909308829, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15208 + }, + { + "epoch": 6.076308429884139, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15209 + }, + { + "epoch": 6.076707950459449, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15210 + }, + { + "epoch": 6.077107471034759, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15211 + }, + { + "epoch": 6.077506991610068, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15212 + }, + { + "epoch": 6.077906512185377, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15213 + }, + { + "epoch": 6.078306032760687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15214 + }, + { + "epoch": 6.078705553335997, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15215 + }, + { + "epoch": 6.079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15216 + }, + { + "epoch": 6.079504594486616, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15217 + }, + { + "epoch": 6.079904115061925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15218 + }, + { + "epoch": 6.080303635637235, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15219 + }, + { + "epoch": 6.080703156212545, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 15220 + }, + { + "epoch": 6.081102676787855, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15221 + }, + { + "epoch": 6.081502197363164, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15222 + }, + { + "epoch": 6.081901717938474, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15223 + }, + { + "epoch": 6.082301238513783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15224 + }, + { + "epoch": 6.082700759089093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15225 + }, + { + "epoch": 6.083100279664403, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15226 + }, + { + "epoch": 6.0834998002397125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15227 + }, + { + "epoch": 6.083899320815022, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15228 + }, + { + "epoch": 6.084298841390331, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15229 + }, + { + "epoch": 6.084698361965641, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15230 + }, + { + "epoch": 6.085097882540951, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15231 + }, + { + "epoch": 6.0854974031162605, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15232 + }, + { + "epoch": 6.08589692369157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15233 + }, + { + "epoch": 6.08629644426688, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15234 + }, + { + "epoch": 6.086695964842189, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15235 + }, + { + "epoch": 6.087095485417499, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15236 + }, + { + "epoch": 6.087495005992809, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15237 + }, + { + "epoch": 6.087894526568118, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15238 + }, + { + "epoch": 6.088294047143428, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15239 + }, + { + "epoch": 6.088693567718738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15240 + }, + { + "epoch": 6.089093088294047, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15241 + }, + { + "epoch": 6.089492608869357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15242 + }, + { + "epoch": 6.089892129444666, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15243 + }, + { + "epoch": 6.090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15244 + }, + { + "epoch": 6.090691170595286, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15245 + }, + { + "epoch": 6.091090691170595, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15246 + }, + { + "epoch": 6.091490211745905, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15247 + }, + { + "epoch": 6.0918897323212144, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15248 + }, + { + "epoch": 6.092289252896524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15249 + }, + { + "epoch": 6.092688773471834, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15250 + }, + { + "epoch": 6.093088294047144, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15251 + }, + { + "epoch": 6.093487814622453, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15252 + }, + { + "epoch": 6.0938873351977625, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15253 + }, + { + "epoch": 6.094286855773072, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15254 + }, + { + "epoch": 6.094686376348382, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15255 + }, + { + "epoch": 6.095085896923692, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15256 + }, + { + "epoch": 6.095485417499002, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15257 + }, + { + "epoch": 6.0958849380743105, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15258 + }, + { + "epoch": 6.09628445864962, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15259 + }, + { + "epoch": 6.09668397922493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15260 + }, + { + "epoch": 6.09708349980024, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15261 + }, + { + "epoch": 6.09748302037555, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15262 + }, + { + "epoch": 6.097882540950859, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15263 + }, + { + "epoch": 6.098282061526168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15264 + }, + { + "epoch": 6.098681582101478, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15265 + }, + { + "epoch": 6.099081102676788, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15266 + }, + { + "epoch": 6.099480623252098, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15267 + }, + { + "epoch": 6.0998801438274075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15268 + }, + { + "epoch": 6.100279664402716, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15269 + }, + { + "epoch": 6.100679184978026, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15270 + }, + { + "epoch": 6.101078705553336, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15271 + }, + { + "epoch": 6.101478226128646, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15272 + }, + { + "epoch": 6.1018777467039556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15273 + }, + { + "epoch": 6.1022772672792644, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15274 + }, + { + "epoch": 6.102676787854574, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15275 + }, + { + "epoch": 6.103076308429884, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15276 + }, + { + "epoch": 6.103475829005194, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15277 + }, + { + "epoch": 6.103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15278 + }, + { + "epoch": 6.104274870155813, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15279 + }, + { + "epoch": 6.104674390731122, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15280 + }, + { + "epoch": 6.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15281 + }, + { + "epoch": 6.105473431881742, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15282 + }, + { + "epoch": 6.105872952457052, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15283 + }, + { + "epoch": 6.106272473032361, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15284 + }, + { + "epoch": 6.106671993607671, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15285 + }, + { + "epoch": 6.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15286 + }, + { + "epoch": 6.10747103475829, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15287 + }, + { + "epoch": 6.1078705553336, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15288 + }, + { + "epoch": 6.1082700759089095, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15289 + }, + { + "epoch": 6.108669596484219, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15290 + }, + { + "epoch": 6.109069117059528, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15291 + }, + { + "epoch": 6.109468637634838, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15292 + }, + { + "epoch": 6.109868158210148, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15293 + }, + { + "epoch": 6.1102676787854575, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15294 + }, + { + "epoch": 6.110667199360767, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15295 + }, + { + "epoch": 6.111066719936077, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15296 + }, + { + "epoch": 6.111466240511386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15297 + }, + { + "epoch": 6.111865761086696, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15298 + }, + { + "epoch": 6.112265281662006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15299 + }, + { + "epoch": 6.112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15300 + }, + { + "epoch": 6.113064322812625, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15301 + }, + { + "epoch": 6.113463843387935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15302 + }, + { + "epoch": 6.113863363963244, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15303 + }, + { + "epoch": 6.114262884538554, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15304 + }, + { + "epoch": 6.114662405113863, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15305 + }, + { + "epoch": 6.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15306 + }, + { + "epoch": 6.115461446264483, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15307 + }, + { + "epoch": 6.115860966839792, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15308 + }, + { + "epoch": 6.116260487415102, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15309 + }, + { + "epoch": 6.116660007990411, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15310 + }, + { + "epoch": 6.117059528565721, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15311 + }, + { + "epoch": 6.117459049141031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 15312 + }, + { + "epoch": 6.117858569716341, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15313 + }, + { + "epoch": 6.11825809029165, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15314 + }, + { + "epoch": 6.1186576108669595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15315 + }, + { + "epoch": 6.119057131442269, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 15316 + }, + { + "epoch": 6.119456652017579, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15317 + }, + { + "epoch": 6.119856172592889, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15318 + }, + { + "epoch": 6.120255693168198, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15319 + }, + { + "epoch": 6.1206552137435075, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15320 + }, + { + "epoch": 6.121054734318817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15321 + }, + { + "epoch": 6.121454254894127, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15322 + }, + { + "epoch": 6.121853775469437, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15323 + }, + { + "epoch": 6.122253296044747, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15324 + }, + { + "epoch": 6.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15325 + }, + { + "epoch": 6.123052337195365, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15326 + }, + { + "epoch": 6.123451857770675, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15327 + }, + { + "epoch": 6.123851378345985, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15328 + }, + { + "epoch": 6.124250898921295, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15329 + }, + { + "epoch": 6.1246504194966045, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15330 + }, + { + "epoch": 6.125049940071913, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 15331 + }, + { + "epoch": 6.125449460647223, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15332 + }, + { + "epoch": 6.125848981222533, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15333 + }, + { + "epoch": 6.126248501797843, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15334 + }, + { + "epoch": 6.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15335 + }, + { + "epoch": 6.127047542948461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15336 + }, + { + "epoch": 6.127447063523771, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15337 + }, + { + "epoch": 6.127846584099081, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15338 + }, + { + "epoch": 6.128246104674391, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 15339 + }, + { + "epoch": 6.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15340 + }, + { + "epoch": 6.12904514582501, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15341 + }, + { + "epoch": 6.129444666400319, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15342 + }, + { + "epoch": 6.129844186975629, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15343 + }, + { + "epoch": 6.130243707550939, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15344 + }, + { + "epoch": 6.130643228126249, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15345 + }, + { + "epoch": 6.131042748701558, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15346 + }, + { + "epoch": 6.131442269276867, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15347 + }, + { + "epoch": 6.131841789852177, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15348 + }, + { + "epoch": 6.132241310427487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15349 + }, + { + "epoch": 6.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15350 + }, + { + "epoch": 6.1330403515781065, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15351 + }, + { + "epoch": 6.133439872153416, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15352 + }, + { + "epoch": 6.133839392728725, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15353 + }, + { + "epoch": 6.134238913304035, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15354 + }, + { + "epoch": 6.134638433879345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15355 + }, + { + "epoch": 6.1350379544546545, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15356 + }, + { + "epoch": 6.135437475029964, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15357 + }, + { + "epoch": 6.135836995605274, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15358 + }, + { + "epoch": 6.136236516180583, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15359 + }, + { + "epoch": 6.136636036755893, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15360 + }, + { + "epoch": 6.1370355573312025, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15361 + }, + { + "epoch": 6.137435077906512, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15362 + }, + { + "epoch": 6.137834598481822, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15363 + }, + { + "epoch": 6.138234119057131, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15364 + }, + { + "epoch": 6.138633639632441, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 15365 + }, + { + "epoch": 6.139033160207751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15366 + }, + { + "epoch": 6.13943268078306, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15367 + }, + { + "epoch": 6.13983220135837, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15368 + }, + { + "epoch": 6.14023172193368, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15369 + }, + { + "epoch": 6.140631242508989, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15370 + }, + { + "epoch": 6.141030763084299, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15371 + }, + { + "epoch": 6.141430283659608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15372 + }, + { + "epoch": 6.141829804234918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15373 + }, + { + "epoch": 6.142229324810228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15374 + }, + { + "epoch": 6.142628845385538, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15375 + }, + { + "epoch": 6.143028365960847, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15376 + }, + { + "epoch": 6.1434278865361565, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15377 + }, + { + "epoch": 6.143827407111466, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15378 + }, + { + "epoch": 6.144226927686776, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15379 + }, + { + "epoch": 6.144626448262086, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15380 + }, + { + "epoch": 6.145025968837395, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15381 + }, + { + "epoch": 6.1454254894127045, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 15382 + }, + { + "epoch": 6.145825009988014, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15383 + }, + { + "epoch": 6.146224530563324, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15384 + }, + { + "epoch": 6.146624051138634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15385 + }, + { + "epoch": 6.147023571713944, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15386 + }, + { + "epoch": 6.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15387 + }, + { + "epoch": 6.147822612864562, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15388 + }, + { + "epoch": 6.148222133439872, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15389 + }, + { + "epoch": 6.148621654015182, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15390 + }, + { + "epoch": 6.149021174590492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15391 + }, + { + "epoch": 6.1494206951658015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15392 + }, + { + "epoch": 6.14982021574111, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15393 + }, + { + "epoch": 6.15021973631642, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15394 + }, + { + "epoch": 6.15061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15395 + }, + { + "epoch": 6.15101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15396 + }, + { + "epoch": 6.1514182980423495, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15397 + }, + { + "epoch": 6.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 15398 + }, + { + "epoch": 6.152217339192968, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15399 + }, + { + "epoch": 6.152616859768278, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15400 + }, + { + "epoch": 6.153016380343588, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15401 + }, + { + "epoch": 6.153415900918898, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15402 + }, + { + "epoch": 6.153815421494207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15403 + }, + { + "epoch": 6.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2963, + "step": 15404 + }, + { + "epoch": 6.154614462644826, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15405 + }, + { + "epoch": 6.155013983220136, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15406 + }, + { + "epoch": 6.155413503795446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15407 + }, + { + "epoch": 6.155813024370755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15408 + }, + { + "epoch": 6.156212544946065, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15409 + }, + { + "epoch": 6.156612065521374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15410 + }, + { + "epoch": 6.157011586096684, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15411 + }, + { + "epoch": 6.157411106671994, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15412 + }, + { + "epoch": 6.157810627247303, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15413 + }, + { + "epoch": 6.158210147822613, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15414 + }, + { + "epoch": 6.158609668397922, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15415 + }, + { + "epoch": 6.159009188973232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15416 + }, + { + "epoch": 6.159408709548542, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15417 + }, + { + "epoch": 6.1598082301238515, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15418 + }, + { + "epoch": 6.160207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15419 + }, + { + "epoch": 6.160607271274471, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.2983, + "step": 15420 + }, + { + "epoch": 6.16100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 15421 + }, + { + "epoch": 6.16140631242509, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15422 + }, + { + "epoch": 6.1618058330003995, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15423 + }, + { + "epoch": 6.162205353575709, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15424 + }, + { + "epoch": 6.162604874151019, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15425 + }, + { + "epoch": 6.163004394726328, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15426 + }, + { + "epoch": 6.163403915301638, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15427 + }, + { + "epoch": 6.163803435876948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15428 + }, + { + "epoch": 6.164202956452257, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15429 + }, + { + "epoch": 6.164602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15430 + }, + { + "epoch": 6.165001997602877, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15431 + }, + { + "epoch": 6.165401518178186, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15432 + }, + { + "epoch": 6.165801038753496, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15433 + }, + { + "epoch": 6.166200559328805, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15434 + }, + { + "epoch": 6.166600079904115, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15435 + }, + { + "epoch": 6.166999600479425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15436 + }, + { + "epoch": 6.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 15437 + }, + { + "epoch": 6.167798641630044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15438 + }, + { + "epoch": 6.168198162205353, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15439 + }, + { + "epoch": 6.168597682780663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15440 + }, + { + "epoch": 6.168997203355973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15441 + }, + { + "epoch": 6.169396723931283, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15442 + }, + { + "epoch": 6.169796244506592, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15443 + }, + { + "epoch": 6.1701957650819015, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 15444 + }, + { + "epoch": 6.170595285657211, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15445 + }, + { + "epoch": 6.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15446 + }, + { + "epoch": 6.171394326807831, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15447 + }, + { + "epoch": 6.171793847383141, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15448 + }, + { + "epoch": 6.1721933679584495, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15449 + }, + { + "epoch": 6.172592888533759, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15450 + }, + { + "epoch": 6.172992409109069, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 15451 + }, + { + "epoch": 6.173391929684379, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15452 + }, + { + "epoch": 6.173791450259689, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15453 + }, + { + "epoch": 6.174190970834998, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15454 + }, + { + "epoch": 6.174590491410307, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15455 + }, + { + "epoch": 6.174990011985617, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15456 + }, + { + "epoch": 6.175389532560927, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15457 + }, + { + "epoch": 6.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15458 + }, + { + "epoch": 6.1761885737115465, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15459 + }, + { + "epoch": 6.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15460 + }, + { + "epoch": 6.176987614862165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15461 + }, + { + "epoch": 6.177387135437475, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 15462 + }, + { + "epoch": 6.177786656012785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15463 + }, + { + "epoch": 6.1781861765880945, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15464 + }, + { + "epoch": 6.178585697163404, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15465 + }, + { + "epoch": 6.178985217738713, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15466 + }, + { + "epoch": 6.179384738314023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15467 + }, + { + "epoch": 6.179784258889333, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15468 + }, + { + "epoch": 6.180183779464643, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15469 + }, + { + "epoch": 6.180583300039952, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15470 + }, + { + "epoch": 6.180982820615261, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15471 + }, + { + "epoch": 6.181382341190571, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15472 + }, + { + "epoch": 6.181781861765881, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15473 + }, + { + "epoch": 6.182181382341191, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15474 + }, + { + "epoch": 6.1825809029165, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15475 + }, + { + "epoch": 6.18298042349181, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15476 + }, + { + "epoch": 6.183379944067119, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15477 + }, + { + "epoch": 6.183779464642429, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 15478 + }, + { + "epoch": 6.184178985217739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15479 + }, + { + "epoch": 6.1845785057930485, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15480 + }, + { + "epoch": 6.184978026368358, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15481 + }, + { + "epoch": 6.185377546943668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15482 + }, + { + "epoch": 6.185777067518977, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15483 + }, + { + "epoch": 6.186176588094287, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15484 + }, + { + "epoch": 6.1865761086695965, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15485 + }, + { + "epoch": 6.186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15486 + }, + { + "epoch": 6.187375149820216, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15487 + }, + { + "epoch": 6.187774670395525, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15488 + }, + { + "epoch": 6.188174190970835, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15489 + }, + { + "epoch": 6.1885737115461446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15490 + }, + { + "epoch": 6.188973232121454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15491 + }, + { + "epoch": 6.189372752696764, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15492 + }, + { + "epoch": 6.189772273272074, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15493 + }, + { + "epoch": 6.190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15494 + }, + { + "epoch": 6.190571314422693, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15495 + }, + { + "epoch": 6.190970834998002, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15496 + }, + { + "epoch": 6.191370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 15497 + }, + { + "epoch": 6.191769876148622, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15498 + }, + { + "epoch": 6.192169396723932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15499 + }, + { + "epoch": 6.192568917299241, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15500 + }, + { + "epoch": 6.19296843787455, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15501 + }, + { + "epoch": 6.19336795844986, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15502 + }, + { + "epoch": 6.19376747902517, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15503 + }, + { + "epoch": 6.19416699960048, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15504 + }, + { + "epoch": 6.194566520175789, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15505 + }, + { + "epoch": 6.1949660407510985, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15506 + }, + { + "epoch": 6.195365561326408, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15507 + }, + { + "epoch": 6.195765081901718, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15508 + }, + { + "epoch": 6.196164602477028, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15509 + }, + { + "epoch": 6.196564123052338, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15510 + }, + { + "epoch": 6.1969636436276465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15511 + }, + { + "epoch": 6.197363164202956, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15512 + }, + { + "epoch": 6.197762684778266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15513 + }, + { + "epoch": 6.198162205353576, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15514 + }, + { + "epoch": 6.198561725928886, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15515 + }, + { + "epoch": 6.1989612465041946, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15516 + }, + { + "epoch": 6.199360767079504, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15517 + }, + { + "epoch": 6.199760287654814, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15518 + }, + { + "epoch": 6.200159808230124, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15519 + }, + { + "epoch": 6.200559328805434, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15520 + }, + { + "epoch": 6.2009588493807435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 15521 + }, + { + "epoch": 6.201358369956052, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15522 + }, + { + "epoch": 6.201757890531362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15523 + }, + { + "epoch": 6.202157411106672, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15524 + }, + { + "epoch": 6.202556931681982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 15525 + }, + { + "epoch": 6.2029564522572915, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15526 + }, + { + "epoch": 6.2033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15527 + }, + { + "epoch": 6.20375549340791, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15528 + }, + { + "epoch": 6.20415501398322, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15529 + }, + { + "epoch": 6.20455453455853, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15530 + }, + { + "epoch": 6.20495405513384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15531 + }, + { + "epoch": 6.205353575709149, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15532 + }, + { + "epoch": 6.205753096284458, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15533 + }, + { + "epoch": 6.206152616859768, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15534 + }, + { + "epoch": 6.206552137435078, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15535 + }, + { + "epoch": 6.206951658010388, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15536 + }, + { + "epoch": 6.207351178585697, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15537 + }, + { + "epoch": 6.207750699161007, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15538 + }, + { + "epoch": 6.208150219736316, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15539 + }, + { + "epoch": 6.208549740311626, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15540 + }, + { + "epoch": 6.208949260886936, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15541 + }, + { + "epoch": 6.2093487814622454, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15542 + }, + { + "epoch": 6.209748302037555, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15543 + }, + { + "epoch": 6.210147822612864, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15544 + }, + { + "epoch": 6.210547343188174, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15545 + }, + { + "epoch": 6.210946863763484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15546 + }, + { + "epoch": 6.2113463843387935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15547 + }, + { + "epoch": 6.211745904914103, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15548 + }, + { + "epoch": 6.212145425489413, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15549 + }, + { + "epoch": 6.212544946064722, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15550 + }, + { + "epoch": 6.212944466640032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15551 + }, + { + "epoch": 6.2133439872153415, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15552 + }, + { + "epoch": 6.213743507790651, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15553 + }, + { + "epoch": 6.214143028365961, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15554 + }, + { + "epoch": 6.214542548941271, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15555 + }, + { + "epoch": 6.21494206951658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15556 + }, + { + "epoch": 6.21534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 15557 + }, + { + "epoch": 6.215741110667199, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 15558 + }, + { + "epoch": 6.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 15559 + }, + { + "epoch": 6.216540151817819, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15560 + }, + { + "epoch": 6.216939672393128, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15561 + }, + { + "epoch": 6.217339192968438, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15562 + }, + { + "epoch": 6.217738713543747, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15563 + }, + { + "epoch": 6.218138234119057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15564 + }, + { + "epoch": 6.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15565 + }, + { + "epoch": 6.218937275269677, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15566 + }, + { + "epoch": 6.219336795844986, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15567 + }, + { + "epoch": 6.2197363164202955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15568 + }, + { + "epoch": 6.220135836995605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15569 + }, + { + "epoch": 6.220535357570915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15570 + }, + { + "epoch": 6.220934878146225, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15571 + }, + { + "epoch": 6.221334398721535, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15572 + }, + { + "epoch": 6.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15573 + }, + { + "epoch": 6.222133439872153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15574 + }, + { + "epoch": 6.222532960447463, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 15575 + }, + { + "epoch": 6.222932481022773, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15576 + }, + { + "epoch": 6.223332001598083, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15577 + }, + { + "epoch": 6.2237315221733915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15578 + }, + { + "epoch": 6.224131042748701, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15579 + }, + { + "epoch": 6.224530563324011, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15580 + }, + { + "epoch": 6.224930083899321, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15581 + }, + { + "epoch": 6.225329604474631, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15582 + }, + { + "epoch": 6.2257291250499405, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15583 + }, + { + "epoch": 6.226128645625249, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15584 + }, + { + "epoch": 6.226528166200559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15585 + }, + { + "epoch": 6.226927686775869, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15586 + }, + { + "epoch": 6.227327207351179, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15587 + }, + { + "epoch": 6.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15588 + }, + { + "epoch": 6.228126248501797, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15589 + }, + { + "epoch": 6.228525769077107, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15590 + }, + { + "epoch": 6.228925289652417, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15591 + }, + { + "epoch": 6.229324810227727, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 15592 + }, + { + "epoch": 6.229724330803037, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15593 + }, + { + "epoch": 6.230123851378346, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15594 + }, + { + "epoch": 6.230523371953655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15595 + }, + { + "epoch": 6.230922892528965, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15596 + }, + { + "epoch": 6.231322413104275, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15597 + }, + { + "epoch": 6.231721933679585, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15598 + }, + { + "epoch": 6.232121454254894, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15599 + }, + { + "epoch": 6.232520974830204, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15600 + }, + { + "epoch": 6.232920495405513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15601 + }, + { + "epoch": 6.233320015980823, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15602 + }, + { + "epoch": 6.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15603 + }, + { + "epoch": 6.234119057131442, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15604 + }, + { + "epoch": 6.234518577706752, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 15605 + }, + { + "epoch": 6.234918098282061, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15606 + }, + { + "epoch": 6.235317618857371, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 15607 + }, + { + "epoch": 6.235717139432681, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15608 + }, + { + "epoch": 6.2361166600079905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15609 + }, + { + "epoch": 6.2365161805833, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15610 + }, + { + "epoch": 6.23691570115861, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15611 + }, + { + "epoch": 6.237315221733919, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15612 + }, + { + "epoch": 6.237714742309229, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15613 + }, + { + "epoch": 6.2381142628845385, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 15614 + }, + { + "epoch": 6.238513783459848, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15615 + }, + { + "epoch": 6.238913304035158, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15616 + }, + { + "epoch": 6.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15617 + }, + { + "epoch": 6.239712345185777, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15618 + }, + { + "epoch": 6.240111865761087, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15619 + }, + { + "epoch": 6.240511386336396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15620 + }, + { + "epoch": 6.240910906911706, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15621 + }, + { + "epoch": 6.241310427487016, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15622 + }, + { + "epoch": 6.241709948062325, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15623 + }, + { + "epoch": 6.242109468637635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15624 + }, + { + "epoch": 6.242508989212944, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15625 + }, + { + "epoch": 6.242908509788254, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15626 + }, + { + "epoch": 6.243308030363564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15627 + }, + { + "epoch": 6.243707550938874, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15628 + }, + { + "epoch": 6.244107071514183, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15629 + }, + { + "epoch": 6.244506592089492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15630 + }, + { + "epoch": 6.244906112664802, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15631 + }, + { + "epoch": 6.245305633240112, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15632 + }, + { + "epoch": 6.245705153815422, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15633 + }, + { + "epoch": 6.246104674390731, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15634 + }, + { + "epoch": 6.2465041949660405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 15635 + }, + { + "epoch": 6.24690371554135, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15636 + }, + { + "epoch": 6.24730323611666, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15637 + }, + { + "epoch": 6.24770275669197, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15638 + }, + { + "epoch": 6.24810227726728, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15639 + }, + { + "epoch": 6.2485017978425885, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15640 + }, + { + "epoch": 6.248901318417898, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15641 + }, + { + "epoch": 6.249300838993208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15642 + }, + { + "epoch": 6.249700359568518, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15643 + }, + { + "epoch": 6.250099880143828, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15644 + }, + { + "epoch": 6.2504994007191375, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15645 + }, + { + "epoch": 6.250898921294446, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15646 + }, + { + "epoch": 6.251298441869756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15647 + }, + { + "epoch": 6.251697962445066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15648 + }, + { + "epoch": 6.252097483020376, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15649 + }, + { + "epoch": 6.2524970035956855, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15650 + }, + { + "epoch": 6.252896524170994, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15651 + }, + { + "epoch": 6.253296044746304, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15652 + }, + { + "epoch": 6.253695565321614, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15653 + }, + { + "epoch": 6.254095085896924, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15654 + }, + { + "epoch": 6.2544946064722335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15655 + }, + { + "epoch": 6.254894127047543, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15656 + }, + { + "epoch": 6.255293647622852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15657 + }, + { + "epoch": 6.255693168198162, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15658 + }, + { + "epoch": 6.256092688773472, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15659 + }, + { + "epoch": 6.256492209348782, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15660 + }, + { + "epoch": 6.256891729924091, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15661 + }, + { + "epoch": 6.257291250499401, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15662 + }, + { + "epoch": 6.25769077107471, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15663 + }, + { + "epoch": 6.25809029165002, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15664 + }, + { + "epoch": 6.25848981222533, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15665 + }, + { + "epoch": 6.258889332800639, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15666 + }, + { + "epoch": 6.259288853375949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 15667 + }, + { + "epoch": 6.259688373951258, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15668 + }, + { + "epoch": 6.260087894526568, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15669 + }, + { + "epoch": 6.260487415101878, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15670 + }, + { + "epoch": 6.2608869356771875, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 15671 + }, + { + "epoch": 6.261286456252497, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15672 + }, + { + "epoch": 6.261685976827807, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15673 + }, + { + "epoch": 6.262085497403116, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15674 + }, + { + "epoch": 6.262485017978426, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15675 + }, + { + "epoch": 6.2628845385537355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15676 + }, + { + "epoch": 6.263284059129045, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15677 + }, + { + "epoch": 6.263683579704355, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 15678 + }, + { + "epoch": 6.264083100279665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 15679 + }, + { + "epoch": 6.264482620854974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15680 + }, + { + "epoch": 6.2648821414302835, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15681 + }, + { + "epoch": 6.265281662005593, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15682 + }, + { + "epoch": 6.265681182580903, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15683 + }, + { + "epoch": 6.266080703156213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15684 + }, + { + "epoch": 6.266480223731522, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15685 + }, + { + "epoch": 6.266879744306832, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15686 + }, + { + "epoch": 6.267279264882141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15687 + }, + { + "epoch": 6.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15688 + }, + { + "epoch": 6.268078306032761, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15689 + }, + { + "epoch": 6.26847782660807, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15690 + }, + { + "epoch": 6.26887734718338, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15691 + }, + { + "epoch": 6.269276867758689, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15692 + }, + { + "epoch": 6.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15693 + }, + { + "epoch": 6.270075908909309, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15694 + }, + { + "epoch": 6.270475429484619, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15695 + }, + { + "epoch": 6.270874950059928, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15696 + }, + { + "epoch": 6.2712744706352375, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15697 + }, + { + "epoch": 6.271673991210547, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15698 + }, + { + "epoch": 6.272073511785857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15699 + }, + { + "epoch": 6.272473032361167, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15700 + }, + { + "epoch": 6.272872552936477, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15701 + }, + { + "epoch": 6.2732720735117855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15702 + }, + { + "epoch": 6.273671594087095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 15703 + }, + { + "epoch": 6.274071114662405, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15704 + }, + { + "epoch": 6.274470635237715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15705 + }, + { + "epoch": 6.274870155813025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15706 + }, + { + "epoch": 6.2752696763883336, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 15707 + }, + { + "epoch": 6.275669196963643, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15708 + }, + { + "epoch": 6.276068717538953, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15709 + }, + { + "epoch": 6.276468238114263, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15710 + }, + { + "epoch": 6.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15711 + }, + { + "epoch": 6.2772672792648825, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15712 + }, + { + "epoch": 6.277666799840191, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15713 + }, + { + "epoch": 6.278066320415501, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15714 + }, + { + "epoch": 6.278465840990811, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15715 + }, + { + "epoch": 6.278865361566121, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15716 + }, + { + "epoch": 6.2792648821414305, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15717 + }, + { + "epoch": 6.27966440271674, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15718 + }, + { + "epoch": 6.280063923292049, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15719 + }, + { + "epoch": 6.280463443867359, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15720 + }, + { + "epoch": 6.280862964442669, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15721 + }, + { + "epoch": 6.281262485017979, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15722 + }, + { + "epoch": 6.281662005593288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15723 + }, + { + "epoch": 6.282061526168597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 15724 + }, + { + "epoch": 6.282461046743907, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15725 + }, + { + "epoch": 6.282860567319217, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15726 + }, + { + "epoch": 6.283260087894527, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15727 + }, + { + "epoch": 6.283659608469836, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15728 + }, + { + "epoch": 6.284059129045146, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15729 + }, + { + "epoch": 6.284458649620455, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15730 + }, + { + "epoch": 6.284858170195765, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 15731 + }, + { + "epoch": 6.285257690771075, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 15732 + }, + { + "epoch": 6.285657211346384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15733 + }, + { + "epoch": 6.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15734 + }, + { + "epoch": 6.286456252497004, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15735 + }, + { + "epoch": 6.286855773072313, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15736 + }, + { + "epoch": 6.287255293647623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15737 + }, + { + "epoch": 6.2876548142229325, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 15738 + }, + { + "epoch": 6.288054334798242, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15739 + }, + { + "epoch": 6.288453855373552, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15740 + }, + { + "epoch": 6.288853375948861, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15741 + }, + { + "epoch": 6.289252896524171, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15742 + }, + { + "epoch": 6.2896524170994805, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15743 + }, + { + "epoch": 6.29005193767479, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15744 + }, + { + "epoch": 6.2904514582501, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15745 + }, + { + "epoch": 6.29085097882541, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15746 + }, + { + "epoch": 6.291250499400719, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15747 + }, + { + "epoch": 6.291650019976029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15748 + }, + { + "epoch": 6.292049540551338, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15749 + }, + { + "epoch": 6.292449061126648, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15750 + }, + { + "epoch": 6.292848581701958, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15751 + }, + { + "epoch": 6.293248102277268, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15752 + }, + { + "epoch": 6.293647622852577, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15753 + }, + { + "epoch": 6.294047143427886, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15754 + }, + { + "epoch": 6.294446664003196, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15755 + }, + { + "epoch": 6.294846184578506, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15756 + }, + { + "epoch": 6.295245705153816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15757 + }, + { + "epoch": 6.295645225729125, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 15758 + }, + { + "epoch": 6.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15759 + }, + { + "epoch": 6.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15760 + }, + { + "epoch": 6.296843787455054, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15761 + }, + { + "epoch": 6.297243308030364, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15762 + }, + { + "epoch": 6.297642828605674, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15763 + }, + { + "epoch": 6.2980423491809825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15764 + }, + { + "epoch": 6.298441869756292, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15765 + }, + { + "epoch": 6.298841390331602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15766 + }, + { + "epoch": 6.299240910906912, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15767 + }, + { + "epoch": 6.299640431482222, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 15768 + }, + { + "epoch": 6.300039952057531, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15769 + }, + { + "epoch": 6.30043947263284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15770 + }, + { + "epoch": 6.30083899320815, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15771 + }, + { + "epoch": 6.30123851378346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15772 + }, + { + "epoch": 6.30163803435877, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15773 + }, + { + "epoch": 6.3020375549340795, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15774 + }, + { + "epoch": 6.302437075509388, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15775 + }, + { + "epoch": 6.302836596084698, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15776 + }, + { + "epoch": 6.303236116660008, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15777 + }, + { + "epoch": 6.303635637235318, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15778 + }, + { + "epoch": 6.3040351578106275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15779 + }, + { + "epoch": 6.304434678385936, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15780 + }, + { + "epoch": 6.304834198961246, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15781 + }, + { + "epoch": 6.305233719536556, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15782 + }, + { + "epoch": 6.305633240111866, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15783 + }, + { + "epoch": 6.3060327606871756, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15784 + }, + { + "epoch": 6.306432281262485, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 15785 + }, + { + "epoch": 6.306831801837794, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15786 + }, + { + "epoch": 6.307231322413104, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15787 + }, + { + "epoch": 6.307630842988414, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15788 + }, + { + "epoch": 6.308030363563724, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15789 + }, + { + "epoch": 6.308429884139033, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15790 + }, + { + "epoch": 6.308829404714343, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15791 + }, + { + "epoch": 6.309228925289652, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15792 + }, + { + "epoch": 6.309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15793 + }, + { + "epoch": 6.310027966440272, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15794 + }, + { + "epoch": 6.310427487015581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15795 + }, + { + "epoch": 6.310827007590891, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15796 + }, + { + "epoch": 6.3112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15797 + }, + { + "epoch": 6.31162604874151, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15798 + }, + { + "epoch": 6.31202556931682, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15799 + }, + { + "epoch": 6.3124250898921295, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 15800 + }, + { + "epoch": 6.312824610467439, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15801 + }, + { + "epoch": 6.313224131042749, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 15802 + }, + { + "epoch": 6.313623651618058, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15803 + }, + { + "epoch": 6.314023172193368, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15804 + }, + { + "epoch": 6.3144226927686775, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15805 + }, + { + "epoch": 6.314822213343987, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15806 + }, + { + "epoch": 6.315221733919297, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15807 + }, + { + "epoch": 6.315621254494607, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15808 + }, + { + "epoch": 6.316020775069916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15809 + }, + { + "epoch": 6.316420295645226, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15810 + }, + { + "epoch": 6.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15811 + }, + { + "epoch": 6.317219336795845, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 15812 + }, + { + "epoch": 6.317618857371155, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15813 + }, + { + "epoch": 6.318018377946464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15814 + }, + { + "epoch": 6.318417898521774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 15815 + }, + { + "epoch": 6.318817419097083, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15816 + }, + { + "epoch": 6.319216939672393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15817 + }, + { + "epoch": 6.319616460247703, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15818 + }, + { + "epoch": 6.320015980823013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15819 + }, + { + "epoch": 6.320415501398322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15820 + }, + { + "epoch": 6.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15821 + }, + { + "epoch": 6.321214542548941, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15822 + }, + { + "epoch": 6.321614063124251, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15823 + }, + { + "epoch": 6.322013583699561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15824 + }, + { + "epoch": 6.322413104274871, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15825 + }, + { + "epoch": 6.3228126248501795, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15826 + }, + { + "epoch": 6.323212145425489, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15827 + }, + { + "epoch": 6.323611666000799, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15828 + }, + { + "epoch": 6.324011186576109, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15829 + }, + { + "epoch": 6.324410707151419, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15830 + }, + { + "epoch": 6.3248102277267275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15831 + }, + { + "epoch": 6.325209748302037, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15832 + }, + { + "epoch": 6.325609268877347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15833 + }, + { + "epoch": 6.326008789452657, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15834 + }, + { + "epoch": 6.326408310027967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15835 + }, + { + "epoch": 6.3268078306032765, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15836 + }, + { + "epoch": 6.327207351178585, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15837 + }, + { + "epoch": 6.327606871753895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15838 + }, + { + "epoch": 6.328006392329205, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15839 + }, + { + "epoch": 6.328405912904515, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15840 + }, + { + "epoch": 6.3288054334798245, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15841 + }, + { + "epoch": 6.329204954055134, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15842 + }, + { + "epoch": 6.329604474630443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15843 + }, + { + "epoch": 6.330003995205753, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 15844 + }, + { + "epoch": 6.330403515781063, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15845 + }, + { + "epoch": 6.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 15846 + }, + { + "epoch": 6.331202556931682, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15847 + }, + { + "epoch": 6.331602077506991, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15848 + }, + { + "epoch": 6.332001598082301, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15849 + }, + { + "epoch": 6.332401118657611, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15850 + }, + { + "epoch": 6.332800639232921, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15851 + }, + { + "epoch": 6.33320015980823, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15852 + }, + { + "epoch": 6.33359968038354, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15853 + }, + { + "epoch": 6.333999200958849, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15854 + }, + { + "epoch": 6.334398721534159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15855 + }, + { + "epoch": 6.334798242109469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15856 + }, + { + "epoch": 6.335197762684778, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15857 + }, + { + "epoch": 6.335597283260088, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15858 + }, + { + "epoch": 6.335996803835398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15859 + }, + { + "epoch": 6.336396324410707, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15860 + }, + { + "epoch": 6.336795844986017, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15861 + }, + { + "epoch": 6.3371953655613265, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15862 + }, + { + "epoch": 6.337594886136636, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15863 + }, + { + "epoch": 6.337994406711946, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15864 + }, + { + "epoch": 6.338393927287255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15865 + }, + { + "epoch": 6.338793447862565, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15866 + }, + { + "epoch": 6.3391929684378745, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15867 + }, + { + "epoch": 6.339592489013184, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15868 + }, + { + "epoch": 6.339992009588494, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15869 + }, + { + "epoch": 6.340391530163803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15870 + }, + { + "epoch": 6.340791050739113, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15871 + }, + { + "epoch": 6.3411905713144225, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15872 + }, + { + "epoch": 6.341590091889732, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15873 + }, + { + "epoch": 6.341989612465042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15874 + }, + { + "epoch": 6.342389133040352, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15875 + }, + { + "epoch": 6.342788653615661, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15876 + }, + { + "epoch": 6.343188174190971, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15877 + }, + { + "epoch": 6.34358769476628, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15878 + }, + { + "epoch": 6.34398721534159, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15879 + }, + { + "epoch": 6.3443867359169, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15880 + }, + { + "epoch": 6.34478625649221, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15881 + }, + { + "epoch": 6.345185777067519, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15882 + }, + { + "epoch": 6.345585297642828, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15883 + }, + { + "epoch": 6.345984818218138, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15884 + }, + { + "epoch": 6.346384338793448, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15885 + }, + { + "epoch": 6.346783859368758, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15886 + }, + { + "epoch": 6.347183379944067, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15887 + }, + { + "epoch": 6.3475829005193765, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15888 + }, + { + "epoch": 6.347982421094686, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15889 + }, + { + "epoch": 6.348381941669996, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15890 + }, + { + "epoch": 6.348781462245306, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 15891 + }, + { + "epoch": 6.349180982820616, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15892 + }, + { + "epoch": 6.3495805033959245, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15893 + }, + { + "epoch": 6.349980023971234, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15894 + }, + { + "epoch": 6.350379544546544, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15895 + }, + { + "epoch": 6.350779065121854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15896 + }, + { + "epoch": 6.351178585697164, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15897 + }, + { + "epoch": 6.351578106272473, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15898 + }, + { + "epoch": 6.351977626847782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15899 + }, + { + "epoch": 6.352377147423092, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15900 + }, + { + "epoch": 6.352776667998402, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15901 + }, + { + "epoch": 6.353176188573712, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 15902 + }, + { + "epoch": 6.3535757091490215, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15903 + }, + { + "epoch": 6.35397522972433, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 15904 + }, + { + "epoch": 6.35437475029964, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15905 + }, + { + "epoch": 6.35477427087495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15906 + }, + { + "epoch": 6.35517379145026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15907 + }, + { + "epoch": 6.3555733120255695, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15908 + }, + { + "epoch": 6.355972832600879, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15909 + }, + { + "epoch": 6.356372353176188, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15910 + }, + { + "epoch": 6.356771873751498, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 15911 + }, + { + "epoch": 6.357171394326808, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15912 + }, + { + "epoch": 6.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15913 + }, + { + "epoch": 6.357970435477427, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15914 + }, + { + "epoch": 6.358369956052737, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15915 + }, + { + "epoch": 6.358769476628046, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15916 + }, + { + "epoch": 6.359168997203356, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15917 + }, + { + "epoch": 6.359568517778666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15918 + }, + { + "epoch": 6.359968038353975, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15919 + }, + { + "epoch": 6.360367558929285, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 15920 + }, + { + "epoch": 6.360767079504594, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15921 + }, + { + "epoch": 6.361166600079904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15922 + }, + { + "epoch": 6.361566120655214, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15923 + }, + { + "epoch": 6.361965641230523, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15924 + }, + { + "epoch": 6.362365161805833, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15925 + }, + { + "epoch": 6.362764682381143, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15926 + }, + { + "epoch": 6.363164202956452, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15927 + }, + { + "epoch": 6.363563723531762, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15928 + }, + { + "epoch": 6.3639632441070715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15929 + }, + { + "epoch": 6.364362764682381, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15930 + }, + { + "epoch": 6.364762285257691, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15931 + }, + { + "epoch": 6.365161805833001, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15932 + }, + { + "epoch": 6.36556132640831, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15933 + }, + { + "epoch": 6.3659608469836195, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15934 + }, + { + "epoch": 6.366360367558929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15935 + }, + { + "epoch": 6.366759888134239, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.297, + "step": 15936 + }, + { + "epoch": 6.367159408709549, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15937 + }, + { + "epoch": 6.367558929284858, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15938 + }, + { + "epoch": 6.367958449860168, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15939 + }, + { + "epoch": 6.368357970435477, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15940 + }, + { + "epoch": 6.368757491010787, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15941 + }, + { + "epoch": 6.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 15942 + }, + { + "epoch": 6.369556532161407, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15943 + }, + { + "epoch": 6.369956052736716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15944 + }, + { + "epoch": 6.370355573312025, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15945 + }, + { + "epoch": 6.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15946 + }, + { + "epoch": 6.371154614462645, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15947 + }, + { + "epoch": 6.371554135037955, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15948 + }, + { + "epoch": 6.3719536556132645, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15949 + }, + { + "epoch": 6.372353176188573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15950 + }, + { + "epoch": 6.372752696763883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15951 + }, + { + "epoch": 6.373152217339193, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15952 + }, + { + "epoch": 6.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 15953 + }, + { + "epoch": 6.373951258489813, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15954 + }, + { + "epoch": 6.3743507790651215, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15955 + }, + { + "epoch": 6.374750299640431, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15956 + }, + { + "epoch": 6.375149820215741, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15957 + }, + { + "epoch": 6.375549340791051, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15958 + }, + { + "epoch": 6.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15959 + }, + { + "epoch": 6.3763483819416695, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15960 + }, + { + "epoch": 6.376747902516979, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 15961 + }, + { + "epoch": 6.377147423092289, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15962 + }, + { + "epoch": 6.377546943667599, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15963 + }, + { + "epoch": 6.377946464242909, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15964 + }, + { + "epoch": 6.3783459848182185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15965 + }, + { + "epoch": 6.378745505393527, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15966 + }, + { + "epoch": 6.379145025968837, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 15967 + }, + { + "epoch": 6.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15968 + }, + { + "epoch": 6.379944067119457, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15969 + }, + { + "epoch": 6.3803435876947665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15970 + }, + { + "epoch": 6.380743108270076, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15971 + }, + { + "epoch": 6.381142628845385, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15972 + }, + { + "epoch": 6.381542149420695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15973 + }, + { + "epoch": 6.381941669996005, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15974 + }, + { + "epoch": 6.3823411905713145, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15975 + }, + { + "epoch": 6.382740711146624, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15976 + }, + { + "epoch": 6.383140231721933, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15977 + }, + { + "epoch": 6.383539752297243, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15978 + }, + { + "epoch": 6.383939272872553, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15979 + }, + { + "epoch": 6.384338793447863, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15980 + }, + { + "epoch": 6.384738314023172, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15981 + }, + { + "epoch": 6.385137834598482, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15982 + }, + { + "epoch": 6.385537355173791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15983 + }, + { + "epoch": 6.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15984 + }, + { + "epoch": 6.386336396324411, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15985 + }, + { + "epoch": 6.38673591689972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15986 + }, + { + "epoch": 6.38713543747503, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15987 + }, + { + "epoch": 6.38753495805034, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 15988 + }, + { + "epoch": 6.387934478625649, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15989 + }, + { + "epoch": 6.388333999200959, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15990 + }, + { + "epoch": 6.3887335197762685, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15991 + }, + { + "epoch": 6.389133040351578, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15992 + }, + { + "epoch": 6.389532560926888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15993 + }, + { + "epoch": 6.389932081502197, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15994 + }, + { + "epoch": 6.390331602077507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15995 + }, + { + "epoch": 6.3907311226528165, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15996 + }, + { + "epoch": 6.391130643228126, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15997 + }, + { + "epoch": 6.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15998 + }, + { + "epoch": 6.391929684378746, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15999 + }, + { + "epoch": 6.392329204954055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16000 + }, + { + "epoch": 6.3927287255293646, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16001 + }, + { + "epoch": 6.393128246104674, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 16002 + }, + { + "epoch": 6.393527766679984, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16003 + }, + { + "epoch": 6.393927287255294, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16004 + }, + { + "epoch": 6.394326807830604, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16005 + }, + { + "epoch": 6.394726328405913, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16006 + }, + { + "epoch": 6.395125848981222, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16007 + }, + { + "epoch": 6.395525369556532, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16008 + }, + { + "epoch": 6.395924890131842, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16009 + }, + { + "epoch": 6.396324410707152, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 16010 + }, + { + "epoch": 6.396723931282461, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16011 + }, + { + "epoch": 6.39712345185777, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16012 + }, + { + "epoch": 6.39752297243308, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16013 + }, + { + "epoch": 6.39792249300839, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16014 + }, + { + "epoch": 6.3983220135837, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16015 + }, + { + "epoch": 6.39872153415901, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16016 + }, + { + "epoch": 6.3991210547343185, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16017 + }, + { + "epoch": 6.399520575309628, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16018 + }, + { + "epoch": 6.399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16019 + }, + { + "epoch": 6.400319616460248, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16020 + }, + { + "epoch": 6.400719137035558, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16021 + }, + { + "epoch": 6.401118657610867, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16022 + }, + { + "epoch": 6.401518178186176, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16023 + }, + { + "epoch": 6.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16024 + }, + { + "epoch": 6.402317219336796, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16025 + }, + { + "epoch": 6.402716739912106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16026 + }, + { + "epoch": 6.4031162604874154, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16027 + }, + { + "epoch": 6.403515781062724, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16028 + }, + { + "epoch": 6.403915301638034, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16029 + }, + { + "epoch": 6.404314822213344, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16030 + }, + { + "epoch": 6.404714342788654, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16031 + }, + { + "epoch": 6.4051138633639635, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16032 + }, + { + "epoch": 6.405513383939273, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16033 + }, + { + "epoch": 6.405912904514582, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16034 + }, + { + "epoch": 6.406312425089892, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16035 + }, + { + "epoch": 6.406711945665202, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16036 + }, + { + "epoch": 6.4071114662405115, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16037 + }, + { + "epoch": 6.407510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16038 + }, + { + "epoch": 6.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16039 + }, + { + "epoch": 6.40831002796644, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16040 + }, + { + "epoch": 6.40870954854175, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16041 + }, + { + "epoch": 6.40910906911706, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16042 + }, + { + "epoch": 6.409508589692369, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16043 + }, + { + "epoch": 6.409908110267679, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16044 + }, + { + "epoch": 6.410307630842988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16045 + }, + { + "epoch": 6.410707151418298, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16046 + }, + { + "epoch": 6.411106671993608, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16047 + }, + { + "epoch": 6.411506192568917, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16048 + }, + { + "epoch": 6.411905713144227, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16049 + }, + { + "epoch": 6.412305233719536, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16050 + }, + { + "epoch": 6.412704754294846, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16051 + }, + { + "epoch": 6.413104274870156, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16052 + }, + { + "epoch": 6.4135037954454654, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16053 + }, + { + "epoch": 6.413903316020775, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16054 + }, + { + "epoch": 6.414302836596085, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16055 + }, + { + "epoch": 6.414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16056 + }, + { + "epoch": 6.415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16057 + }, + { + "epoch": 6.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16058 + }, + { + "epoch": 6.415900918897323, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16059 + }, + { + "epoch": 6.416300439472633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16060 + }, + { + "epoch": 6.416699960047943, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16061 + }, + { + "epoch": 6.417099480623252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 16062 + }, + { + "epoch": 6.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16063 + }, + { + "epoch": 6.417898521773871, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 16064 + }, + { + "epoch": 6.418298042349181, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16065 + }, + { + "epoch": 6.418697562924491, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16066 + }, + { + "epoch": 6.4190970834998, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16067 + }, + { + "epoch": 6.41949660407511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16068 + }, + { + "epoch": 6.419896124650419, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16069 + }, + { + "epoch": 6.420295645225729, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16070 + }, + { + "epoch": 6.420695165801039, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16071 + }, + { + "epoch": 6.421094686376349, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16072 + }, + { + "epoch": 6.421494206951658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16073 + }, + { + "epoch": 6.421893727526967, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16074 + }, + { + "epoch": 6.422293248102277, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16075 + }, + { + "epoch": 6.422692768677587, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16076 + }, + { + "epoch": 6.423092289252897, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16077 + }, + { + "epoch": 6.4234918098282066, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16078 + }, + { + "epoch": 6.4238913304035155, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16079 + }, + { + "epoch": 6.424290850978825, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16080 + }, + { + "epoch": 6.424690371554135, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16081 + }, + { + "epoch": 6.425089892129445, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16082 + }, + { + "epoch": 6.425489412704755, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16083 + }, + { + "epoch": 6.4258889332800635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16084 + }, + { + "epoch": 6.426288453855373, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16085 + }, + { + "epoch": 6.426687974430683, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16086 + }, + { + "epoch": 6.427087495005993, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16087 + }, + { + "epoch": 6.427487015581303, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16088 + }, + { + "epoch": 6.427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16089 + }, + { + "epoch": 6.428286056731921, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16090 + }, + { + "epoch": 6.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16091 + }, + { + "epoch": 6.429085097882541, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16092 + }, + { + "epoch": 6.429484618457851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16093 + }, + { + "epoch": 6.4298841390331605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16094 + }, + { + "epoch": 6.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16095 + }, + { + "epoch": 6.430683180183779, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16096 + }, + { + "epoch": 6.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16097 + }, + { + "epoch": 6.431482221334399, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16098 + }, + { + "epoch": 6.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16099 + }, + { + "epoch": 6.432281262485018, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16100 + }, + { + "epoch": 6.432680783060327, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16101 + }, + { + "epoch": 6.433080303635637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16102 + }, + { + "epoch": 6.433479824210947, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 16103 + }, + { + "epoch": 6.433879344786257, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16104 + }, + { + "epoch": 6.434278865361566, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16105 + }, + { + "epoch": 6.434678385936876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16106 + }, + { + "epoch": 6.435077906512185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16107 + }, + { + "epoch": 6.435477427087495, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16108 + }, + { + "epoch": 6.435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16109 + }, + { + "epoch": 6.436276468238114, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16110 + }, + { + "epoch": 6.436675988813424, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16111 + }, + { + "epoch": 6.437075509388734, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16112 + }, + { + "epoch": 6.437475029964043, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16113 + }, + { + "epoch": 6.437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16114 + }, + { + "epoch": 6.438274071114662, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 16115 + }, + { + "epoch": 6.438673591689972, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16116 + }, + { + "epoch": 6.439073112265282, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 16117 + }, + { + "epoch": 6.439472632840591, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16118 + }, + { + "epoch": 6.439872153415901, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16119 + }, + { + "epoch": 6.4402716739912105, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16120 + }, + { + "epoch": 6.44067119456652, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16121 + }, + { + "epoch": 6.44107071514183, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16122 + }, + { + "epoch": 6.44147023571714, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16123 + }, + { + "epoch": 6.441869756292449, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16124 + }, + { + "epoch": 6.4422692768677585, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16125 + }, + { + "epoch": 6.442668797443068, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16126 + }, + { + "epoch": 6.443068318018378, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16127 + }, + { + "epoch": 6.443467838593688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16128 + }, + { + "epoch": 6.443867359168998, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16129 + }, + { + "epoch": 6.444266879744307, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16130 + }, + { + "epoch": 6.444666400319616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16131 + }, + { + "epoch": 6.445065920894926, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16132 + }, + { + "epoch": 6.445465441470236, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16133 + }, + { + "epoch": 6.445864962045546, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16134 + }, + { + "epoch": 6.446264482620855, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 16135 + }, + { + "epoch": 6.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16136 + }, + { + "epoch": 6.447063523771474, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16137 + }, + { + "epoch": 6.447463044346784, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16138 + }, + { + "epoch": 6.447862564922094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16139 + }, + { + "epoch": 6.448262085497403, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16140 + }, + { + "epoch": 6.448661606072712, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16141 + }, + { + "epoch": 6.449061126648022, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16142 + }, + { + "epoch": 6.449460647223332, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16143 + }, + { + "epoch": 6.449860167798642, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16144 + }, + { + "epoch": 6.450259688373952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16145 + }, + { + "epoch": 6.4506592089492605, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16146 + }, + { + "epoch": 6.45105872952457, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16147 + }, + { + "epoch": 6.45145825009988, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16148 + }, + { + "epoch": 6.45185777067519, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16149 + }, + { + "epoch": 6.4522572912505, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16150 + }, + { + "epoch": 6.452656811825809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16151 + }, + { + "epoch": 6.453056332401118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16152 + }, + { + "epoch": 6.453455852976428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16153 + }, + { + "epoch": 6.453855373551738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16154 + }, + { + "epoch": 6.454254894127048, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16155 + }, + { + "epoch": 6.4546544147023575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16156 + }, + { + "epoch": 6.455053935277666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16157 + }, + { + "epoch": 6.455453455852976, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16158 + }, + { + "epoch": 6.455852976428286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16159 + }, + { + "epoch": 6.456252497003596, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16160 + }, + { + "epoch": 6.4566520175789055, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16161 + }, + { + "epoch": 6.457051538154215, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16162 + }, + { + "epoch": 6.457451058729524, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16163 + }, + { + "epoch": 6.457850579304834, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16164 + }, + { + "epoch": 6.458250099880144, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16165 + }, + { + "epoch": 6.4586496204554535, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16166 + }, + { + "epoch": 6.459049141030763, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16167 + }, + { + "epoch": 6.459448661606073, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16168 + }, + { + "epoch": 6.459848182181382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16169 + }, + { + "epoch": 6.460247702756692, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 16170 + }, + { + "epoch": 6.460647223332002, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16171 + }, + { + "epoch": 6.461046743907311, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16172 + }, + { + "epoch": 6.461446264482621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16173 + }, + { + "epoch": 6.46184578505793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16174 + }, + { + "epoch": 6.46224530563324, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16175 + }, + { + "epoch": 6.46264482620855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16176 + }, + { + "epoch": 6.463044346783859, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16177 + }, + { + "epoch": 6.463443867359169, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16178 + }, + { + "epoch": 6.463843387934479, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16179 + }, + { + "epoch": 6.464242908509788, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16180 + }, + { + "epoch": 6.464642429085098, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16181 + }, + { + "epoch": 6.4650419496604075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16182 + }, + { + "epoch": 6.465441470235717, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16183 + }, + { + "epoch": 6.465840990811027, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16184 + }, + { + "epoch": 6.466240511386337, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16185 + }, + { + "epoch": 6.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16186 + }, + { + "epoch": 6.4670395525369555, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16187 + }, + { + "epoch": 6.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 16188 + }, + { + "epoch": 6.467838593687575, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16189 + }, + { + "epoch": 6.468238114262885, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16190 + }, + { + "epoch": 6.468637634838194, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16191 + }, + { + "epoch": 6.4690371554135035, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16192 + }, + { + "epoch": 6.469436675988813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16193 + }, + { + "epoch": 6.469836196564123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16194 + }, + { + "epoch": 6.470235717139433, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16195 + }, + { + "epoch": 6.470635237714743, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16196 + }, + { + "epoch": 6.471034758290052, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16197 + }, + { + "epoch": 6.471434278865361, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16198 + }, + { + "epoch": 6.471833799440671, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16199 + }, + { + "epoch": 6.472233320015981, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16200 + }, + { + "epoch": 6.472632840591291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.2986, + "step": 16201 + }, + { + "epoch": 6.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16202 + }, + { + "epoch": 6.473431881741909, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16203 + }, + { + "epoch": 6.473831402317219, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16204 + }, + { + "epoch": 6.474230922892529, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16205 + }, + { + "epoch": 6.474630443467839, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16206 + }, + { + "epoch": 6.475029964043149, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16207 + }, + { + "epoch": 6.4754294846184575, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16208 + }, + { + "epoch": 6.475829005193767, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16209 + }, + { + "epoch": 6.476228525769077, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16210 + }, + { + "epoch": 6.476628046344387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16211 + }, + { + "epoch": 6.477027566919697, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16212 + }, + { + "epoch": 6.477427087495006, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16213 + }, + { + "epoch": 6.477826608070315, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16214 + }, + { + "epoch": 6.478226128645625, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16215 + }, + { + "epoch": 6.478625649220935, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16216 + }, + { + "epoch": 6.479025169796245, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16217 + }, + { + "epoch": 6.479424690371554, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16218 + }, + { + "epoch": 6.479824210946864, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16219 + }, + { + "epoch": 6.480223731522173, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16220 + }, + { + "epoch": 6.480623252097483, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16221 + }, + { + "epoch": 6.481022772672793, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16222 + }, + { + "epoch": 6.4814222932481025, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16223 + }, + { + "epoch": 6.481821813823412, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16224 + }, + { + "epoch": 6.482221334398721, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16225 + }, + { + "epoch": 6.482620854974031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16226 + }, + { + "epoch": 6.483020375549341, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16227 + }, + { + "epoch": 6.4834198961246505, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16228 + }, + { + "epoch": 6.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16229 + }, + { + "epoch": 6.484218937275269, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16230 + }, + { + "epoch": 6.484618457850579, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16231 + }, + { + "epoch": 6.485017978425889, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16232 + }, + { + "epoch": 6.485417499001199, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16233 + }, + { + "epoch": 6.485817019576508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16234 + }, + { + "epoch": 6.486216540151818, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16235 + }, + { + "epoch": 6.486616060727127, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16236 + }, + { + "epoch": 6.487015581302437, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16237 + }, + { + "epoch": 6.487415101877747, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16238 + }, + { + "epoch": 6.487814622453056, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16239 + }, + { + "epoch": 6.488214143028366, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16240 + }, + { + "epoch": 6.488613663603676, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16241 + }, + { + "epoch": 6.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16242 + }, + { + "epoch": 6.489412704754295, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16243 + }, + { + "epoch": 6.4898122253296044, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16244 + }, + { + "epoch": 6.490211745904914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16245 + }, + { + "epoch": 6.490611266480224, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16246 + }, + { + "epoch": 6.491010787055533, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16247 + }, + { + "epoch": 6.491410307630843, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16248 + }, + { + "epoch": 6.4918098282061525, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16249 + }, + { + "epoch": 6.492209348781462, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16250 + }, + { + "epoch": 6.492608869356772, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 16251 + }, + { + "epoch": 6.493008389932082, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16252 + }, + { + "epoch": 6.493407910507391, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16253 + }, + { + "epoch": 6.4938074310827005, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16254 + }, + { + "epoch": 6.49420695165801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16255 + }, + { + "epoch": 6.49460647223332, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16256 + }, + { + "epoch": 6.49500599280863, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16257 + }, + { + "epoch": 6.49540551338394, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16258 + }, + { + "epoch": 6.495805033959249, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16259 + }, + { + "epoch": 6.496204554534558, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16260 + }, + { + "epoch": 6.496604075109868, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16261 + }, + { + "epoch": 6.497003595685178, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16262 + }, + { + "epoch": 6.497403116260488, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16263 + }, + { + "epoch": 6.497802636835797, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16264 + }, + { + "epoch": 6.498202157411106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16265 + }, + { + "epoch": 6.498601677986416, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16266 + }, + { + "epoch": 6.499001198561726, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16267 + }, + { + "epoch": 6.499400719137036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 16268 + }, + { + "epoch": 6.4998002397123456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16269 + }, + { + "epoch": 6.5001997602876544, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16270 + }, + { + "epoch": 6.500599280862964, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16271 + }, + { + "epoch": 6.500998801438274, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16272 + }, + { + "epoch": 6.501398322013584, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16273 + }, + { + "epoch": 6.501797842588894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16274 + }, + { + "epoch": 6.502197363164203, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16275 + }, + { + "epoch": 6.502596883739512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16276 + }, + { + "epoch": 6.502996404314822, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16277 + }, + { + "epoch": 6.503395924890132, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16278 + }, + { + "epoch": 6.503795445465442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16279 + }, + { + "epoch": 6.504194966040751, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16280 + }, + { + "epoch": 6.50459448661606, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.2957, + "step": 16281 + }, + { + "epoch": 6.50499400719137, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16282 + }, + { + "epoch": 6.50539352776668, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16283 + }, + { + "epoch": 6.50579304834199, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 16284 + }, + { + "epoch": 6.5061925689172995, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16285 + }, + { + "epoch": 6.506592089492609, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16286 + }, + { + "epoch": 6.506991610067918, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16287 + }, + { + "epoch": 6.507391130643228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16288 + }, + { + "epoch": 6.507790651218538, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16289 + }, + { + "epoch": 6.5081901717938475, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 16290 + }, + { + "epoch": 6.508589692369157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16291 + }, + { + "epoch": 6.508989212944467, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16292 + }, + { + "epoch": 6.509388733519776, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16293 + }, + { + "epoch": 6.509788254095086, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16294 + }, + { + "epoch": 6.5101877746703956, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16295 + }, + { + "epoch": 6.510587295245705, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 16296 + }, + { + "epoch": 6.510986815821015, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16297 + }, + { + "epoch": 6.511386336396324, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16298 + }, + { + "epoch": 6.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16299 + }, + { + "epoch": 6.512185377546944, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16300 + }, + { + "epoch": 6.512584898122253, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16301 + }, + { + "epoch": 6.512984418697563, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16302 + }, + { + "epoch": 6.513383939272872, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16303 + }, + { + "epoch": 6.513783459848182, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16304 + }, + { + "epoch": 6.514182980423492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16305 + }, + { + "epoch": 6.514582500998801, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16306 + }, + { + "epoch": 6.514982021574111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16307 + }, + { + "epoch": 6.515381542149421, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16308 + }, + { + "epoch": 6.515781062724731, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16309 + }, + { + "epoch": 6.51618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16310 + }, + { + "epoch": 6.5165801038753495, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16311 + }, + { + "epoch": 6.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16312 + }, + { + "epoch": 6.517379145025969, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16313 + }, + { + "epoch": 6.517778665601279, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16314 + }, + { + "epoch": 6.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16315 + }, + { + "epoch": 6.5185777067518975, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16316 + }, + { + "epoch": 6.518977227327207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16317 + }, + { + "epoch": 6.519376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 16318 + }, + { + "epoch": 6.519776268477827, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16319 + }, + { + "epoch": 6.520175789053136, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16320 + }, + { + "epoch": 6.520575309628446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16321 + }, + { + "epoch": 6.520974830203755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16322 + }, + { + "epoch": 6.521374350779065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16323 + }, + { + "epoch": 6.521773871354375, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16324 + }, + { + "epoch": 6.522173391929685, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16325 + }, + { + "epoch": 6.5225729125049945, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16326 + }, + { + "epoch": 6.522972433080303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16327 + }, + { + "epoch": 6.523371953655613, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16328 + }, + { + "epoch": 6.523771474230923, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 16329 + }, + { + "epoch": 6.524170994806233, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16330 + }, + { + "epoch": 6.5245705153815425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16331 + }, + { + "epoch": 6.524970035956851, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16332 + }, + { + "epoch": 6.525369556532161, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 16333 + }, + { + "epoch": 6.525769077107471, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16334 + }, + { + "epoch": 6.526168597682781, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16335 + }, + { + "epoch": 6.526568118258091, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16336 + }, + { + "epoch": 6.5269676388333995, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16337 + }, + { + "epoch": 6.527367159408709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16338 + }, + { + "epoch": 6.527766679984019, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16339 + }, + { + "epoch": 6.528166200559329, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16340 + }, + { + "epoch": 6.528565721134639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16341 + }, + { + "epoch": 6.528965241709948, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16342 + }, + { + "epoch": 6.529364762285257, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16343 + }, + { + "epoch": 6.529764282860567, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16344 + }, + { + "epoch": 6.530163803435877, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16345 + }, + { + "epoch": 6.530563324011187, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 16346 + }, + { + "epoch": 6.5309628445864965, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 16347 + }, + { + "epoch": 6.531362365161806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16348 + }, + { + "epoch": 6.531761885737115, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16349 + }, + { + "epoch": 6.532161406312425, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16350 + }, + { + "epoch": 6.532560926887735, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16351 + }, + { + "epoch": 6.5329604474630445, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16352 + }, + { + "epoch": 6.533359968038354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16353 + }, + { + "epoch": 6.533759488613663, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 16354 + }, + { + "epoch": 6.534159009188973, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16355 + }, + { + "epoch": 6.534558529764283, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16356 + }, + { + "epoch": 6.5349580503395925, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16357 + }, + { + "epoch": 6.535357570914902, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 16358 + }, + { + "epoch": 6.535757091490212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16359 + }, + { + "epoch": 6.536156612065521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16360 + }, + { + "epoch": 6.536556132640831, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16361 + }, + { + "epoch": 6.536955653216141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 16362 + }, + { + "epoch": 6.53735517379145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16363 + }, + { + "epoch": 6.53775469436676, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16364 + }, + { + "epoch": 6.53815421494207, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 16365 + }, + { + "epoch": 6.538553735517379, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16366 + }, + { + "epoch": 6.538953256092689, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16367 + }, + { + "epoch": 6.539352776667998, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16368 + }, + { + "epoch": 6.539752297243308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16369 + }, + { + "epoch": 6.540151817818618, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16370 + }, + { + "epoch": 6.540551338393927, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16371 + }, + { + "epoch": 6.540950858969237, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16372 + }, + { + "epoch": 6.5413503795445465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16373 + }, + { + "epoch": 6.541749900119856, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16374 + }, + { + "epoch": 6.542149420695166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16375 + }, + { + "epoch": 6.542548941270476, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16376 + }, + { + "epoch": 6.542948461845785, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16377 + }, + { + "epoch": 6.5433479824210945, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16378 + }, + { + "epoch": 6.543747502996404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16379 + }, + { + "epoch": 6.544147023571714, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16380 + }, + { + "epoch": 6.544546544147024, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16381 + }, + { + "epoch": 6.544946064722334, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16382 + }, + { + "epoch": 6.5453455852976425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16383 + }, + { + "epoch": 6.545745105872952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16384 + }, + { + "epoch": 6.546144626448262, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16385 + }, + { + "epoch": 6.546544147023572, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16386 + }, + { + "epoch": 6.546943667598882, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16387 + }, + { + "epoch": 6.547343188174191, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 16388 + }, + { + "epoch": 6.5477427087495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16389 + }, + { + "epoch": 6.54814222932481, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16390 + }, + { + "epoch": 6.54854174990012, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16391 + }, + { + "epoch": 6.54894127047543, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16392 + }, + { + "epoch": 6.549340791050739, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16393 + }, + { + "epoch": 6.549740311626048, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16394 + }, + { + "epoch": 6.550139832201358, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16395 + }, + { + "epoch": 6.550539352776668, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16396 + }, + { + "epoch": 6.550938873351978, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 16397 + }, + { + "epoch": 6.551338393927288, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16398 + }, + { + "epoch": 6.551737914502597, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16399 + }, + { + "epoch": 6.552137435077906, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16400 + }, + { + "epoch": 6.552536955653216, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16401 + }, + { + "epoch": 6.552936476228526, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16402 + }, + { + "epoch": 6.553335996803836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16403 + }, + { + "epoch": 6.553735517379145, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 16404 + }, + { + "epoch": 6.554135037954454, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16405 + }, + { + "epoch": 6.554534558529764, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16406 + }, + { + "epoch": 6.554934079105074, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16407 + }, + { + "epoch": 6.555333599680384, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16408 + }, + { + "epoch": 6.555733120255693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16409 + }, + { + "epoch": 6.556132640831002, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16410 + }, + { + "epoch": 6.556532161406312, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16411 + }, + { + "epoch": 6.556931681981622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16412 + }, + { + "epoch": 6.557331202556932, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16413 + }, + { + "epoch": 6.5577307231322415, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16414 + }, + { + "epoch": 6.558130243707551, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16415 + }, + { + "epoch": 6.558529764282861, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16416 + }, + { + "epoch": 6.55892928485817, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16417 + }, + { + "epoch": 6.55932880543348, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16418 + }, + { + "epoch": 6.5597283260087895, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16419 + }, + { + "epoch": 6.560127846584099, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16420 + }, + { + "epoch": 6.560527367159409, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16421 + }, + { + "epoch": 6.560926887734718, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16422 + }, + { + "epoch": 6.561326408310028, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16423 + }, + { + "epoch": 6.561725928885338, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16424 + }, + { + "epoch": 6.562125449460647, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16425 + }, + { + "epoch": 6.562524970035957, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16426 + }, + { + "epoch": 6.562924490611266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16427 + }, + { + "epoch": 6.563324011186576, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16428 + }, + { + "epoch": 6.563723531761886, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16429 + }, + { + "epoch": 6.564123052337195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16430 + }, + { + "epoch": 6.564522572912505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16431 + }, + { + "epoch": 6.564922093487815, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16432 + }, + { + "epoch": 6.565321614063124, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16433 + }, + { + "epoch": 6.565721134638434, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16434 + }, + { + "epoch": 6.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16435 + }, + { + "epoch": 6.566520175789053, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16436 + }, + { + "epoch": 6.566919696364363, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16437 + }, + { + "epoch": 6.567319216939673, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16438 + }, + { + "epoch": 6.567718737514982, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 16439 + }, + { + "epoch": 6.5681182580902915, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16440 + }, + { + "epoch": 6.568517778665601, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16441 + }, + { + "epoch": 6.568917299240911, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16442 + }, + { + "epoch": 6.569316819816221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16443 + }, + { + "epoch": 6.56971634039153, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16444 + }, + { + "epoch": 6.5701158609668395, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16445 + }, + { + "epoch": 6.570515381542149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16446 + }, + { + "epoch": 6.570914902117459, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 16447 + }, + { + "epoch": 6.571314422692769, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16448 + }, + { + "epoch": 6.571713943268079, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16449 + }, + { + "epoch": 6.572113463843388, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16450 + }, + { + "epoch": 6.572512984418697, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16451 + }, + { + "epoch": 6.572912504994007, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16452 + }, + { + "epoch": 6.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16453 + }, + { + "epoch": 6.573711546144627, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16454 + }, + { + "epoch": 6.5741110667199365, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16455 + }, + { + "epoch": 6.574510587295245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16456 + }, + { + "epoch": 6.574910107870555, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16457 + }, + { + "epoch": 6.575309628445865, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16458 + }, + { + "epoch": 6.575709149021175, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16459 + }, + { + "epoch": 6.5761086695964845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16460 + }, + { + "epoch": 6.5765081901717934, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16461 + }, + { + "epoch": 6.576907710747103, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16462 + }, + { + "epoch": 6.577307231322413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16463 + }, + { + "epoch": 6.577706751897723, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16464 + }, + { + "epoch": 6.578106272473033, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16465 + }, + { + "epoch": 6.578505793048342, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16466 + }, + { + "epoch": 6.578905313623651, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16467 + }, + { + "epoch": 6.579304834198961, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16468 + }, + { + "epoch": 6.579704354774271, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16469 + }, + { + "epoch": 6.580103875349581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16470 + }, + { + "epoch": 6.58050339592489, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16471 + }, + { + "epoch": 6.5809029165002, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16472 + }, + { + "epoch": 6.581302437075509, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16473 + }, + { + "epoch": 6.581701957650819, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16474 + }, + { + "epoch": 6.582101478226129, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16475 + }, + { + "epoch": 6.5825009988014385, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16476 + }, + { + "epoch": 6.582900519376748, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16477 + }, + { + "epoch": 6.583300039952057, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16478 + }, + { + "epoch": 6.583699560527367, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16479 + }, + { + "epoch": 6.584099081102677, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16480 + }, + { + "epoch": 6.5844986016779865, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16481 + }, + { + "epoch": 6.584898122253296, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16482 + }, + { + "epoch": 6.585297642828605, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16483 + }, + { + "epoch": 6.585697163403915, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16484 + }, + { + "epoch": 6.586096683979225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16485 + }, + { + "epoch": 6.5864962045545346, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16486 + }, + { + "epoch": 6.586895725129844, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16487 + }, + { + "epoch": 6.587295245705154, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16488 + }, + { + "epoch": 6.587694766280464, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16489 + }, + { + "epoch": 6.588094286855773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16490 + }, + { + "epoch": 6.588493807431083, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16491 + }, + { + "epoch": 6.588893328006392, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16492 + }, + { + "epoch": 6.589292848581702, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16493 + }, + { + "epoch": 6.589692369157012, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16494 + }, + { + "epoch": 6.590091889732321, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16495 + }, + { + "epoch": 6.590491410307631, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16496 + }, + { + "epoch": 6.59089093088294, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16497 + }, + { + "epoch": 6.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16498 + }, + { + "epoch": 6.59168997203356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16499 + }, + { + "epoch": 6.592089492608869, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16500 + }, + { + "epoch": 6.592489013184179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16501 + }, + { + "epoch": 6.5928885337594885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16502 + }, + { + "epoch": 6.593288054334798, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16503 + }, + { + "epoch": 6.593687574910108, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16504 + }, + { + "epoch": 6.594087095485418, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16505 + }, + { + "epoch": 6.594486616060728, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16506 + }, + { + "epoch": 6.5948861366360365, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16507 + }, + { + "epoch": 6.595285657211346, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16508 + }, + { + "epoch": 6.595685177786656, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16509 + }, + { + "epoch": 6.596084698361966, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16510 + }, + { + "epoch": 6.596484218937276, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16511 + }, + { + "epoch": 6.5968837395125846, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16512 + }, + { + "epoch": 6.597283260087894, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16513 + }, + { + "epoch": 6.597682780663204, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16514 + }, + { + "epoch": 6.598082301238514, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16515 + }, + { + "epoch": 6.598481821813824, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 16516 + }, + { + "epoch": 6.598881342389133, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 16517 + }, + { + "epoch": 6.599280862964442, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16518 + }, + { + "epoch": 6.599680383539752, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16519 + }, + { + "epoch": 6.600079904115062, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16520 + }, + { + "epoch": 6.600479424690372, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16521 + }, + { + "epoch": 6.6008789452656815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16522 + }, + { + "epoch": 6.60127846584099, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16523 + }, + { + "epoch": 6.6016779864163, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16524 + }, + { + "epoch": 6.60207750699161, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16525 + }, + { + "epoch": 6.60247702756692, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16526 + }, + { + "epoch": 6.60287654814223, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16527 + }, + { + "epoch": 6.603276068717539, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16528 + }, + { + "epoch": 6.603675589292848, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16529 + }, + { + "epoch": 6.604075109868158, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16530 + }, + { + "epoch": 6.604474630443468, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16531 + }, + { + "epoch": 6.604874151018778, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16532 + }, + { + "epoch": 6.605273671594087, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16533 + }, + { + "epoch": 6.605673192169396, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16534 + }, + { + "epoch": 6.606072712744706, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16535 + }, + { + "epoch": 6.606472233320016, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16536 + }, + { + "epoch": 6.606871753895326, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16537 + }, + { + "epoch": 6.6072712744706354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16538 + }, + { + "epoch": 6.607670795045945, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16539 + }, + { + "epoch": 6.608070315621254, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16540 + }, + { + "epoch": 6.608469836196564, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16541 + }, + { + "epoch": 6.608869356771874, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 16542 + }, + { + "epoch": 6.6092688773471835, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16543 + }, + { + "epoch": 6.609668397922493, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16544 + }, + { + "epoch": 6.610067918497803, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16545 + }, + { + "epoch": 6.610467439073112, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16546 + }, + { + "epoch": 6.610866959648422, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16547 + }, + { + "epoch": 6.6112664802237315, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16548 + }, + { + "epoch": 6.611666000799041, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16549 + }, + { + "epoch": 6.612065521374351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16550 + }, + { + "epoch": 6.61246504194966, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16551 + }, + { + "epoch": 6.61286456252497, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16552 + }, + { + "epoch": 6.61326408310028, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16553 + }, + { + "epoch": 6.613663603675589, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16554 + }, + { + "epoch": 6.614063124250899, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16555 + }, + { + "epoch": 6.614462644826209, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16556 + }, + { + "epoch": 6.614862165401518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16557 + }, + { + "epoch": 6.615261685976828, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16558 + }, + { + "epoch": 6.615661206552137, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16559 + }, + { + "epoch": 6.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16560 + }, + { + "epoch": 6.616460247702757, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16561 + }, + { + "epoch": 6.616859768278067, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16562 + }, + { + "epoch": 6.617259288853376, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16563 + }, + { + "epoch": 6.6176588094286855, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16564 + }, + { + "epoch": 6.618058330003995, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16565 + }, + { + "epoch": 6.618457850579305, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16566 + }, + { + "epoch": 6.618857371154615, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16567 + }, + { + "epoch": 6.619256891729924, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16568 + }, + { + "epoch": 6.6196564123052335, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16569 + }, + { + "epoch": 6.620055932880543, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16570 + }, + { + "epoch": 6.620455453455853, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16571 + }, + { + "epoch": 6.620854974031163, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16572 + }, + { + "epoch": 6.621254494606472, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 16573 + }, + { + "epoch": 6.6216540151817815, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16574 + }, + { + "epoch": 6.622053535757091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 16575 + }, + { + "epoch": 6.622453056332401, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16576 + }, + { + "epoch": 6.622852576907711, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16577 + }, + { + "epoch": 6.623252097483021, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16578 + }, + { + "epoch": 6.6236516180583305, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16579 + }, + { + "epoch": 6.624051138633639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16580 + }, + { + "epoch": 6.624450659208949, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16581 + }, + { + "epoch": 6.624850179784259, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16582 + }, + { + "epoch": 6.625249700359569, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16583 + }, + { + "epoch": 6.6256492209348785, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16584 + }, + { + "epoch": 6.626048741510187, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16585 + }, + { + "epoch": 6.626448262085497, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16586 + }, + { + "epoch": 6.626847782660807, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16587 + }, + { + "epoch": 6.627247303236117, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16588 + }, + { + "epoch": 6.627646823811427, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16589 + }, + { + "epoch": 6.6280463443867355, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16590 + }, + { + "epoch": 6.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16591 + }, + { + "epoch": 6.628845385537355, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16592 + }, + { + "epoch": 6.629244906112665, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16593 + }, + { + "epoch": 6.629644426687975, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16594 + }, + { + "epoch": 6.630043947263284, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16595 + }, + { + "epoch": 6.630443467838594, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16596 + }, + { + "epoch": 6.630842988413903, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16597 + }, + { + "epoch": 6.631242508989213, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16598 + }, + { + "epoch": 6.631642029564523, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 16599 + }, + { + "epoch": 6.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16600 + }, + { + "epoch": 6.632441070715142, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16601 + }, + { + "epoch": 6.632840591290451, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16602 + }, + { + "epoch": 6.633240111865761, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16603 + }, + { + "epoch": 6.633639632441071, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16604 + }, + { + "epoch": 6.6340391530163805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16605 + }, + { + "epoch": 6.63443867359169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16606 + }, + { + "epoch": 6.634838194166999, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16607 + }, + { + "epoch": 6.635237714742309, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16608 + }, + { + "epoch": 6.635637235317619, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 16609 + }, + { + "epoch": 6.6360367558929285, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16610 + }, + { + "epoch": 6.636436276468238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16611 + }, + { + "epoch": 6.636835797043548, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16612 + }, + { + "epoch": 6.637235317618857, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16613 + }, + { + "epoch": 6.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16614 + }, + { + "epoch": 6.638034358769477, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16615 + }, + { + "epoch": 6.638433879344786, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16616 + }, + { + "epoch": 6.638833399920096, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 16617 + }, + { + "epoch": 6.639232920495406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16618 + }, + { + "epoch": 6.639632441070715, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16619 + }, + { + "epoch": 6.640031961646025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16620 + }, + { + "epoch": 6.640431482221334, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16621 + }, + { + "epoch": 6.640831002796644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16622 + }, + { + "epoch": 6.641230523371954, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16623 + }, + { + "epoch": 6.641630043947263, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16624 + }, + { + "epoch": 6.642029564522573, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16625 + }, + { + "epoch": 6.642429085097882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16626 + }, + { + "epoch": 6.642828605673192, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16627 + }, + { + "epoch": 6.643228126248502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16628 + }, + { + "epoch": 6.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16629 + }, + { + "epoch": 6.644027167399121, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16630 + }, + { + "epoch": 6.6444266879744305, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16631 + }, + { + "epoch": 6.64482620854974, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16632 + }, + { + "epoch": 6.64522572912505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16633 + }, + { + "epoch": 6.64562524970036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16634 + }, + { + "epoch": 6.64602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16635 + }, + { + "epoch": 6.6464242908509785, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16636 + }, + { + "epoch": 6.646823811426288, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16637 + }, + { + "epoch": 6.647223332001598, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16638 + }, + { + "epoch": 6.647622852576908, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16639 + }, + { + "epoch": 6.648022373152218, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16640 + }, + { + "epoch": 6.648421893727527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16641 + }, + { + "epoch": 6.648821414302836, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16642 + }, + { + "epoch": 6.649220934878146, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16643 + }, + { + "epoch": 6.649620455453456, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16644 + }, + { + "epoch": 6.650019976028766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16645 + }, + { + "epoch": 6.6504194966040755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16646 + }, + { + "epoch": 6.650819017179384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16647 + }, + { + "epoch": 6.651218537754694, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16648 + }, + { + "epoch": 6.651618058330004, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16649 + }, + { + "epoch": 6.652017578905314, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16650 + }, + { + "epoch": 6.6524170994806235, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16651 + }, + { + "epoch": 6.652816620055933, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16652 + }, + { + "epoch": 6.653216140631242, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16653 + }, + { + "epoch": 6.653615661206552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16654 + }, + { + "epoch": 6.654015181781862, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16655 + }, + { + "epoch": 6.654414702357172, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16656 + }, + { + "epoch": 6.654814222932481, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16657 + }, + { + "epoch": 6.65521374350779, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16658 + }, + { + "epoch": 6.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16659 + }, + { + "epoch": 6.65601278465841, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16660 + }, + { + "epoch": 6.65641230523372, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16661 + }, + { + "epoch": 6.656811825809029, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16662 + }, + { + "epoch": 6.657211346384338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16663 + }, + { + "epoch": 6.657610866959648, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16664 + }, + { + "epoch": 6.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16665 + }, + { + "epoch": 6.658409908110268, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16666 + }, + { + "epoch": 6.6588094286855775, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16667 + }, + { + "epoch": 6.659208949260887, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16668 + }, + { + "epoch": 6.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16669 + }, + { + "epoch": 6.660007990411506, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16670 + }, + { + "epoch": 6.660407510986816, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16671 + }, + { + "epoch": 6.6608070315621255, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16672 + }, + { + "epoch": 6.661206552137435, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16673 + }, + { + "epoch": 6.661606072712745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16674 + }, + { + "epoch": 6.662005593288054, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16675 + }, + { + "epoch": 6.662405113863364, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16676 + }, + { + "epoch": 6.6628046344386735, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16677 + }, + { + "epoch": 6.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16678 + }, + { + "epoch": 6.663603675589293, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16679 + }, + { + "epoch": 6.664003196164602, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16680 + }, + { + "epoch": 6.664402716739912, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16681 + }, + { + "epoch": 6.664802237315222, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16682 + }, + { + "epoch": 6.665201757890531, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16683 + }, + { + "epoch": 6.665601278465841, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16684 + }, + { + "epoch": 6.666000799041151, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16685 + }, + { + "epoch": 6.666400319616461, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16686 + }, + { + "epoch": 6.66679984019177, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16687 + }, + { + "epoch": 6.667199360767079, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16688 + }, + { + "epoch": 6.667598881342389, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16689 + }, + { + "epoch": 6.667998401917699, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16690 + }, + { + "epoch": 6.668397922493009, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16691 + }, + { + "epoch": 6.668797443068318, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16692 + }, + { + "epoch": 6.6691969636436275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16693 + }, + { + "epoch": 6.669596484218937, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16694 + }, + { + "epoch": 6.669996004794247, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16695 + }, + { + "epoch": 6.670395525369557, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16696 + }, + { + "epoch": 6.670795045944866, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16697 + }, + { + "epoch": 6.6711945665201755, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16698 + }, + { + "epoch": 6.671594087095485, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16699 + }, + { + "epoch": 6.671993607670795, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16700 + }, + { + "epoch": 6.672393128246105, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16701 + }, + { + "epoch": 6.672792648821415, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16702 + }, + { + "epoch": 6.6731921693967235, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16703 + }, + { + "epoch": 6.673591689972033, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16704 + }, + { + "epoch": 6.673991210547343, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16705 + }, + { + "epoch": 6.674390731122653, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 16706 + }, + { + "epoch": 6.674790251697963, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16707 + }, + { + "epoch": 6.6751897722732725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16708 + }, + { + "epoch": 6.675589292848581, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16709 + }, + { + "epoch": 6.675988813423891, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16710 + }, + { + "epoch": 6.676388333999201, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16711 + }, + { + "epoch": 6.676787854574511, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16712 + }, + { + "epoch": 6.6771873751498205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16713 + }, + { + "epoch": 6.677586895725129, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16714 + }, + { + "epoch": 6.677986416300439, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16715 + }, + { + "epoch": 6.678385936875749, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16716 + }, + { + "epoch": 6.678785457451059, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16717 + }, + { + "epoch": 6.679184978026369, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16718 + }, + { + "epoch": 6.679584498601678, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16719 + }, + { + "epoch": 6.679984019176987, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 16720 + }, + { + "epoch": 6.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16721 + }, + { + "epoch": 6.680783060327607, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16722 + }, + { + "epoch": 6.681182580902917, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16723 + }, + { + "epoch": 6.681582101478226, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16724 + }, + { + "epoch": 6.681981622053536, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16725 + }, + { + "epoch": 6.682381142628845, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16726 + }, + { + "epoch": 6.682780663204155, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16727 + }, + { + "epoch": 6.683180183779465, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16728 + }, + { + "epoch": 6.683579704354774, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16729 + }, + { + "epoch": 6.683979224930084, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16730 + }, + { + "epoch": 6.684378745505393, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16731 + }, + { + "epoch": 6.684778266080703, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16732 + }, + { + "epoch": 6.685177786656013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16733 + }, + { + "epoch": 6.6855773072313225, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16734 + }, + { + "epoch": 6.685976827806632, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16735 + }, + { + "epoch": 6.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16736 + }, + { + "epoch": 6.686775868957251, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16737 + }, + { + "epoch": 6.687175389532561, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 16738 + }, + { + "epoch": 6.6875749101078705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16739 + }, + { + "epoch": 6.68797443068318, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16740 + }, + { + "epoch": 6.68837395125849, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16741 + }, + { + "epoch": 6.6887734718338, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16742 + }, + { + "epoch": 6.689172992409109, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16743 + }, + { + "epoch": 6.689572512984419, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16744 + }, + { + "epoch": 6.689972033559728, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16745 + }, + { + "epoch": 6.690371554135038, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16746 + }, + { + "epoch": 6.690771074710348, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16747 + }, + { + "epoch": 6.691170595285657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16748 + }, + { + "epoch": 6.691570115860967, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16749 + }, + { + "epoch": 6.691969636436276, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16750 + }, + { + "epoch": 6.692369157011586, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 16751 + }, + { + "epoch": 6.692768677586896, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 16752 + }, + { + "epoch": 6.693168198162205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16753 + }, + { + "epoch": 6.693567718737515, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16754 + }, + { + "epoch": 6.6939672393128244, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16755 + }, + { + "epoch": 6.694366759888134, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16756 + }, + { + "epoch": 6.694766280463444, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16757 + }, + { + "epoch": 6.695165801038754, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16758 + }, + { + "epoch": 6.695565321614064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16759 + }, + { + "epoch": 6.6959648421893725, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16760 + }, + { + "epoch": 6.696364362764682, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16761 + }, + { + "epoch": 6.696763883339992, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16762 + }, + { + "epoch": 6.697163403915302, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16763 + }, + { + "epoch": 6.697562924490612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16764 + }, + { + "epoch": 6.6979624450659205, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16765 + }, + { + "epoch": 6.69836196564123, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16766 + }, + { + "epoch": 6.69876148621654, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16767 + }, + { + "epoch": 6.69916100679185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16768 + }, + { + "epoch": 6.69956052736716, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16769 + }, + { + "epoch": 6.699960047942469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16770 + }, + { + "epoch": 6.700359568517778, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16771 + }, + { + "epoch": 6.700759089093088, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16772 + }, + { + "epoch": 6.701158609668398, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16773 + }, + { + "epoch": 6.701558130243708, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16774 + }, + { + "epoch": 6.7019576508190175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16775 + }, + { + "epoch": 6.702357171394327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16776 + }, + { + "epoch": 6.702756691969636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16777 + }, + { + "epoch": 6.703156212544946, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16778 + }, + { + "epoch": 6.703555733120256, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16779 + }, + { + "epoch": 6.7039552536955656, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16780 + }, + { + "epoch": 6.704354774270875, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16781 + }, + { + "epoch": 6.704754294846184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16782 + }, + { + "epoch": 6.705153815421494, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16783 + }, + { + "epoch": 6.705553335996804, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16784 + }, + { + "epoch": 6.705952856572114, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16785 + }, + { + "epoch": 6.706352377147423, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16786 + }, + { + "epoch": 6.706751897722732, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16787 + }, + { + "epoch": 6.707151418298042, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16788 + }, + { + "epoch": 6.707550938873352, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16789 + }, + { + "epoch": 6.707950459448662, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16790 + }, + { + "epoch": 6.708349980023971, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16791 + }, + { + "epoch": 6.708749500599281, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16792 + }, + { + "epoch": 6.70914902117459, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16793 + }, + { + "epoch": 6.7095485417499, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16794 + }, + { + "epoch": 6.70994806232521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16795 + }, + { + "epoch": 6.7103475829005195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16796 + }, + { + "epoch": 6.710747103475829, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16797 + }, + { + "epoch": 6.711146624051139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16798 + }, + { + "epoch": 6.711546144626448, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16799 + }, + { + "epoch": 6.711945665201758, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16800 + }, + { + "epoch": 6.7123451857770675, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16801 + }, + { + "epoch": 6.712744706352377, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.2968, + "step": 16802 + }, + { + "epoch": 6.713144226927687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16803 + }, + { + "epoch": 6.713543747502996, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16804 + }, + { + "epoch": 6.713943268078306, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16805 + }, + { + "epoch": 6.714342788653616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16806 + }, + { + "epoch": 6.714742309228925, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16807 + }, + { + "epoch": 6.715141829804235, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 16808 + }, + { + "epoch": 6.715541350379545, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16809 + }, + { + "epoch": 6.715940870954854, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 16810 + }, + { + "epoch": 6.716340391530164, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16811 + }, + { + "epoch": 6.716739912105473, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16812 + }, + { + "epoch": 6.717139432680783, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16813 + }, + { + "epoch": 6.717538953256093, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16814 + }, + { + "epoch": 6.717938473831403, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16815 + }, + { + "epoch": 6.718337994406712, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16816 + }, + { + "epoch": 6.718737514982021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16817 + }, + { + "epoch": 6.719137035557331, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16818 + }, + { + "epoch": 6.719536556132641, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16819 + }, + { + "epoch": 6.719936076707951, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16820 + }, + { + "epoch": 6.72033559728326, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16821 + }, + { + "epoch": 6.7207351178585695, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 16822 + }, + { + "epoch": 6.721134638433879, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16823 + }, + { + "epoch": 6.721534159009189, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 16824 + }, + { + "epoch": 6.721933679584499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16825 + }, + { + "epoch": 6.722333200159808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16826 + }, + { + "epoch": 6.7227327207351175, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16827 + }, + { + "epoch": 6.723132241310427, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16828 + }, + { + "epoch": 6.723531761885737, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16829 + }, + { + "epoch": 6.723931282461047, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16830 + }, + { + "epoch": 6.724330803036357, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16831 + }, + { + "epoch": 6.7247303236116664, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16832 + }, + { + "epoch": 6.725129844186975, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16833 + }, + { + "epoch": 6.725529364762285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16834 + }, + { + "epoch": 6.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16835 + }, + { + "epoch": 6.726328405912905, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16836 + }, + { + "epoch": 6.7267279264882145, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16837 + }, + { + "epoch": 6.727127447063523, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16838 + }, + { + "epoch": 6.727526967638833, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16839 + }, + { + "epoch": 6.727926488214143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16840 + }, + { + "epoch": 6.728326008789453, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16841 + }, + { + "epoch": 6.7287255293647625, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16842 + }, + { + "epoch": 6.729125049940071, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16843 + }, + { + "epoch": 6.729524570515381, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16844 + }, + { + "epoch": 6.729924091090691, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16845 + }, + { + "epoch": 6.730323611666001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16846 + }, + { + "epoch": 6.730723132241311, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16847 + }, + { + "epoch": 6.73112265281662, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16848 + }, + { + "epoch": 6.73152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16849 + }, + { + "epoch": 6.731921693967239, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16850 + }, + { + "epoch": 6.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16851 + }, + { + "epoch": 6.732720735117859, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16852 + }, + { + "epoch": 6.733120255693168, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16853 + }, + { + "epoch": 6.733519776268478, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16854 + }, + { + "epoch": 6.733919296843787, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16855 + }, + { + "epoch": 6.734318817419097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16856 + }, + { + "epoch": 6.734718337994407, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16857 + }, + { + "epoch": 6.7351178585697165, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16858 + }, + { + "epoch": 6.735517379145026, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16859 + }, + { + "epoch": 6.735916899720335, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16860 + }, + { + "epoch": 6.736316420295645, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16861 + }, + { + "epoch": 6.736715940870955, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16862 + }, + { + "epoch": 6.7371154614462645, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 16863 + }, + { + "epoch": 6.737514982021574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16864 + }, + { + "epoch": 6.737914502596884, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16865 + }, + { + "epoch": 6.738314023172194, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16866 + }, + { + "epoch": 6.738713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16867 + }, + { + "epoch": 6.7391130643228125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16868 + }, + { + "epoch": 6.739512584898122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16869 + }, + { + "epoch": 6.739912105473432, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16870 + }, + { + "epoch": 6.740311626048742, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16871 + }, + { + "epoch": 6.740711146624051, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16872 + }, + { + "epoch": 6.741110667199361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16873 + }, + { + "epoch": 6.74151018777467, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16874 + }, + { + "epoch": 6.74190970834998, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16875 + }, + { + "epoch": 6.74230922892529, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16876 + }, + { + "epoch": 6.742708749500599, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16877 + }, + { + "epoch": 6.743108270075909, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16878 + }, + { + "epoch": 6.743507790651218, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16879 + }, + { + "epoch": 6.743907311226528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16880 + }, + { + "epoch": 6.744306831801838, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16881 + }, + { + "epoch": 6.744706352377148, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16882 + }, + { + "epoch": 6.745105872952457, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16883 + }, + { + "epoch": 6.7455053935277665, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16884 + }, + { + "epoch": 6.745904914103076, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16885 + }, + { + "epoch": 6.746304434678386, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 16886 + }, + { + "epoch": 6.746703955253696, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16887 + }, + { + "epoch": 6.747103475829006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16888 + }, + { + "epoch": 6.7475029964043145, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16889 + }, + { + "epoch": 6.747902516979624, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16890 + }, + { + "epoch": 6.748302037554934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16891 + }, + { + "epoch": 6.748701558130244, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16892 + }, + { + "epoch": 6.749101078705554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16893 + }, + { + "epoch": 6.7495005992808625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16894 + }, + { + "epoch": 6.749900119856172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16895 + }, + { + "epoch": 6.750299640431482, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16896 + }, + { + "epoch": 6.750699161006792, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16897 + }, + { + "epoch": 6.751098681582102, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16898 + }, + { + "epoch": 6.7514982021574115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.296, + "step": 16899 + }, + { + "epoch": 6.75189772273272, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16900 + }, + { + "epoch": 6.75229724330803, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16901 + }, + { + "epoch": 6.75269676388334, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16902 + }, + { + "epoch": 6.75309628445865, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16903 + }, + { + "epoch": 6.7534958050339595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16904 + }, + { + "epoch": 6.753895325609269, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16905 + }, + { + "epoch": 6.754294846184578, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16906 + }, + { + "epoch": 6.754694366759888, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 16907 + }, + { + "epoch": 6.755093887335198, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16908 + }, + { + "epoch": 6.755493407910508, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16909 + }, + { + "epoch": 6.755892928485817, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 16910 + }, + { + "epoch": 6.756292449061126, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16911 + }, + { + "epoch": 6.756691969636436, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16912 + }, + { + "epoch": 6.757091490211746, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16913 + }, + { + "epoch": 6.757491010787056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16914 + }, + { + "epoch": 6.757890531362365, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16915 + }, + { + "epoch": 6.758290051937674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16916 + }, + { + "epoch": 6.758689572512984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16917 + }, + { + "epoch": 6.759089093088294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16918 + }, + { + "epoch": 6.759488613663604, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16919 + }, + { + "epoch": 6.759888134238913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16920 + }, + { + "epoch": 6.760287654814223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16921 + }, + { + "epoch": 6.760687175389533, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16922 + }, + { + "epoch": 6.761086695964842, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16923 + }, + { + "epoch": 6.761486216540152, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16924 + }, + { + "epoch": 6.7618857371154615, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16925 + }, + { + "epoch": 6.762285257690771, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16926 + }, + { + "epoch": 6.762684778266081, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16927 + }, + { + "epoch": 6.76308429884139, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16928 + }, + { + "epoch": 6.7634838194167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16929 + }, + { + "epoch": 6.7638833399920095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16930 + }, + { + "epoch": 6.764282860567319, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16931 + }, + { + "epoch": 6.764682381142629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16932 + }, + { + "epoch": 6.765081901717938, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16933 + }, + { + "epoch": 6.765481422293248, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16934 + }, + { + "epoch": 6.765880942868558, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16935 + }, + { + "epoch": 6.766280463443867, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16936 + }, + { + "epoch": 6.766679984019177, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16937 + }, + { + "epoch": 6.767079504594487, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16938 + }, + { + "epoch": 6.767479025169797, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16939 + }, + { + "epoch": 6.767878545745106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16940 + }, + { + "epoch": 6.768278066320415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16941 + }, + { + "epoch": 6.768677586895725, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16942 + }, + { + "epoch": 6.769077107471035, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16943 + }, + { + "epoch": 6.769476628046345, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16944 + }, + { + "epoch": 6.769876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16945 + }, + { + "epoch": 6.770275669196963, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16946 + }, + { + "epoch": 6.770675189772273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16947 + }, + { + "epoch": 6.771074710347583, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16948 + }, + { + "epoch": 6.771474230922893, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16949 + }, + { + "epoch": 6.771873751498202, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16950 + }, + { + "epoch": 6.7722732720735115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16951 + }, + { + "epoch": 6.772672792648821, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16952 + }, + { + "epoch": 6.773072313224131, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16953 + }, + { + "epoch": 6.773471833799441, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16954 + }, + { + "epoch": 6.773871354374751, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16955 + }, + { + "epoch": 6.77427087495006, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16956 + }, + { + "epoch": 6.774670395525369, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16957 + }, + { + "epoch": 6.775069916100679, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16958 + }, + { + "epoch": 6.775469436675989, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16959 + }, + { + "epoch": 6.775868957251299, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16960 + }, + { + "epoch": 6.7762684778266085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16961 + }, + { + "epoch": 6.776667998401917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16962 + }, + { + "epoch": 6.777067518977227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16963 + }, + { + "epoch": 6.777467039552537, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 16964 + }, + { + "epoch": 6.777866560127847, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16965 + }, + { + "epoch": 6.7782660807031565, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16966 + }, + { + "epoch": 6.778665601278465, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 16967 + }, + { + "epoch": 6.779065121853775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16968 + }, + { + "epoch": 6.779464642429085, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16969 + }, + { + "epoch": 6.779864163004395, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 16970 + }, + { + "epoch": 6.7802636835797045, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16971 + }, + { + "epoch": 6.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16972 + }, + { + "epoch": 6.781062724730323, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16973 + }, + { + "epoch": 6.781462245305633, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16974 + }, + { + "epoch": 6.781861765880943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16975 + }, + { + "epoch": 6.782261286456253, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16976 + }, + { + "epoch": 6.782660807031562, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16977 + }, + { + "epoch": 6.783060327606872, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16978 + }, + { + "epoch": 6.783459848182181, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16979 + }, + { + "epoch": 6.783859368757491, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16980 + }, + { + "epoch": 6.784258889332801, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16981 + }, + { + "epoch": 6.78465840990811, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16982 + }, + { + "epoch": 6.78505793048342, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16983 + }, + { + "epoch": 6.785457451058729, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16984 + }, + { + "epoch": 6.785856971634039, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16985 + }, + { + "epoch": 6.786256492209349, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16986 + }, + { + "epoch": 6.7866560127846585, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16987 + }, + { + "epoch": 6.787055533359968, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16988 + }, + { + "epoch": 6.787455053935278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16989 + }, + { + "epoch": 6.787854574510587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16990 + }, + { + "epoch": 6.788254095085897, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16991 + }, + { + "epoch": 6.7886536156612065, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16992 + }, + { + "epoch": 6.789053136236516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16993 + }, + { + "epoch": 6.789452656811826, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16994 + }, + { + "epoch": 6.789852177387136, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16995 + }, + { + "epoch": 6.790251697962445, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16996 + }, + { + "epoch": 6.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16997 + }, + { + "epoch": 6.791050739113064, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16998 + }, + { + "epoch": 6.791450259688374, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16999 + }, + { + "epoch": 6.791849780263684, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17000 + }, + { + "epoch": 6.792249300838993, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17001 + }, + { + "epoch": 6.792648821414303, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17002 + }, + { + "epoch": 6.793048341989612, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17003 + }, + { + "epoch": 6.793447862564922, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17004 + }, + { + "epoch": 6.793847383140232, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17005 + }, + { + "epoch": 6.794246903715541, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17006 + }, + { + "epoch": 6.794646424290851, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17007 + }, + { + "epoch": 6.79504594486616, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17008 + }, + { + "epoch": 6.79544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17009 + }, + { + "epoch": 6.79584498601678, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17010 + }, + { + "epoch": 6.79624450659209, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17011 + }, + { + "epoch": 6.7966440271674, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17012 + }, + { + "epoch": 6.7970435477427085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17013 + }, + { + "epoch": 6.797443068318018, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17014 + }, + { + "epoch": 6.797842588893328, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17015 + }, + { + "epoch": 6.798242109468638, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17016 + }, + { + "epoch": 6.798641630043948, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17017 + }, + { + "epoch": 6.7990411506192565, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17018 + }, + { + "epoch": 6.799440671194566, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17019 + }, + { + "epoch": 6.799840191769876, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17020 + }, + { + "epoch": 6.800239712345186, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17021 + }, + { + "epoch": 6.800639232920496, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17022 + }, + { + "epoch": 6.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17023 + }, + { + "epoch": 6.801438274071114, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17024 + }, + { + "epoch": 6.801837794646424, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17025 + }, + { + "epoch": 6.802237315221734, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17026 + }, + { + "epoch": 6.802636835797044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17027 + }, + { + "epoch": 6.8030363563723535, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17028 + }, + { + "epoch": 6.803435876947663, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17029 + }, + { + "epoch": 6.803835397522972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17030 + }, + { + "epoch": 6.804234918098282, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17031 + }, + { + "epoch": 6.804634438673592, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17032 + }, + { + "epoch": 6.8050339592489015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17033 + }, + { + "epoch": 6.805433479824211, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17034 + }, + { + "epoch": 6.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17035 + }, + { + "epoch": 6.80623252097483, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17036 + }, + { + "epoch": 6.80663204155014, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17037 + }, + { + "epoch": 6.80703156212545, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17038 + }, + { + "epoch": 6.807431082700759, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17039 + }, + { + "epoch": 6.807830603276068, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17040 + }, + { + "epoch": 6.808230123851378, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17041 + }, + { + "epoch": 6.808629644426688, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17042 + }, + { + "epoch": 6.809029165001998, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17043 + }, + { + "epoch": 6.809428685577307, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17044 + }, + { + "epoch": 6.809828206152617, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17045 + }, + { + "epoch": 6.810227726727927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17046 + }, + { + "epoch": 6.810627247303236, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17047 + }, + { + "epoch": 6.811026767878546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17048 + }, + { + "epoch": 6.8114262884538554, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17049 + }, + { + "epoch": 6.811825809029165, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17050 + }, + { + "epoch": 6.812225329604475, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17051 + }, + { + "epoch": 6.812624850179784, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17052 + }, + { + "epoch": 6.813024370755094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17053 + }, + { + "epoch": 6.8134238913304035, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17054 + }, + { + "epoch": 6.813823411905713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 17055 + }, + { + "epoch": 6.814222932481023, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17056 + }, + { + "epoch": 6.814622453056332, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 17057 + }, + { + "epoch": 6.815021973631642, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17058 + }, + { + "epoch": 6.8154214942069515, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17059 + }, + { + "epoch": 6.815821014782261, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17060 + }, + { + "epoch": 6.816220535357571, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 17061 + }, + { + "epoch": 6.816620055932881, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 17062 + }, + { + "epoch": 6.81701957650819, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17063 + }, + { + "epoch": 6.8174190970835, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17064 + }, + { + "epoch": 6.817818617658809, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17065 + }, + { + "epoch": 6.818218138234119, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17066 + }, + { + "epoch": 6.818617658809429, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17067 + }, + { + "epoch": 6.819017179384739, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 17068 + }, + { + "epoch": 6.819416699960048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17069 + }, + { + "epoch": 6.819816220535357, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17070 + }, + { + "epoch": 6.820215741110667, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17071 + }, + { + "epoch": 6.820615261685977, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17072 + }, + { + "epoch": 6.821014782261287, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17073 + }, + { + "epoch": 6.821414302836596, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17074 + }, + { + "epoch": 6.8218138234119055, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 17075 + }, + { + "epoch": 6.822213343987215, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17076 + }, + { + "epoch": 6.822612864562525, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 17077 + }, + { + "epoch": 6.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17078 + }, + { + "epoch": 6.823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17079 + }, + { + "epoch": 6.8238114262884535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17080 + }, + { + "epoch": 6.824210946863763, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17081 + }, + { + "epoch": 6.824610467439073, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17082 + }, + { + "epoch": 6.825009988014383, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17083 + }, + { + "epoch": 6.825409508589693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17084 + }, + { + "epoch": 6.825809029165002, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17085 + }, + { + "epoch": 6.826208549740311, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17086 + }, + { + "epoch": 6.826608070315621, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17087 + }, + { + "epoch": 6.827007590890931, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17088 + }, + { + "epoch": 6.827407111466241, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17089 + }, + { + "epoch": 6.8278066320415505, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17090 + }, + { + "epoch": 6.828206152616859, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17091 + }, + { + "epoch": 6.828605673192169, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 17092 + }, + { + "epoch": 6.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17093 + }, + { + "epoch": 6.829404714342789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17094 + }, + { + "epoch": 6.8298042349180985, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17095 + }, + { + "epoch": 6.830203755493407, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17096 + }, + { + "epoch": 6.830603276068717, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17097 + }, + { + "epoch": 6.831002796644027, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17098 + }, + { + "epoch": 6.831402317219337, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17099 + }, + { + "epoch": 6.831801837794647, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17100 + }, + { + "epoch": 6.832201358369956, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17101 + }, + { + "epoch": 6.832600878945266, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17102 + }, + { + "epoch": 6.833000399520575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17103 + }, + { + "epoch": 6.833399920095885, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17104 + }, + { + "epoch": 6.833799440671195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17105 + }, + { + "epoch": 6.834198961246504, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17106 + }, + { + "epoch": 6.834598481821814, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17107 + }, + { + "epoch": 6.834998002397123, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17108 + }, + { + "epoch": 6.835397522972433, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17109 + }, + { + "epoch": 6.835797043547743, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17110 + }, + { + "epoch": 6.836196564123052, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17111 + }, + { + "epoch": 6.836596084698362, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17112 + }, + { + "epoch": 6.836995605273671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17113 + }, + { + "epoch": 6.837395125848981, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17114 + }, + { + "epoch": 6.837794646424291, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17115 + }, + { + "epoch": 6.8381941669996005, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17116 + }, + { + "epoch": 6.83859368757491, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17117 + }, + { + "epoch": 6.83899320815022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17118 + }, + { + "epoch": 6.83939272872553, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17119 + }, + { + "epoch": 6.839792249300839, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17120 + }, + { + "epoch": 6.8401917698761485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17121 + }, + { + "epoch": 6.840591290451458, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17122 + }, + { + "epoch": 6.840990811026768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17123 + }, + { + "epoch": 6.841390331602078, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17124 + }, + { + "epoch": 6.841789852177387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17125 + }, + { + "epoch": 6.842189372752697, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17126 + }, + { + "epoch": 6.842588893328006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17127 + }, + { + "epoch": 6.842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17128 + }, + { + "epoch": 6.843387934478626, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17129 + }, + { + "epoch": 6.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 17130 + }, + { + "epoch": 6.844186975629245, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17131 + }, + { + "epoch": 6.844586496204554, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17132 + }, + { + "epoch": 6.844986016779864, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17133 + }, + { + "epoch": 6.845385537355174, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17134 + }, + { + "epoch": 6.845785057930484, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17135 + }, + { + "epoch": 6.8461845785057935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17136 + }, + { + "epoch": 6.846584099081102, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17137 + }, + { + "epoch": 6.846983619656412, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17138 + }, + { + "epoch": 6.847383140231722, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17139 + }, + { + "epoch": 6.847782660807032, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17140 + }, + { + "epoch": 6.848182181382342, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17141 + }, + { + "epoch": 6.8485817019576505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17142 + }, + { + "epoch": 6.84898122253296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17143 + }, + { + "epoch": 6.84938074310827, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17144 + }, + { + "epoch": 6.84978026368358, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17145 + }, + { + "epoch": 6.85017978425889, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17146 + }, + { + "epoch": 6.8505793048341985, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17147 + }, + { + "epoch": 6.850978825409508, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17148 + }, + { + "epoch": 6.851378345984818, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17149 + }, + { + "epoch": 6.851777866560128, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17150 + }, + { + "epoch": 6.852177387135438, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 17151 + }, + { + "epoch": 6.8525769077107475, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.2944, + "step": 17152 + }, + { + "epoch": 6.852976428286056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17153 + }, + { + "epoch": 6.853375948861366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17154 + }, + { + "epoch": 6.853775469436676, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17155 + }, + { + "epoch": 6.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17156 + }, + { + "epoch": 6.8545745105872955, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17157 + }, + { + "epoch": 6.854974031162605, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17158 + }, + { + "epoch": 6.855373551737914, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17159 + }, + { + "epoch": 6.855773072313224, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17160 + }, + { + "epoch": 6.856172592888534, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17161 + }, + { + "epoch": 6.8565721134638435, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 17162 + }, + { + "epoch": 6.856971634039153, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17163 + }, + { + "epoch": 6.857371154614462, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17164 + }, + { + "epoch": 6.857770675189772, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17165 + }, + { + "epoch": 6.858170195765082, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17166 + }, + { + "epoch": 6.858569716340392, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17167 + }, + { + "epoch": 6.858969236915701, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17168 + }, + { + "epoch": 6.859368757491011, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17169 + }, + { + "epoch": 6.85976827806632, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17170 + }, + { + "epoch": 6.86016779864163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17171 + }, + { + "epoch": 6.86056731921694, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17172 + }, + { + "epoch": 6.860966839792249, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17173 + }, + { + "epoch": 6.861366360367559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17174 + }, + { + "epoch": 6.861765880942869, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17175 + }, + { + "epoch": 6.862165401518178, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17176 + }, + { + "epoch": 6.862564922093488, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17177 + }, + { + "epoch": 6.8629644426687975, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17178 + }, + { + "epoch": 6.863363963244107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17179 + }, + { + "epoch": 6.863763483819417, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17180 + }, + { + "epoch": 6.864163004394726, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17181 + }, + { + "epoch": 6.864562524970036, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17182 + }, + { + "epoch": 6.8649620455453455, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17183 + }, + { + "epoch": 6.865361566120655, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17184 + }, + { + "epoch": 6.865761086695965, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17185 + }, + { + "epoch": 6.866160607271274, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17186 + }, + { + "epoch": 6.866560127846584, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17187 + }, + { + "epoch": 6.8669596484218935, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17188 + }, + { + "epoch": 6.867359168997203, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17189 + }, + { + "epoch": 6.867758689572513, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17190 + }, + { + "epoch": 6.868158210147823, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17191 + }, + { + "epoch": 6.868557730723133, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17192 + }, + { + "epoch": 6.868957251298442, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17193 + }, + { + "epoch": 6.869356771873751, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17194 + }, + { + "epoch": 6.869756292449061, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17195 + }, + { + "epoch": 6.870155813024371, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17196 + }, + { + "epoch": 6.870555333599681, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17197 + }, + { + "epoch": 6.87095485417499, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17198 + }, + { + "epoch": 6.871354374750299, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17199 + }, + { + "epoch": 6.871753895325609, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17200 + }, + { + "epoch": 6.872153415900919, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17201 + }, + { + "epoch": 6.872552936476229, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 17202 + }, + { + "epoch": 6.872952457051538, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17203 + }, + { + "epoch": 6.8733519776268475, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17204 + }, + { + "epoch": 6.873751498202157, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17205 + }, + { + "epoch": 6.874151018777467, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17206 + }, + { + "epoch": 6.874550539352777, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 17207 + }, + { + "epoch": 6.874950059928087, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17208 + }, + { + "epoch": 6.875349580503396, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17209 + }, + { + "epoch": 6.875749101078705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17210 + }, + { + "epoch": 6.876148621654015, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17211 + }, + { + "epoch": 6.876548142229325, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17212 + }, + { + "epoch": 6.876947662804635, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17213 + }, + { + "epoch": 6.877347183379944, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17214 + }, + { + "epoch": 6.877746703955253, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17215 + }, + { + "epoch": 6.878146224530563, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17216 + }, + { + "epoch": 6.878545745105873, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17217 + }, + { + "epoch": 6.878945265681183, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17218 + }, + { + "epoch": 6.8793447862564925, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17219 + }, + { + "epoch": 6.879744306831801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17220 + }, + { + "epoch": 6.880143827407111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17221 + }, + { + "epoch": 6.880543347982421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17222 + }, + { + "epoch": 6.880942868557731, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17223 + }, + { + "epoch": 6.8813423891330405, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17224 + }, + { + "epoch": 6.88174190970835, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17225 + }, + { + "epoch": 6.88214143028366, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17226 + }, + { + "epoch": 6.882540950858969, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17227 + }, + { + "epoch": 6.882940471434279, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17228 + }, + { + "epoch": 6.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17229 + }, + { + "epoch": 6.883739512584898, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17230 + }, + { + "epoch": 6.884139033160208, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17231 + }, + { + "epoch": 6.884538553735517, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17232 + }, + { + "epoch": 6.884938074310827, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17233 + }, + { + "epoch": 6.885337594886137, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17234 + }, + { + "epoch": 6.885737115461446, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17235 + }, + { + "epoch": 6.886136636036756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17236 + }, + { + "epoch": 6.886536156612065, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17237 + }, + { + "epoch": 6.886935677187375, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17238 + }, + { + "epoch": 6.887335197762685, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17239 + }, + { + "epoch": 6.887734718337994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17240 + }, + { + "epoch": 6.888134238913304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17241 + }, + { + "epoch": 6.888533759488614, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17242 + }, + { + "epoch": 6.888933280063923, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17243 + }, + { + "epoch": 6.889332800639233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17244 + }, + { + "epoch": 6.8897323212145425, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17245 + }, + { + "epoch": 6.890131841789852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17246 + }, + { + "epoch": 6.890531362365162, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17247 + }, + { + "epoch": 6.890930882940472, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17248 + }, + { + "epoch": 6.891330403515781, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17249 + }, + { + "epoch": 6.8917299240910905, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17250 + }, + { + "epoch": 6.8921294446664, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17251 + }, + { + "epoch": 6.89252896524171, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 17252 + }, + { + "epoch": 6.89292848581702, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17253 + }, + { + "epoch": 6.893328006392329, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17254 + }, + { + "epoch": 6.893727526967639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17255 + }, + { + "epoch": 6.894127047542948, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17256 + }, + { + "epoch": 6.894526568118258, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17257 + }, + { + "epoch": 6.894926088693568, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17258 + }, + { + "epoch": 6.895325609268878, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17259 + }, + { + "epoch": 6.895725129844187, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17260 + }, + { + "epoch": 6.896124650419496, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17261 + }, + { + "epoch": 6.896524170994806, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17262 + }, + { + "epoch": 6.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17263 + }, + { + "epoch": 6.897323212145426, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17264 + }, + { + "epoch": 6.8977227327207356, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17265 + }, + { + "epoch": 6.8981222532960444, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17266 + }, + { + "epoch": 6.898521773871354, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17267 + }, + { + "epoch": 6.898921294446664, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17268 + }, + { + "epoch": 6.899320815021974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17269 + }, + { + "epoch": 6.899720335597284, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17270 + }, + { + "epoch": 6.9001198561725925, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17271 + }, + { + "epoch": 6.900519376747902, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17272 + }, + { + "epoch": 6.900918897323212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17273 + }, + { + "epoch": 6.901318417898522, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17274 + }, + { + "epoch": 6.901717938473832, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17275 + }, + { + "epoch": 6.9021174590491405, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.2994, + "step": 17276 + }, + { + "epoch": 6.90251697962445, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17277 + }, + { + "epoch": 6.90291650019976, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17278 + }, + { + "epoch": 6.90331602077507, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17279 + }, + { + "epoch": 6.90371554135038, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17280 + }, + { + "epoch": 6.9041150619256895, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17281 + }, + { + "epoch": 6.904514582500999, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17282 + }, + { + "epoch": 6.904914103076308, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17283 + }, + { + "epoch": 6.905313623651618, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17284 + }, + { + "epoch": 6.905713144226928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17285 + }, + { + "epoch": 6.9061126648022375, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17286 + }, + { + "epoch": 6.906512185377547, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17287 + }, + { + "epoch": 6.906911705952856, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17288 + }, + { + "epoch": 6.907311226528166, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17289 + }, + { + "epoch": 6.907710747103476, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17290 + }, + { + "epoch": 6.9081102676787856, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17291 + }, + { + "epoch": 6.908509788254095, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17292 + }, + { + "epoch": 6.908909308829404, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17293 + }, + { + "epoch": 6.909308829404714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17294 + }, + { + "epoch": 6.909708349980024, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17295 + }, + { + "epoch": 6.910107870555334, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17296 + }, + { + "epoch": 6.910507391130643, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17297 + }, + { + "epoch": 6.910906911705953, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17298 + }, + { + "epoch": 6.911306432281263, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17299 + }, + { + "epoch": 6.911705952856572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17300 + }, + { + "epoch": 6.912105473431882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17301 + }, + { + "epoch": 6.912504994007191, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17302 + }, + { + "epoch": 6.912904514582501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17303 + }, + { + "epoch": 6.913304035157811, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17304 + }, + { + "epoch": 6.91370355573312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17305 + }, + { + "epoch": 6.91410307630843, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17306 + }, + { + "epoch": 6.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17307 + }, + { + "epoch": 6.914902117459049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 17308 + }, + { + "epoch": 6.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17309 + }, + { + "epoch": 6.915701158609668, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17310 + }, + { + "epoch": 6.916100679184978, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17311 + }, + { + "epoch": 6.9165001997602875, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17312 + }, + { + "epoch": 6.916899720335597, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17313 + }, + { + "epoch": 6.917299240910907, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17314 + }, + { + "epoch": 6.917698761486217, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 17315 + }, + { + "epoch": 6.918098282061527, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17316 + }, + { + "epoch": 6.918497802636836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17317 + }, + { + "epoch": 6.918897323212145, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17318 + }, + { + "epoch": 6.919296843787455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17319 + }, + { + "epoch": 6.919696364362765, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17320 + }, + { + "epoch": 6.920095884938075, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17321 + }, + { + "epoch": 6.920495405513384, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17322 + }, + { + "epoch": 6.920894926088693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17323 + }, + { + "epoch": 6.921294446664003, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17324 + }, + { + "epoch": 6.921693967239313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 17325 + }, + { + "epoch": 6.922093487814623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17326 + }, + { + "epoch": 6.922493008389932, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17327 + }, + { + "epoch": 6.922892528965241, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17328 + }, + { + "epoch": 6.923292049540551, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17329 + }, + { + "epoch": 6.923691570115861, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 17330 + }, + { + "epoch": 6.924091090691171, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17331 + }, + { + "epoch": 6.924490611266481, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17332 + }, + { + "epoch": 6.9248901318417895, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17333 + }, + { + "epoch": 6.925289652417099, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17334 + }, + { + "epoch": 6.925689172992409, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17335 + }, + { + "epoch": 6.926088693567719, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17336 + }, + { + "epoch": 6.926488214143029, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17337 + }, + { + "epoch": 6.926887734718338, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17338 + }, + { + "epoch": 6.927287255293647, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17339 + }, + { + "epoch": 6.927686775868957, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 17340 + }, + { + "epoch": 6.928086296444267, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17341 + }, + { + "epoch": 6.928485817019577, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17342 + }, + { + "epoch": 6.9288853375948865, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17343 + }, + { + "epoch": 6.929284858170195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17344 + }, + { + "epoch": 6.929684378745505, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17345 + }, + { + "epoch": 6.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 17346 + }, + { + "epoch": 6.930483419896125, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17347 + }, + { + "epoch": 6.9308829404714345, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 17348 + }, + { + "epoch": 6.931282461046744, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17349 + }, + { + "epoch": 6.931681981622053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17350 + }, + { + "epoch": 6.932081502197363, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 17351 + }, + { + "epoch": 6.932481022772673, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17352 + }, + { + "epoch": 6.9328805433479825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17353 + }, + { + "epoch": 6.933280063923292, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17354 + }, + { + "epoch": 6.933679584498602, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17355 + }, + { + "epoch": 6.934079105073911, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17356 + }, + { + "epoch": 6.934478625649221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17357 + }, + { + "epoch": 6.934878146224531, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17358 + }, + { + "epoch": 6.93527766679984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17359 + }, + { + "epoch": 6.93567718737515, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17360 + }, + { + "epoch": 6.936076707950459, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17361 + }, + { + "epoch": 6.936476228525769, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 17362 + }, + { + "epoch": 6.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17363 + }, + { + "epoch": 6.937275269676388, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17364 + }, + { + "epoch": 6.937674790251698, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17365 + }, + { + "epoch": 6.938074310827007, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17366 + }, + { + "epoch": 6.938473831402317, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17367 + }, + { + "epoch": 6.938873351977627, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 17368 + }, + { + "epoch": 6.9392728725529365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17369 + }, + { + "epoch": 6.939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17370 + }, + { + "epoch": 6.940071913703556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17371 + }, + { + "epoch": 6.940471434278866, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17372 + }, + { + "epoch": 6.940870954854175, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17373 + }, + { + "epoch": 6.9412704754294845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17374 + }, + { + "epoch": 6.941669996004794, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17375 + }, + { + "epoch": 6.942069516580104, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17376 + }, + { + "epoch": 6.942469037155414, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17377 + }, + { + "epoch": 6.942868557730723, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17378 + }, + { + "epoch": 6.9432680783060325, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17379 + }, + { + "epoch": 6.943667598881342, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 17380 + }, + { + "epoch": 6.944067119456652, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17381 + }, + { + "epoch": 6.944466640031962, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17382 + }, + { + "epoch": 6.944866160607271, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17383 + }, + { + "epoch": 6.945265681182581, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 17384 + }, + { + "epoch": 6.94566520175789, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17385 + }, + { + "epoch": 6.9460647223332, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17386 + }, + { + "epoch": 6.94646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17387 + }, + { + "epoch": 6.94686376348382, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 17388 + }, + { + "epoch": 6.9472632840591295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17389 + }, + { + "epoch": 6.947662804634438, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17390 + }, + { + "epoch": 6.948062325209748, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17391 + }, + { + "epoch": 6.948461845785058, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17392 + }, + { + "epoch": 6.948861366360368, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17393 + }, + { + "epoch": 6.949260886935678, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17394 + }, + { + "epoch": 6.9496604075109865, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17395 + }, + { + "epoch": 6.950059928086296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 17396 + }, + { + "epoch": 6.950459448661606, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 17397 + }, + { + "epoch": 6.950858969236916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17398 + }, + { + "epoch": 6.951258489812226, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17399 + }, + { + "epoch": 6.9516580103875345, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17400 + }, + { + "epoch": 6.952057530962844, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17401 + }, + { + "epoch": 6.952457051538154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17402 + }, + { + "epoch": 6.952856572113464, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17403 + }, + { + "epoch": 6.953256092688774, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17404 + }, + { + "epoch": 6.953655613264083, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17405 + }, + { + "epoch": 6.954055133839393, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17406 + }, + { + "epoch": 6.954454654414702, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17407 + }, + { + "epoch": 6.954854174990012, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17408 + }, + { + "epoch": 6.955253695565322, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17409 + }, + { + "epoch": 6.9556532161406315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17410 + }, + { + "epoch": 6.956052736715941, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 17411 + }, + { + "epoch": 6.95645225729125, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17412 + }, + { + "epoch": 6.95685177786656, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17413 + }, + { + "epoch": 6.95725129844187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17414 + }, + { + "epoch": 6.9576508190171795, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17415 + }, + { + "epoch": 6.958050339592489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17416 + }, + { + "epoch": 6.958449860167798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17417 + }, + { + "epoch": 6.958849380743108, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17418 + }, + { + "epoch": 6.959248901318418, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17419 + }, + { + "epoch": 6.959648421893728, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17420 + }, + { + "epoch": 6.960047942469037, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17421 + }, + { + "epoch": 6.960447463044347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17422 + }, + { + "epoch": 6.960846983619656, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17423 + }, + { + "epoch": 6.961246504194966, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17424 + }, + { + "epoch": 6.961646024770276, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17425 + }, + { + "epoch": 6.962045545345585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17426 + }, + { + "epoch": 6.962445065920895, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17427 + }, + { + "epoch": 6.962844586496205, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17428 + }, + { + "epoch": 6.963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17429 + }, + { + "epoch": 6.963643627646824, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17430 + }, + { + "epoch": 6.964043148222133, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17431 + }, + { + "epoch": 6.964442668797443, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17432 + }, + { + "epoch": 6.964842189372753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17433 + }, + { + "epoch": 6.965241709948062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17434 + }, + { + "epoch": 6.965641230523372, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17435 + }, + { + "epoch": 6.9660407510986815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17436 + }, + { + "epoch": 6.966440271673991, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17437 + }, + { + "epoch": 6.966839792249301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17438 + }, + { + "epoch": 6.967239312824611, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 17439 + }, + { + "epoch": 6.96763883339992, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17440 + }, + { + "epoch": 6.9680383539752295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17441 + }, + { + "epoch": 6.968437874550539, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17442 + }, + { + "epoch": 6.968837395125849, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17443 + }, + { + "epoch": 6.969236915701159, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17444 + }, + { + "epoch": 6.969636436276469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17445 + }, + { + "epoch": 6.970035956851778, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17446 + }, + { + "epoch": 6.970435477427087, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17447 + }, + { + "epoch": 6.970834998002397, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17448 + }, + { + "epoch": 6.971234518577707, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17449 + }, + { + "epoch": 6.971634039153017, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17450 + }, + { + "epoch": 6.972033559728326, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17451 + }, + { + "epoch": 6.972433080303635, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17452 + }, + { + "epoch": 6.972832600878945, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17453 + }, + { + "epoch": 6.973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17454 + }, + { + "epoch": 6.973631642029565, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17455 + }, + { + "epoch": 6.974031162604874, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17456 + }, + { + "epoch": 6.974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17457 + }, + { + "epoch": 6.974830203755493, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17458 + }, + { + "epoch": 6.975229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17459 + }, + { + "epoch": 6.975629244906113, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17460 + }, + { + "epoch": 6.976028765481423, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17461 + }, + { + "epoch": 6.976428286056732, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 17462 + }, + { + "epoch": 6.976827806632041, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17463 + }, + { + "epoch": 6.977227327207351, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17464 + }, + { + "epoch": 6.977626847782661, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17465 + }, + { + "epoch": 6.978026368357971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17466 + }, + { + "epoch": 6.97842588893328, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17467 + }, + { + "epoch": 6.978825409508589, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17468 + }, + { + "epoch": 6.979224930083899, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17469 + }, + { + "epoch": 6.979624450659209, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17470 + }, + { + "epoch": 6.980023971234519, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17471 + }, + { + "epoch": 6.9804234918098285, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17472 + }, + { + "epoch": 6.980823012385137, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17473 + }, + { + "epoch": 6.981222532960447, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17474 + }, + { + "epoch": 6.981622053535757, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17475 + }, + { + "epoch": 6.982021574111067, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17476 + }, + { + "epoch": 6.9824210946863765, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17477 + }, + { + "epoch": 6.982820615261686, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17478 + }, + { + "epoch": 6.983220135836996, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17479 + }, + { + "epoch": 6.983619656412305, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17480 + }, + { + "epoch": 6.984019176987615, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17481 + }, + { + "epoch": 6.9844186975629245, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17482 + }, + { + "epoch": 6.984818218138234, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17483 + }, + { + "epoch": 6.985217738713544, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17484 + }, + { + "epoch": 6.985617259288853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17485 + }, + { + "epoch": 6.986016779864163, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17486 + }, + { + "epoch": 6.986416300439473, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17487 + }, + { + "epoch": 6.986815821014782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17488 + }, + { + "epoch": 6.987215341590092, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17489 + }, + { + "epoch": 6.987614862165401, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17490 + }, + { + "epoch": 6.988014382740711, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17491 + }, + { + "epoch": 6.988413903316021, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17492 + }, + { + "epoch": 6.98881342389133, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17493 + }, + { + "epoch": 6.98921294446664, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17494 + }, + { + "epoch": 6.98961246504195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17495 + }, + { + "epoch": 6.99001198561726, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17496 + }, + { + "epoch": 6.990411506192569, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17497 + }, + { + "epoch": 6.9908110267678785, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17498 + }, + { + "epoch": 6.991210547343188, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17499 + }, + { + "epoch": 6.991610067918498, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17500 + }, + { + "epoch": 6.992009588493808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17501 + }, + { + "epoch": 6.992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 17502 + }, + { + "epoch": 6.9928086296444265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17503 + }, + { + "epoch": 6.993208150219736, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17504 + }, + { + "epoch": 6.993607670795046, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 17505 + }, + { + "epoch": 6.994007191370356, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17506 + }, + { + "epoch": 6.994406711945665, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17507 + }, + { + "epoch": 6.9948062325209746, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17508 + }, + { + "epoch": 6.995205753096284, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17509 + }, + { + "epoch": 6.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17510 + }, + { + "epoch": 6.996004794246904, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17511 + }, + { + "epoch": 6.996404314822214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17512 + }, + { + "epoch": 6.996803835397523, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 17513 + }, + { + "epoch": 6.997203355972832, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17514 + }, + { + "epoch": 6.997602876548142, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17515 + }, + { + "epoch": 6.998002397123452, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17516 + }, + { + "epoch": 6.998401917698762, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 17517 + }, + { + "epoch": 6.9988014382740715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17518 + }, + { + "epoch": 6.99920095884938, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 17519 + }, + { + "epoch": 6.99960047942469, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17520 + }, + { + "epoch": 7.0, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17521 + }, + { + "epoch": 7.00039952057531, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17522 + }, + { + "epoch": 7.00079904115062, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17523 + }, + { + "epoch": 7.0011985617259285, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17524 + }, + { + "epoch": 7.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17525 + }, + { + "epoch": 7.001997602876548, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17526 + }, + { + "epoch": 7.002397123451858, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17527 + }, + { + "epoch": 7.002796644027168, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17528 + }, + { + "epoch": 7.003196164602477, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17529 + }, + { + "epoch": 7.003595685177786, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17530 + }, + { + "epoch": 7.003995205753096, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17531 + }, + { + "epoch": 7.004394726328406, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17532 + }, + { + "epoch": 7.004794246903716, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17533 + }, + { + "epoch": 7.0051937674790254, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17534 + }, + { + "epoch": 7.005593288054335, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17535 + }, + { + "epoch": 7.005992808629644, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17536 + }, + { + "epoch": 7.006392329204954, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17537 + }, + { + "epoch": 7.006791849780264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17538 + }, + { + "epoch": 7.0071913703555735, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17539 + }, + { + "epoch": 7.007590890930883, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17540 + }, + { + "epoch": 7.007990411506192, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17541 + }, + { + "epoch": 7.008389932081502, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17542 + }, + { + "epoch": 7.008789452656812, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17543 + }, + { + "epoch": 7.0091889732321215, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17544 + }, + { + "epoch": 7.009588493807431, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17545 + }, + { + "epoch": 7.009988014382741, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17546 + }, + { + "epoch": 7.01038753495805, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17547 + }, + { + "epoch": 7.01078705553336, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17548 + }, + { + "epoch": 7.01118657610867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 17549 + }, + { + "epoch": 7.011586096683979, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17550 + }, + { + "epoch": 7.011985617259289, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17551 + }, + { + "epoch": 7.012385137834598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17552 + }, + { + "epoch": 7.012784658409908, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17553 + }, + { + "epoch": 7.013184178985218, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17554 + }, + { + "epoch": 7.013583699560527, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17555 + }, + { + "epoch": 7.013983220135837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17556 + }, + { + "epoch": 7.014382740711147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17557 + }, + { + "epoch": 7.014782261286456, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17558 + }, + { + "epoch": 7.015181781861766, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17559 + }, + { + "epoch": 7.0155813024370755, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17560 + }, + { + "epoch": 7.015980823012385, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17561 + }, + { + "epoch": 7.016380343587695, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17562 + }, + { + "epoch": 7.016779864163005, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17563 + }, + { + "epoch": 7.017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17564 + }, + { + "epoch": 7.0175789053136235, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 17565 + }, + { + "epoch": 7.017978425888933, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17566 + }, + { + "epoch": 7.018377946464243, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 17567 + }, + { + "epoch": 7.018777467039553, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17568 + }, + { + "epoch": 7.019176987614862, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17569 + }, + { + "epoch": 7.0195765081901715, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17570 + }, + { + "epoch": 7.019976028765481, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17571 + }, + { + "epoch": 7.020375549340791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17572 + }, + { + "epoch": 7.020775069916101, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17573 + }, + { + "epoch": 7.021174590491411, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17574 + }, + { + "epoch": 7.02157411106672, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 17575 + }, + { + "epoch": 7.021973631642029, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17576 + }, + { + "epoch": 7.022373152217339, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17577 + }, + { + "epoch": 7.022772672792649, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17578 + }, + { + "epoch": 7.023172193367959, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17579 + }, + { + "epoch": 7.0235717139432685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 17580 + }, + { + "epoch": 7.023971234518577, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17581 + }, + { + "epoch": 7.024370755093887, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17582 + }, + { + "epoch": 7.024770275669197, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17583 + }, + { + "epoch": 7.025169796244507, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17584 + }, + { + "epoch": 7.025569316819817, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17585 + }, + { + "epoch": 7.0259688373951255, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17586 + }, + { + "epoch": 7.026368357970435, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17587 + }, + { + "epoch": 7.026767878545745, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17588 + }, + { + "epoch": 7.027167399121055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17589 + }, + { + "epoch": 7.027566919696365, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 17590 + }, + { + "epoch": 7.027966440271674, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17591 + }, + { + "epoch": 7.028365960846983, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17592 + }, + { + "epoch": 7.028765481422293, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17593 + }, + { + "epoch": 7.029165001997603, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17594 + }, + { + "epoch": 7.029564522572913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17595 + }, + { + "epoch": 7.029964043148222, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17596 + }, + { + "epoch": 7.030363563723531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17597 + }, + { + "epoch": 7.030763084298841, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 17598 + }, + { + "epoch": 7.031162604874151, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17599 + }, + { + "epoch": 7.031562125449461, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17600 + }, + { + "epoch": 7.0319616460247705, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 17601 + }, + { + "epoch": 7.03236116660008, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17602 + }, + { + "epoch": 7.032760687175389, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17603 + }, + { + "epoch": 7.033160207750699, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17604 + }, + { + "epoch": 7.033559728326009, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17605 + }, + { + "epoch": 7.0339592489013185, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17606 + }, + { + "epoch": 7.034358769476628, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17607 + }, + { + "epoch": 7.034758290051938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17608 + }, + { + "epoch": 7.035157810627247, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17609 + }, + { + "epoch": 7.035557331202557, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17610 + }, + { + "epoch": 7.035956851777867, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17611 + }, + { + "epoch": 7.036356372353176, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17612 + }, + { + "epoch": 7.036755892928486, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17613 + }, + { + "epoch": 7.037155413503795, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17614 + }, + { + "epoch": 7.037554934079105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17615 + }, + { + "epoch": 7.037954454654415, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 17616 + }, + { + "epoch": 7.038353975229724, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17617 + }, + { + "epoch": 7.038753495805034, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17618 + }, + { + "epoch": 7.039153016380344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17619 + }, + { + "epoch": 7.039552536955653, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17620 + }, + { + "epoch": 7.039952057530963, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17621 + }, + { + "epoch": 7.040351578106272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 17622 + }, + { + "epoch": 7.040751098681582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17623 + }, + { + "epoch": 7.041150619256892, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17624 + }, + { + "epoch": 7.041550139832202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 17625 + }, + { + "epoch": 7.041949660407511, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17626 + }, + { + "epoch": 7.0423491809828205, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17627 + }, + { + "epoch": 7.04274870155813, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17628 + }, + { + "epoch": 7.04314822213344, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17629 + }, + { + "epoch": 7.04354774270875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17630 + }, + { + "epoch": 7.043947263284059, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17631 + }, + { + "epoch": 7.0443467838593685, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17632 + }, + { + "epoch": 7.044746304434678, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17633 + }, + { + "epoch": 7.045145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17634 + }, + { + "epoch": 7.045545345585298, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17635 + }, + { + "epoch": 7.045944866160608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17636 + }, + { + "epoch": 7.046344386735917, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.2991, + "step": 17637 + }, + { + "epoch": 7.046743907311226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17638 + }, + { + "epoch": 7.047143427886536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17639 + }, + { + "epoch": 7.047542948461846, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17640 + }, + { + "epoch": 7.047942469037156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17641 + }, + { + "epoch": 7.048341989612465, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17642 + }, + { + "epoch": 7.048741510187774, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17643 + }, + { + "epoch": 7.049141030763084, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17644 + }, + { + "epoch": 7.049540551338394, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17645 + }, + { + "epoch": 7.049940071913704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17646 + }, + { + "epoch": 7.0503395924890135, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17647 + }, + { + "epoch": 7.050739113064322, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17648 + }, + { + "epoch": 7.051138633639632, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17649 + }, + { + "epoch": 7.051538154214942, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17650 + }, + { + "epoch": 7.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17651 + }, + { + "epoch": 7.052337195365562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17652 + }, + { + "epoch": 7.052736715940871, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17653 + }, + { + "epoch": 7.05313623651618, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17654 + }, + { + "epoch": 7.05353575709149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17655 + }, + { + "epoch": 7.0539352776668, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17656 + }, + { + "epoch": 7.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17657 + }, + { + "epoch": 7.054734318817419, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17658 + }, + { + "epoch": 7.055133839392728, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17659 + }, + { + "epoch": 7.055533359968038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17660 + }, + { + "epoch": 7.055932880543348, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17661 + }, + { + "epoch": 7.056332401118658, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17662 + }, + { + "epoch": 7.0567319216939675, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17663 + }, + { + "epoch": 7.057131442269277, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17664 + }, + { + "epoch": 7.057530962844586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17665 + }, + { + "epoch": 7.057930483419896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17666 + }, + { + "epoch": 7.058330003995206, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17667 + }, + { + "epoch": 7.0587295245705155, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17668 + }, + { + "epoch": 7.059129045145825, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17669 + }, + { + "epoch": 7.059528565721135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17670 + }, + { + "epoch": 7.059928086296444, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17671 + }, + { + "epoch": 7.060327606871754, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17672 + }, + { + "epoch": 7.0607271274470635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17673 + }, + { + "epoch": 7.061126648022373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17674 + }, + { + "epoch": 7.061526168597683, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17675 + }, + { + "epoch": 7.061925689172992, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17676 + }, + { + "epoch": 7.062325209748302, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17677 + }, + { + "epoch": 7.062724730323612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17678 + }, + { + "epoch": 7.063124250898921, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17679 + }, + { + "epoch": 7.063523771474231, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17680 + }, + { + "epoch": 7.063923292049541, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17681 + }, + { + "epoch": 7.06432281262485, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 17682 + }, + { + "epoch": 7.06472233320016, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17683 + }, + { + "epoch": 7.065121853775469, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17684 + }, + { + "epoch": 7.065521374350779, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17685 + }, + { + "epoch": 7.065920894926089, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17686 + }, + { + "epoch": 7.066320415501398, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17687 + }, + { + "epoch": 7.066719936076708, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17688 + }, + { + "epoch": 7.0671194566520175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 17689 + }, + { + "epoch": 7.067518977227327, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17690 + }, + { + "epoch": 7.067918497802637, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17691 + }, + { + "epoch": 7.068318018377947, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17692 + }, + { + "epoch": 7.068717538953256, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17693 + }, + { + "epoch": 7.0691170595285655, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17694 + }, + { + "epoch": 7.069516580103875, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17695 + }, + { + "epoch": 7.069916100679185, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17696 + }, + { + "epoch": 7.070315621254495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17697 + }, + { + "epoch": 7.070715141829805, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17698 + }, + { + "epoch": 7.0711146624051135, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 17699 + }, + { + "epoch": 7.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17700 + }, + { + "epoch": 7.071913703555733, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17701 + }, + { + "epoch": 7.072313224131043, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 17702 + }, + { + "epoch": 7.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17703 + }, + { + "epoch": 7.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17704 + }, + { + "epoch": 7.073511785856971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17705 + }, + { + "epoch": 7.073911306432281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17706 + }, + { + "epoch": 7.074310827007591, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17707 + }, + { + "epoch": 7.074710347582901, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17708 + }, + { + "epoch": 7.0751098681582105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17709 + }, + { + "epoch": 7.075509388733519, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17710 + }, + { + "epoch": 7.075908909308829, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17711 + }, + { + "epoch": 7.076308429884139, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17712 + }, + { + "epoch": 7.076707950459449, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17713 + }, + { + "epoch": 7.077107471034759, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17714 + }, + { + "epoch": 7.077506991610068, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17715 + }, + { + "epoch": 7.077906512185377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17716 + }, + { + "epoch": 7.078306032760687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17717 + }, + { + "epoch": 7.078705553335997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17718 + }, + { + "epoch": 7.079105073911307, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17719 + }, + { + "epoch": 7.079504594486616, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17720 + }, + { + "epoch": 7.079904115061925, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17721 + }, + { + "epoch": 7.080303635637235, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17722 + }, + { + "epoch": 7.080703156212545, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17723 + }, + { + "epoch": 7.081102676787855, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17724 + }, + { + "epoch": 7.081502197363164, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17725 + }, + { + "epoch": 7.081901717938474, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17726 + }, + { + "epoch": 7.082301238513783, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17727 + }, + { + "epoch": 7.082700759089093, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17728 + }, + { + "epoch": 7.083100279664403, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17729 + }, + { + "epoch": 7.0834998002397125, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17730 + }, + { + "epoch": 7.083899320815022, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 17731 + }, + { + "epoch": 7.084298841390331, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17732 + }, + { + "epoch": 7.084698361965641, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17733 + }, + { + "epoch": 7.085097882540951, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17734 + }, + { + "epoch": 7.0854974031162605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17735 + }, + { + "epoch": 7.08589692369157, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17736 + }, + { + "epoch": 7.08629644426688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17737 + }, + { + "epoch": 7.086695964842189, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17738 + }, + { + "epoch": 7.087095485417499, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17739 + }, + { + "epoch": 7.087495005992809, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17740 + }, + { + "epoch": 7.087894526568118, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17741 + }, + { + "epoch": 7.088294047143428, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17742 + }, + { + "epoch": 7.088693567718738, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17743 + }, + { + "epoch": 7.089093088294047, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17744 + }, + { + "epoch": 7.089492608869357, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17745 + }, + { + "epoch": 7.089892129444666, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17746 + }, + { + "epoch": 7.090291650019976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17747 + }, + { + "epoch": 7.090691170595286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17748 + }, + { + "epoch": 7.091090691170595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 17749 + }, + { + "epoch": 7.091490211745905, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17750 + }, + { + "epoch": 7.0918897323212144, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17751 + }, + { + "epoch": 7.092289252896524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17752 + }, + { + "epoch": 7.092688773471834, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17753 + }, + { + "epoch": 7.093088294047144, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17754 + }, + { + "epoch": 7.093487814622453, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17755 + }, + { + "epoch": 7.0938873351977625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17756 + }, + { + "epoch": 7.094286855773072, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17757 + }, + { + "epoch": 7.094686376348382, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17758 + }, + { + "epoch": 7.095085896923692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17759 + }, + { + "epoch": 7.095485417499002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17760 + }, + { + "epoch": 7.0958849380743105, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17761 + }, + { + "epoch": 7.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17762 + }, + { + "epoch": 7.09668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17763 + }, + { + "epoch": 7.09708349980024, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17764 + }, + { + "epoch": 7.09748302037555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17765 + }, + { + "epoch": 7.097882540950859, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17766 + }, + { + "epoch": 7.098282061526168, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17767 + }, + { + "epoch": 7.098681582101478, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17768 + }, + { + "epoch": 7.099081102676788, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17769 + }, + { + "epoch": 7.099480623252098, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17770 + }, + { + "epoch": 7.0998801438274075, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17771 + }, + { + "epoch": 7.100279664402716, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17772 + }, + { + "epoch": 7.100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17773 + }, + { + "epoch": 7.101078705553336, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17774 + }, + { + "epoch": 7.101478226128646, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17775 + }, + { + "epoch": 7.1018777467039556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17776 + }, + { + "epoch": 7.1022772672792644, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17777 + }, + { + "epoch": 7.102676787854574, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17778 + }, + { + "epoch": 7.103076308429884, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17779 + }, + { + "epoch": 7.103475829005194, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17780 + }, + { + "epoch": 7.103875349580504, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17781 + }, + { + "epoch": 7.104274870155813, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17782 + }, + { + "epoch": 7.104674390731122, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17783 + }, + { + "epoch": 7.105073911306432, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17784 + }, + { + "epoch": 7.105473431881742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 17785 + }, + { + "epoch": 7.105872952457052, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17786 + }, + { + "epoch": 7.106272473032361, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17787 + }, + { + "epoch": 7.106671993607671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17788 + }, + { + "epoch": 7.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17789 + }, + { + "epoch": 7.10747103475829, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17790 + }, + { + "epoch": 7.1078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17791 + }, + { + "epoch": 7.1082700759089095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17792 + }, + { + "epoch": 7.108669596484219, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17793 + }, + { + "epoch": 7.109069117059528, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17794 + }, + { + "epoch": 7.109468637634838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17795 + }, + { + "epoch": 7.109868158210148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17796 + }, + { + "epoch": 7.1102676787854575, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17797 + }, + { + "epoch": 7.110667199360767, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17798 + }, + { + "epoch": 7.111066719936077, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17799 + }, + { + "epoch": 7.111466240511386, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17800 + }, + { + "epoch": 7.111865761086696, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17801 + }, + { + "epoch": 7.112265281662006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 17802 + }, + { + "epoch": 7.112664802237315, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17803 + }, + { + "epoch": 7.113064322812625, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17804 + }, + { + "epoch": 7.113463843387935, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17805 + }, + { + "epoch": 7.113863363963244, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17806 + }, + { + "epoch": 7.114262884538554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17807 + }, + { + "epoch": 7.114662405113863, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17808 + }, + { + "epoch": 7.115061925689173, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17809 + }, + { + "epoch": 7.115461446264483, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17810 + }, + { + "epoch": 7.115860966839792, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17811 + }, + { + "epoch": 7.116260487415102, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 17812 + }, + { + "epoch": 7.116660007990411, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17813 + }, + { + "epoch": 7.117059528565721, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17814 + }, + { + "epoch": 7.117459049141031, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17815 + }, + { + "epoch": 7.117858569716341, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17816 + }, + { + "epoch": 7.11825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17817 + }, + { + "epoch": 7.1186576108669595, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 17818 + }, + { + "epoch": 7.119057131442269, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17819 + }, + { + "epoch": 7.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17820 + }, + { + "epoch": 7.119856172592889, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17821 + }, + { + "epoch": 7.120255693168198, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17822 + }, + { + "epoch": 7.1206552137435075, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17823 + }, + { + "epoch": 7.121054734318817, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17824 + }, + { + "epoch": 7.121454254894127, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17825 + }, + { + "epoch": 7.121853775469437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17826 + }, + { + "epoch": 7.122253296044747, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17827 + }, + { + "epoch": 7.122652816620056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17828 + }, + { + "epoch": 7.123052337195365, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 17829 + }, + { + "epoch": 7.123451857770675, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 17830 + }, + { + "epoch": 7.123851378345985, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17831 + }, + { + "epoch": 7.124250898921295, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17832 + }, + { + "epoch": 7.1246504194966045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17833 + }, + { + "epoch": 7.125049940071913, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17834 + }, + { + "epoch": 7.125449460647223, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17835 + }, + { + "epoch": 7.125848981222533, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17836 + }, + { + "epoch": 7.126248501797843, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17837 + }, + { + "epoch": 7.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17838 + }, + { + "epoch": 7.127047542948461, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17839 + }, + { + "epoch": 7.127447063523771, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17840 + }, + { + "epoch": 7.127846584099081, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17841 + }, + { + "epoch": 7.128246104674391, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17842 + }, + { + "epoch": 7.128645625249701, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17843 + }, + { + "epoch": 7.12904514582501, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17844 + }, + { + "epoch": 7.129444666400319, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 17845 + }, + { + "epoch": 7.129844186975629, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17846 + }, + { + "epoch": 7.130243707550939, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17847 + }, + { + "epoch": 7.130643228126249, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17848 + }, + { + "epoch": 7.131042748701558, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17849 + }, + { + "epoch": 7.131442269276867, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17850 + }, + { + "epoch": 7.131841789852177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17851 + }, + { + "epoch": 7.132241310427487, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17852 + }, + { + "epoch": 7.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17853 + }, + { + "epoch": 7.1330403515781065, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17854 + }, + { + "epoch": 7.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17855 + }, + { + "epoch": 7.133839392728725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17856 + }, + { + "epoch": 7.134238913304035, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17857 + }, + { + "epoch": 7.134638433879345, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17858 + }, + { + "epoch": 7.1350379544546545, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17859 + }, + { + "epoch": 7.135437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17860 + }, + { + "epoch": 7.135836995605274, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17861 + }, + { + "epoch": 7.136236516180583, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17862 + }, + { + "epoch": 7.136636036755893, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17863 + }, + { + "epoch": 7.1370355573312025, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17864 + }, + { + "epoch": 7.137435077906512, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17865 + }, + { + "epoch": 7.137834598481822, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17866 + }, + { + "epoch": 7.138234119057131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 17867 + }, + { + "epoch": 7.138633639632441, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17868 + }, + { + "epoch": 7.139033160207751, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17869 + }, + { + "epoch": 7.13943268078306, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17870 + }, + { + "epoch": 7.13983220135837, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17871 + }, + { + "epoch": 7.14023172193368, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17872 + }, + { + "epoch": 7.140631242508989, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17873 + }, + { + "epoch": 7.141030763084299, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17874 + }, + { + "epoch": 7.141430283659608, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17875 + }, + { + "epoch": 7.141829804234918, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17876 + }, + { + "epoch": 7.142229324810228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17877 + }, + { + "epoch": 7.142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17878 + }, + { + "epoch": 7.143028365960847, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17879 + }, + { + "epoch": 7.1434278865361565, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17880 + }, + { + "epoch": 7.143827407111466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17881 + }, + { + "epoch": 7.144226927686776, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17882 + }, + { + "epoch": 7.144626448262086, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17883 + }, + { + "epoch": 7.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17884 + }, + { + "epoch": 7.1454254894127045, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17885 + }, + { + "epoch": 7.145825009988014, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 17886 + }, + { + "epoch": 7.146224530563324, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17887 + }, + { + "epoch": 7.146624051138634, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17888 + }, + { + "epoch": 7.147023571713944, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17889 + }, + { + "epoch": 7.1474230922892525, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17890 + }, + { + "epoch": 7.147822612864562, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17891 + }, + { + "epoch": 7.148222133439872, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17892 + }, + { + "epoch": 7.148621654015182, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 17893 + }, + { + "epoch": 7.149021174590492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17894 + }, + { + "epoch": 7.1494206951658015, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17895 + }, + { + "epoch": 7.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17896 + }, + { + "epoch": 7.15021973631642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17897 + }, + { + "epoch": 7.15061925689173, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17898 + }, + { + "epoch": 7.15101877746704, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17899 + }, + { + "epoch": 7.1514182980423495, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17900 + }, + { + "epoch": 7.151817818617658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17901 + }, + { + "epoch": 7.152217339192968, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17902 + }, + { + "epoch": 7.152616859768278, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17903 + }, + { + "epoch": 7.153016380343588, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17904 + }, + { + "epoch": 7.153415900918898, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17905 + }, + { + "epoch": 7.153815421494207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17906 + }, + { + "epoch": 7.154214942069516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 17907 + }, + { + "epoch": 7.154614462644826, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17908 + }, + { + "epoch": 7.155013983220136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17909 + }, + { + "epoch": 7.155413503795446, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17910 + }, + { + "epoch": 7.155813024370755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17911 + }, + { + "epoch": 7.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17912 + }, + { + "epoch": 7.156612065521374, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17913 + }, + { + "epoch": 7.157011586096684, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17914 + }, + { + "epoch": 7.157411106671994, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17915 + }, + { + "epoch": 7.157810627247303, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17916 + }, + { + "epoch": 7.158210147822613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17917 + }, + { + "epoch": 7.158609668397922, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17918 + }, + { + "epoch": 7.159009188973232, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17919 + }, + { + "epoch": 7.159408709548542, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17920 + }, + { + "epoch": 7.1598082301238515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17921 + }, + { + "epoch": 7.160207750699161, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17922 + }, + { + "epoch": 7.160607271274471, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17923 + }, + { + "epoch": 7.16100679184978, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17924 + }, + { + "epoch": 7.16140631242509, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17925 + }, + { + "epoch": 7.1618058330003995, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17926 + }, + { + "epoch": 7.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17927 + }, + { + "epoch": 7.162604874151019, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17928 + }, + { + "epoch": 7.163004394726328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17929 + }, + { + "epoch": 7.163403915301638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17930 + }, + { + "epoch": 7.163803435876948, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17931 + }, + { + "epoch": 7.164202956452257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17932 + }, + { + "epoch": 7.164602477027567, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17933 + }, + { + "epoch": 7.165001997602877, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17934 + }, + { + "epoch": 7.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17935 + }, + { + "epoch": 7.165801038753496, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17936 + }, + { + "epoch": 7.166200559328805, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17937 + }, + { + "epoch": 7.166600079904115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17938 + }, + { + "epoch": 7.166999600479425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17939 + }, + { + "epoch": 7.167399121054734, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17940 + }, + { + "epoch": 7.167798641630044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17941 + }, + { + "epoch": 7.168198162205353, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17942 + }, + { + "epoch": 7.168597682780663, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17943 + }, + { + "epoch": 7.168997203355973, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17944 + }, + { + "epoch": 7.169396723931283, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 17945 + }, + { + "epoch": 7.169796244506592, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17946 + }, + { + "epoch": 7.1701957650819015, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17947 + }, + { + "epoch": 7.170595285657211, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17948 + }, + { + "epoch": 7.170994806232521, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17949 + }, + { + "epoch": 7.171394326807831, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17950 + }, + { + "epoch": 7.171793847383141, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17951 + }, + { + "epoch": 7.1721933679584495, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17952 + }, + { + "epoch": 7.172592888533759, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17953 + }, + { + "epoch": 7.172992409109069, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17954 + }, + { + "epoch": 7.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17955 + }, + { + "epoch": 7.173791450259689, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17956 + }, + { + "epoch": 7.174190970834998, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17957 + }, + { + "epoch": 7.174590491410307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17958 + }, + { + "epoch": 7.174990011985617, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17959 + }, + { + "epoch": 7.175389532560927, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17960 + }, + { + "epoch": 7.175789053136237, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17961 + }, + { + "epoch": 7.1761885737115465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17962 + }, + { + "epoch": 7.176588094286855, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17963 + }, + { + "epoch": 7.176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17964 + }, + { + "epoch": 7.177387135437475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17965 + }, + { + "epoch": 7.177786656012785, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17966 + }, + { + "epoch": 7.1781861765880945, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17967 + }, + { + "epoch": 7.178585697163404, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17968 + }, + { + "epoch": 7.178985217738713, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17969 + }, + { + "epoch": 7.179384738314023, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17970 + }, + { + "epoch": 7.179784258889333, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17971 + }, + { + "epoch": 7.180183779464643, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17972 + }, + { + "epoch": 7.180583300039952, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17973 + }, + { + "epoch": 7.180982820615261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17974 + }, + { + "epoch": 7.181382341190571, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17975 + }, + { + "epoch": 7.181781861765881, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17976 + }, + { + "epoch": 7.182181382341191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17977 + }, + { + "epoch": 7.1825809029165, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17978 + }, + { + "epoch": 7.18298042349181, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17979 + }, + { + "epoch": 7.183379944067119, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17980 + }, + { + "epoch": 7.183779464642429, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17981 + }, + { + "epoch": 7.184178985217739, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17982 + }, + { + "epoch": 7.1845785057930485, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17983 + }, + { + "epoch": 7.184978026368358, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17984 + }, + { + "epoch": 7.185377546943668, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17985 + }, + { + "epoch": 7.185777067518977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17986 + }, + { + "epoch": 7.186176588094287, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17987 + }, + { + "epoch": 7.1865761086695965, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17988 + }, + { + "epoch": 7.186975629244906, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17989 + }, + { + "epoch": 7.187375149820216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17990 + }, + { + "epoch": 7.187774670395525, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17991 + }, + { + "epoch": 7.188174190970835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17992 + }, + { + "epoch": 7.1885737115461446, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17993 + }, + { + "epoch": 7.188973232121454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17994 + }, + { + "epoch": 7.189372752696764, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17995 + }, + { + "epoch": 7.189772273272074, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17996 + }, + { + "epoch": 7.190171793847383, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17997 + }, + { + "epoch": 7.190571314422693, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17998 + }, + { + "epoch": 7.190970834998002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17999 + }, + { + "epoch": 7.191370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 18000 + }, + { + "epoch": 7.191769876148622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18001 + }, + { + "epoch": 7.192169396723932, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18002 + }, + { + "epoch": 7.192568917299241, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18003 + }, + { + "epoch": 7.19296843787455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18004 + }, + { + "epoch": 7.19336795844986, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18005 + }, + { + "epoch": 7.19376747902517, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18006 + }, + { + "epoch": 7.19416699960048, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18007 + }, + { + "epoch": 7.194566520175789, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18008 + }, + { + "epoch": 7.1949660407510985, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18009 + }, + { + "epoch": 7.195365561326408, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18010 + }, + { + "epoch": 7.195765081901718, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18011 + }, + { + "epoch": 7.196164602477028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18012 + }, + { + "epoch": 7.196564123052338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18013 + }, + { + "epoch": 7.1969636436276465, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18014 + }, + { + "epoch": 7.197363164202956, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18015 + }, + { + "epoch": 7.197762684778266, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18016 + }, + { + "epoch": 7.198162205353576, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18017 + }, + { + "epoch": 7.198561725928886, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18018 + }, + { + "epoch": 7.1989612465041946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18019 + }, + { + "epoch": 7.199360767079504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18020 + }, + { + "epoch": 7.199760287654814, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18021 + }, + { + "epoch": 7.200159808230124, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18022 + }, + { + "epoch": 7.200559328805434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18023 + }, + { + "epoch": 7.2009588493807435, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18024 + }, + { + "epoch": 7.201358369956052, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18025 + }, + { + "epoch": 7.201757890531362, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18026 + }, + { + "epoch": 7.202157411106672, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18027 + }, + { + "epoch": 7.202556931681982, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18028 + }, + { + "epoch": 7.2029564522572915, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18029 + }, + { + "epoch": 7.2033559728326, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18030 + }, + { + "epoch": 7.20375549340791, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18031 + }, + { + "epoch": 7.20415501398322, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18032 + }, + { + "epoch": 7.20455453455853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18033 + }, + { + "epoch": 7.20495405513384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 18034 + }, + { + "epoch": 7.205353575709149, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18035 + }, + { + "epoch": 7.205753096284458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18036 + }, + { + "epoch": 7.206152616859768, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18037 + }, + { + "epoch": 7.206552137435078, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18038 + }, + { + "epoch": 7.206951658010388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18039 + }, + { + "epoch": 7.207351178585697, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18040 + }, + { + "epoch": 7.207750699161007, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18041 + }, + { + "epoch": 7.208150219736316, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18042 + }, + { + "epoch": 7.208549740311626, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18043 + }, + { + "epoch": 7.208949260886936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18044 + }, + { + "epoch": 7.2093487814622454, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18045 + }, + { + "epoch": 7.209748302037555, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18046 + }, + { + "epoch": 7.210147822612864, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18047 + }, + { + "epoch": 7.210547343188174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18048 + }, + { + "epoch": 7.210946863763484, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18049 + }, + { + "epoch": 7.2113463843387935, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18050 + }, + { + "epoch": 7.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18051 + }, + { + "epoch": 7.212145425489413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18052 + }, + { + "epoch": 7.212544946064722, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18053 + }, + { + "epoch": 7.212944466640032, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18054 + }, + { + "epoch": 7.2133439872153415, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18055 + }, + { + "epoch": 7.213743507790651, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18056 + }, + { + "epoch": 7.214143028365961, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 18057 + }, + { + "epoch": 7.214542548941271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18058 + }, + { + "epoch": 7.21494206951658, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18059 + }, + { + "epoch": 7.21534159009189, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18060 + }, + { + "epoch": 7.215741110667199, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18061 + }, + { + "epoch": 7.216140631242509, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18062 + }, + { + "epoch": 7.216540151817819, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18063 + }, + { + "epoch": 7.216939672393128, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18064 + }, + { + "epoch": 7.217339192968438, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18065 + }, + { + "epoch": 7.217738713543747, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18066 + }, + { + "epoch": 7.218138234119057, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18067 + }, + { + "epoch": 7.218537754694367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18068 + }, + { + "epoch": 7.218937275269677, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18069 + }, + { + "epoch": 7.219336795844986, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18070 + }, + { + "epoch": 7.2197363164202955, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18071 + }, + { + "epoch": 7.220135836995605, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18072 + }, + { + "epoch": 7.220535357570915, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18073 + }, + { + "epoch": 7.220934878146225, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18074 + }, + { + "epoch": 7.221334398721535, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18075 + }, + { + "epoch": 7.2217339192968435, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18076 + }, + { + "epoch": 7.222133439872153, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18077 + }, + { + "epoch": 7.222532960447463, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 18078 + }, + { + "epoch": 7.222932481022773, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18079 + }, + { + "epoch": 7.223332001598083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18080 + }, + { + "epoch": 7.2237315221733915, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18081 + }, + { + "epoch": 7.224131042748701, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 18082 + }, + { + "epoch": 7.224530563324011, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18083 + }, + { + "epoch": 7.224930083899321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18084 + }, + { + "epoch": 7.225329604474631, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18085 + }, + { + "epoch": 7.2257291250499405, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18086 + }, + { + "epoch": 7.226128645625249, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18087 + }, + { + "epoch": 7.226528166200559, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18088 + }, + { + "epoch": 7.226927686775869, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18089 + }, + { + "epoch": 7.227327207351179, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18090 + }, + { + "epoch": 7.2277267279264885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18091 + }, + { + "epoch": 7.228126248501797, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18092 + }, + { + "epoch": 7.228525769077107, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18093 + }, + { + "epoch": 7.228925289652417, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18094 + }, + { + "epoch": 7.229324810227727, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18095 + }, + { + "epoch": 7.229724330803037, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18096 + }, + { + "epoch": 7.230123851378346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18097 + }, + { + "epoch": 7.230523371953655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18098 + }, + { + "epoch": 7.230922892528965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18099 + }, + { + "epoch": 7.231322413104275, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18100 + }, + { + "epoch": 7.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18101 + }, + { + "epoch": 7.232121454254894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18102 + }, + { + "epoch": 7.232520974830204, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18103 + }, + { + "epoch": 7.232920495405513, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18104 + }, + { + "epoch": 7.233320015980823, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18105 + }, + { + "epoch": 7.233719536556133, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18106 + }, + { + "epoch": 7.234119057131442, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18107 + }, + { + "epoch": 7.234518577706752, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18108 + }, + { + "epoch": 7.234918098282061, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18109 + }, + { + "epoch": 7.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18110 + }, + { + "epoch": 7.235717139432681, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18111 + }, + { + "epoch": 7.2361166600079905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18112 + }, + { + "epoch": 7.2365161805833, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18113 + }, + { + "epoch": 7.23691570115861, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18114 + }, + { + "epoch": 7.237315221733919, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18115 + }, + { + "epoch": 7.237714742309229, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 18116 + }, + { + "epoch": 7.2381142628845385, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18117 + }, + { + "epoch": 7.238513783459848, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18118 + }, + { + "epoch": 7.238913304035158, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18119 + }, + { + "epoch": 7.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18120 + }, + { + "epoch": 7.239712345185777, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18121 + }, + { + "epoch": 7.240111865761087, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18122 + }, + { + "epoch": 7.240511386336396, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18123 + }, + { + "epoch": 7.240910906911706, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18124 + }, + { + "epoch": 7.241310427487016, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18125 + }, + { + "epoch": 7.241709948062325, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18126 + }, + { + "epoch": 7.242109468637635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18127 + }, + { + "epoch": 7.242508989212944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18128 + }, + { + "epoch": 7.242908509788254, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18129 + }, + { + "epoch": 7.243308030363564, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 18130 + }, + { + "epoch": 7.243707550938874, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18131 + }, + { + "epoch": 7.244107071514183, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18132 + }, + { + "epoch": 7.244506592089492, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18133 + }, + { + "epoch": 7.244906112664802, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18134 + }, + { + "epoch": 7.245305633240112, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 18135 + }, + { + "epoch": 7.245705153815422, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 18136 + }, + { + "epoch": 7.246104674390731, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18137 + }, + { + "epoch": 7.2465041949660405, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18138 + }, + { + "epoch": 7.24690371554135, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18139 + }, + { + "epoch": 7.24730323611666, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18140 + }, + { + "epoch": 7.24770275669197, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18141 + }, + { + "epoch": 7.24810227726728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18142 + }, + { + "epoch": 7.2485017978425885, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18143 + }, + { + "epoch": 7.248901318417898, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18144 + }, + { + "epoch": 7.249300838993208, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18145 + }, + { + "epoch": 7.249700359568518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18146 + }, + { + "epoch": 7.250099880143828, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18147 + }, + { + "epoch": 7.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18148 + }, + { + "epoch": 7.250898921294446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18149 + }, + { + "epoch": 7.251298441869756, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18150 + }, + { + "epoch": 7.251697962445066, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18151 + }, + { + "epoch": 7.252097483020376, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18152 + }, + { + "epoch": 7.2524970035956855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18153 + }, + { + "epoch": 7.252896524170994, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18154 + }, + { + "epoch": 7.253296044746304, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18155 + }, + { + "epoch": 7.253695565321614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18156 + }, + { + "epoch": 7.254095085896924, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18157 + }, + { + "epoch": 7.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18158 + }, + { + "epoch": 7.254894127047543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18159 + }, + { + "epoch": 7.255293647622852, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18160 + }, + { + "epoch": 7.255693168198162, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18161 + }, + { + "epoch": 7.256092688773472, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18162 + }, + { + "epoch": 7.256492209348782, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18163 + }, + { + "epoch": 7.256891729924091, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18164 + }, + { + "epoch": 7.257291250499401, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 18165 + }, + { + "epoch": 7.25769077107471, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18166 + }, + { + "epoch": 7.25809029165002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18167 + }, + { + "epoch": 7.25848981222533, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18168 + }, + { + "epoch": 7.258889332800639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18169 + }, + { + "epoch": 7.259288853375949, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 18170 + }, + { + "epoch": 7.259688373951258, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18171 + }, + { + "epoch": 7.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18172 + }, + { + "epoch": 7.260487415101878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18173 + }, + { + "epoch": 7.2608869356771875, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 18174 + }, + { + "epoch": 7.261286456252497, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18175 + }, + { + "epoch": 7.261685976827807, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18176 + }, + { + "epoch": 7.262085497403116, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18177 + }, + { + "epoch": 7.262485017978426, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18178 + }, + { + "epoch": 7.2628845385537355, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18179 + }, + { + "epoch": 7.263284059129045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18180 + }, + { + "epoch": 7.263683579704355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18181 + }, + { + "epoch": 7.264083100279665, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18182 + }, + { + "epoch": 7.264482620854974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18183 + }, + { + "epoch": 7.2648821414302835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18184 + }, + { + "epoch": 7.265281662005593, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18185 + }, + { + "epoch": 7.265681182580903, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18186 + }, + { + "epoch": 7.266080703156213, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18187 + }, + { + "epoch": 7.266480223731522, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 18188 + }, + { + "epoch": 7.266879744306832, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18189 + }, + { + "epoch": 7.267279264882141, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18190 + }, + { + "epoch": 7.267678785457451, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18191 + }, + { + "epoch": 7.268078306032761, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18192 + }, + { + "epoch": 7.26847782660807, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 18193 + }, + { + "epoch": 7.26887734718338, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18194 + }, + { + "epoch": 7.269276867758689, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18195 + }, + { + "epoch": 7.269676388333999, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18196 + }, + { + "epoch": 7.270075908909309, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18197 + }, + { + "epoch": 7.270475429484619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18198 + }, + { + "epoch": 7.270874950059928, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18199 + }, + { + "epoch": 7.2712744706352375, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18200 + }, + { + "epoch": 7.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18201 + }, + { + "epoch": 7.272073511785857, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 18202 + }, + { + "epoch": 7.272473032361167, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18203 + }, + { + "epoch": 7.272872552936477, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18204 + }, + { + "epoch": 7.2732720735117855, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18205 + }, + { + "epoch": 7.273671594087095, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18206 + }, + { + "epoch": 7.274071114662405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18207 + }, + { + "epoch": 7.274470635237715, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18208 + }, + { + "epoch": 7.274870155813025, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18209 + }, + { + "epoch": 7.2752696763883336, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18210 + }, + { + "epoch": 7.275669196963643, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18211 + }, + { + "epoch": 7.276068717538953, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2981, + "step": 18212 + }, + { + "epoch": 7.276468238114263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18213 + }, + { + "epoch": 7.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18214 + }, + { + "epoch": 7.2772672792648825, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18215 + }, + { + "epoch": 7.277666799840191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18216 + }, + { + "epoch": 7.278066320415501, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18217 + }, + { + "epoch": 7.278465840990811, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18218 + }, + { + "epoch": 7.278865361566121, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18219 + }, + { + "epoch": 7.2792648821414305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18220 + }, + { + "epoch": 7.27966440271674, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18221 + }, + { + "epoch": 7.280063923292049, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 18222 + }, + { + "epoch": 7.280463443867359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18223 + }, + { + "epoch": 7.280862964442669, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18224 + }, + { + "epoch": 7.281262485017979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18225 + }, + { + "epoch": 7.281662005593288, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18226 + }, + { + "epoch": 7.282061526168597, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18227 + }, + { + "epoch": 7.282461046743907, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18228 + }, + { + "epoch": 7.282860567319217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 18229 + }, + { + "epoch": 7.283260087894527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18230 + }, + { + "epoch": 7.283659608469836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 18231 + }, + { + "epoch": 7.284059129045146, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18232 + }, + { + "epoch": 7.284458649620455, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18233 + }, + { + "epoch": 7.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18234 + }, + { + "epoch": 7.285257690771075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18235 + }, + { + "epoch": 7.285657211346384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18236 + }, + { + "epoch": 7.286056731921694, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18237 + }, + { + "epoch": 7.286456252497004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18238 + }, + { + "epoch": 7.286855773072313, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18239 + }, + { + "epoch": 7.287255293647623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18240 + }, + { + "epoch": 7.2876548142229325, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18241 + }, + { + "epoch": 7.288054334798242, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18242 + }, + { + "epoch": 7.288453855373552, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18243 + }, + { + "epoch": 7.288853375948861, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18244 + }, + { + "epoch": 7.289252896524171, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18245 + }, + { + "epoch": 7.2896524170994805, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18246 + }, + { + "epoch": 7.29005193767479, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18247 + }, + { + "epoch": 7.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18248 + }, + { + "epoch": 7.29085097882541, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18249 + }, + { + "epoch": 7.291250499400719, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18250 + }, + { + "epoch": 7.291650019976029, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18251 + }, + { + "epoch": 7.292049540551338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18252 + }, + { + "epoch": 7.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18253 + }, + { + "epoch": 7.292848581701958, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18254 + }, + { + "epoch": 7.293248102277268, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18255 + }, + { + "epoch": 7.293647622852577, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18256 + }, + { + "epoch": 7.294047143427886, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18257 + }, + { + "epoch": 7.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18258 + }, + { + "epoch": 7.294846184578506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18259 + }, + { + "epoch": 7.295245705153816, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18260 + }, + { + "epoch": 7.295645225729125, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18261 + }, + { + "epoch": 7.2960447463044344, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18262 + }, + { + "epoch": 7.296444266879744, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18263 + }, + { + "epoch": 7.296843787455054, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18264 + }, + { + "epoch": 7.297243308030364, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18265 + }, + { + "epoch": 7.297642828605674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18266 + }, + { + "epoch": 7.2980423491809825, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18267 + }, + { + "epoch": 7.298441869756292, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 18268 + }, + { + "epoch": 7.298841390331602, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18269 + }, + { + "epoch": 7.299240910906912, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18270 + }, + { + "epoch": 7.299640431482222, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18271 + }, + { + "epoch": 7.300039952057531, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18272 + }, + { + "epoch": 7.30043947263284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18273 + }, + { + "epoch": 7.30083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18274 + }, + { + "epoch": 7.30123851378346, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18275 + }, + { + "epoch": 7.30163803435877, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18276 + }, + { + "epoch": 7.3020375549340795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18277 + }, + { + "epoch": 7.302437075509388, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18278 + }, + { + "epoch": 7.302836596084698, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18279 + }, + { + "epoch": 7.303236116660008, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18280 + }, + { + "epoch": 7.303635637235318, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18281 + }, + { + "epoch": 7.3040351578106275, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18282 + }, + { + "epoch": 7.304434678385936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18283 + }, + { + "epoch": 7.304834198961246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 18284 + }, + { + "epoch": 7.305233719536556, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18285 + }, + { + "epoch": 7.305633240111866, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18286 + }, + { + "epoch": 7.3060327606871756, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18287 + }, + { + "epoch": 7.306432281262485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18288 + }, + { + "epoch": 7.306831801837794, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18289 + }, + { + "epoch": 7.307231322413104, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18290 + }, + { + "epoch": 7.307630842988414, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18291 + }, + { + "epoch": 7.308030363563724, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18292 + }, + { + "epoch": 7.308429884139033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18293 + }, + { + "epoch": 7.308829404714343, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18294 + }, + { + "epoch": 7.309228925289652, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18295 + }, + { + "epoch": 7.309628445864962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18296 + }, + { + "epoch": 7.310027966440272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18297 + }, + { + "epoch": 7.310427487015581, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18298 + }, + { + "epoch": 7.310827007590891, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18299 + }, + { + "epoch": 7.3112265281662, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18300 + }, + { + "epoch": 7.31162604874151, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18301 + }, + { + "epoch": 7.31202556931682, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18302 + }, + { + "epoch": 7.3124250898921295, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18303 + }, + { + "epoch": 7.312824610467439, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18304 + }, + { + "epoch": 7.313224131042749, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18305 + }, + { + "epoch": 7.313623651618058, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18306 + }, + { + "epoch": 7.314023172193368, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 18307 + }, + { + "epoch": 7.3144226927686775, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18308 + }, + { + "epoch": 7.314822213343987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18309 + }, + { + "epoch": 7.315221733919297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18310 + }, + { + "epoch": 7.315621254494607, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18311 + }, + { + "epoch": 7.316020775069916, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18312 + }, + { + "epoch": 7.316420295645226, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 18313 + }, + { + "epoch": 7.316819816220535, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18314 + }, + { + "epoch": 7.317219336795845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18315 + }, + { + "epoch": 7.317618857371155, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18316 + }, + { + "epoch": 7.318018377946464, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 18317 + }, + { + "epoch": 7.318417898521774, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18318 + }, + { + "epoch": 7.318817419097083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18319 + }, + { + "epoch": 7.319216939672393, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18320 + }, + { + "epoch": 7.319616460247703, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18321 + }, + { + "epoch": 7.320015980823013, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18322 + }, + { + "epoch": 7.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18323 + }, + { + "epoch": 7.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18324 + }, + { + "epoch": 7.321214542548941, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18325 + }, + { + "epoch": 7.321614063124251, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18326 + }, + { + "epoch": 7.322013583699561, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18327 + }, + { + "epoch": 7.322413104274871, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18328 + }, + { + "epoch": 7.3228126248501795, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18329 + }, + { + "epoch": 7.323212145425489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18330 + }, + { + "epoch": 7.323611666000799, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18331 + }, + { + "epoch": 7.324011186576109, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18332 + }, + { + "epoch": 7.324410707151419, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18333 + }, + { + "epoch": 7.3248102277267275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18334 + }, + { + "epoch": 7.325209748302037, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18335 + }, + { + "epoch": 7.325609268877347, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18336 + }, + { + "epoch": 7.326008789452657, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18337 + }, + { + "epoch": 7.326408310027967, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18338 + }, + { + "epoch": 7.3268078306032765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18339 + }, + { + "epoch": 7.327207351178585, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18340 + }, + { + "epoch": 7.327606871753895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18341 + }, + { + "epoch": 7.328006392329205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18342 + }, + { + "epoch": 7.328405912904515, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18343 + }, + { + "epoch": 7.3288054334798245, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18344 + }, + { + "epoch": 7.329204954055134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18345 + }, + { + "epoch": 7.329604474630443, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18346 + }, + { + "epoch": 7.330003995205753, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18347 + }, + { + "epoch": 7.330403515781063, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 18348 + }, + { + "epoch": 7.3308030363563725, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18349 + }, + { + "epoch": 7.331202556931682, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18350 + }, + { + "epoch": 7.331602077506991, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18351 + }, + { + "epoch": 7.332001598082301, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18352 + }, + { + "epoch": 7.332401118657611, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18353 + }, + { + "epoch": 7.332800639232921, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18354 + }, + { + "epoch": 7.33320015980823, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18355 + }, + { + "epoch": 7.33359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18356 + }, + { + "epoch": 7.333999200958849, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18357 + }, + { + "epoch": 7.334398721534159, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18358 + }, + { + "epoch": 7.334798242109469, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18359 + }, + { + "epoch": 7.335197762684778, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18360 + }, + { + "epoch": 7.335597283260088, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18361 + }, + { + "epoch": 7.335996803835398, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18362 + }, + { + "epoch": 7.336396324410707, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18363 + }, + { + "epoch": 7.336795844986017, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18364 + }, + { + "epoch": 7.3371953655613265, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18365 + }, + { + "epoch": 7.337594886136636, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18366 + }, + { + "epoch": 7.337994406711946, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18367 + }, + { + "epoch": 7.338393927287255, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18368 + }, + { + "epoch": 7.338793447862565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18369 + }, + { + "epoch": 7.3391929684378745, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18370 + }, + { + "epoch": 7.339592489013184, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18371 + }, + { + "epoch": 7.339992009588494, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18372 + }, + { + "epoch": 7.340391530163803, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18373 + }, + { + "epoch": 7.340791050739113, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18374 + }, + { + "epoch": 7.3411905713144225, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18375 + }, + { + "epoch": 7.341590091889732, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18376 + }, + { + "epoch": 7.341989612465042, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18377 + }, + { + "epoch": 7.342389133040352, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18378 + }, + { + "epoch": 7.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18379 + }, + { + "epoch": 7.343188174190971, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18380 + }, + { + "epoch": 7.34358769476628, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18381 + }, + { + "epoch": 7.34398721534159, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18382 + }, + { + "epoch": 7.3443867359169, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 18383 + }, + { + "epoch": 7.34478625649221, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18384 + }, + { + "epoch": 7.345185777067519, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18385 + }, + { + "epoch": 7.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18386 + }, + { + "epoch": 7.345984818218138, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18387 + }, + { + "epoch": 7.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18388 + }, + { + "epoch": 7.346783859368758, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18389 + }, + { + "epoch": 7.347183379944067, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18390 + }, + { + "epoch": 7.3475829005193765, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18391 + }, + { + "epoch": 7.347982421094686, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18392 + }, + { + "epoch": 7.348381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18393 + }, + { + "epoch": 7.348781462245306, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18394 + }, + { + "epoch": 7.349180982820616, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18395 + }, + { + "epoch": 7.3495805033959245, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18396 + }, + { + "epoch": 7.349980023971234, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18397 + }, + { + "epoch": 7.350379544546544, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18398 + }, + { + "epoch": 7.350779065121854, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18399 + }, + { + "epoch": 7.351178585697164, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18400 + }, + { + "epoch": 7.351578106272473, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18401 + }, + { + "epoch": 7.351977626847782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18402 + }, + { + "epoch": 7.352377147423092, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 18403 + }, + { + "epoch": 7.352776667998402, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18404 + }, + { + "epoch": 7.353176188573712, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18405 + }, + { + "epoch": 7.3535757091490215, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18406 + }, + { + "epoch": 7.35397522972433, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18407 + }, + { + "epoch": 7.35437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18408 + }, + { + "epoch": 7.35477427087495, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18409 + }, + { + "epoch": 7.35517379145026, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18410 + }, + { + "epoch": 7.3555733120255695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18411 + }, + { + "epoch": 7.355972832600879, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18412 + }, + { + "epoch": 7.356372353176188, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18413 + }, + { + "epoch": 7.356771873751498, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18414 + }, + { + "epoch": 7.357171394326808, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 18415 + }, + { + "epoch": 7.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18416 + }, + { + "epoch": 7.357970435477427, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18417 + }, + { + "epoch": 7.358369956052737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18418 + }, + { + "epoch": 7.358769476628046, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18419 + }, + { + "epoch": 7.359168997203356, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18420 + }, + { + "epoch": 7.359568517778666, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 18421 + }, + { + "epoch": 7.359968038353975, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18422 + }, + { + "epoch": 7.360367558929285, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 18423 + }, + { + "epoch": 7.360767079504594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18424 + }, + { + "epoch": 7.361166600079904, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18425 + }, + { + "epoch": 7.361566120655214, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18426 + }, + { + "epoch": 7.361965641230523, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18427 + }, + { + "epoch": 7.362365161805833, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18428 + }, + { + "epoch": 7.362764682381143, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18429 + }, + { + "epoch": 7.363164202956452, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18430 + }, + { + "epoch": 7.363563723531762, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18431 + }, + { + "epoch": 7.3639632441070715, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18432 + }, + { + "epoch": 7.364362764682381, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18433 + }, + { + "epoch": 7.364762285257691, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18434 + }, + { + "epoch": 7.365161805833001, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18435 + }, + { + "epoch": 7.36556132640831, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18436 + }, + { + "epoch": 7.3659608469836195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18437 + }, + { + "epoch": 7.366360367558929, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18438 + }, + { + "epoch": 7.366759888134239, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18439 + }, + { + "epoch": 7.367159408709549, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18440 + }, + { + "epoch": 7.367558929284858, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18441 + }, + { + "epoch": 7.367958449860168, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18442 + }, + { + "epoch": 7.368357970435477, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18443 + }, + { + "epoch": 7.368757491010787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18444 + }, + { + "epoch": 7.369157011586097, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18445 + }, + { + "epoch": 7.369556532161407, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18446 + }, + { + "epoch": 7.369956052736716, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18447 + }, + { + "epoch": 7.370355573312025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18448 + }, + { + "epoch": 7.370755093887335, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18449 + }, + { + "epoch": 7.371154614462645, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18450 + }, + { + "epoch": 7.371554135037955, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18451 + }, + { + "epoch": 7.3719536556132645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18452 + }, + { + "epoch": 7.372353176188573, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18453 + }, + { + "epoch": 7.372752696763883, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18454 + }, + { + "epoch": 7.373152217339193, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18455 + }, + { + "epoch": 7.373551737914503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18456 + }, + { + "epoch": 7.373951258489813, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 18457 + }, + { + "epoch": 7.3743507790651215, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18458 + }, + { + "epoch": 7.374750299640431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18459 + }, + { + "epoch": 7.375149820215741, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18460 + }, + { + "epoch": 7.375549340791051, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18461 + }, + { + "epoch": 7.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18462 + }, + { + "epoch": 7.3763483819416695, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18463 + }, + { + "epoch": 7.376747902516979, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18464 + }, + { + "epoch": 7.377147423092289, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18465 + }, + { + "epoch": 7.377546943667599, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 18466 + }, + { + "epoch": 7.377946464242909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18467 + }, + { + "epoch": 7.3783459848182185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18468 + }, + { + "epoch": 7.378745505393527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18469 + }, + { + "epoch": 7.379145025968837, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18470 + }, + { + "epoch": 7.379544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18471 + }, + { + "epoch": 7.379944067119457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18472 + }, + { + "epoch": 7.3803435876947665, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18473 + }, + { + "epoch": 7.380743108270076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18474 + }, + { + "epoch": 7.381142628845385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18475 + }, + { + "epoch": 7.381542149420695, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18476 + }, + { + "epoch": 7.381941669996005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18477 + }, + { + "epoch": 7.3823411905713145, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18478 + }, + { + "epoch": 7.382740711146624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18479 + }, + { + "epoch": 7.383140231721933, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18480 + }, + { + "epoch": 7.383539752297243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18481 + }, + { + "epoch": 7.383939272872553, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18482 + }, + { + "epoch": 7.384338793447863, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18483 + }, + { + "epoch": 7.384738314023172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18484 + }, + { + "epoch": 7.385137834598482, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18485 + }, + { + "epoch": 7.385537355173791, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18486 + }, + { + "epoch": 7.385936875749101, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 18487 + }, + { + "epoch": 7.386336396324411, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18488 + }, + { + "epoch": 7.38673591689972, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18489 + }, + { + "epoch": 7.38713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18490 + }, + { + "epoch": 7.38753495805034, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18491 + }, + { + "epoch": 7.387934478625649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18492 + }, + { + "epoch": 7.388333999200959, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18493 + }, + { + "epoch": 7.3887335197762685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18494 + }, + { + "epoch": 7.389133040351578, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18495 + }, + { + "epoch": 7.389532560926888, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18496 + }, + { + "epoch": 7.389932081502197, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18497 + }, + { + "epoch": 7.390331602077507, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18498 + }, + { + "epoch": 7.3907311226528165, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18499 + }, + { + "epoch": 7.391130643228126, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18500 + }, + { + "epoch": 7.391530163803436, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18501 + }, + { + "epoch": 7.391929684378746, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 18502 + }, + { + "epoch": 7.392329204954055, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18503 + }, + { + "epoch": 7.3927287255293646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18504 + }, + { + "epoch": 7.393128246104674, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18505 + }, + { + "epoch": 7.393527766679984, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18506 + }, + { + "epoch": 7.393927287255294, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18507 + }, + { + "epoch": 7.394326807830604, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18508 + }, + { + "epoch": 7.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18509 + }, + { + "epoch": 7.395125848981222, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18510 + }, + { + "epoch": 7.395525369556532, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18511 + }, + { + "epoch": 7.395924890131842, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18512 + }, + { + "epoch": 7.396324410707152, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18513 + }, + { + "epoch": 7.396723931282461, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 18514 + }, + { + "epoch": 7.39712345185777, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18515 + }, + { + "epoch": 7.39752297243308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18516 + }, + { + "epoch": 7.39792249300839, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18517 + }, + { + "epoch": 7.3983220135837, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18518 + }, + { + "epoch": 7.39872153415901, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18519 + }, + { + "epoch": 7.3991210547343185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 18520 + }, + { + "epoch": 7.399520575309628, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18521 + }, + { + "epoch": 7.399920095884938, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18522 + }, + { + "epoch": 7.400319616460248, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18523 + }, + { + "epoch": 7.400719137035558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18524 + }, + { + "epoch": 7.401118657610867, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18525 + }, + { + "epoch": 7.401518178186176, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18526 + }, + { + "epoch": 7.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18527 + }, + { + "epoch": 7.402317219336796, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18528 + }, + { + "epoch": 7.402716739912106, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18529 + }, + { + "epoch": 7.4031162604874154, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18530 + }, + { + "epoch": 7.403515781062724, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 18531 + }, + { + "epoch": 7.403915301638034, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18532 + }, + { + "epoch": 7.404314822213344, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18533 + }, + { + "epoch": 7.404714342788654, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18534 + }, + { + "epoch": 7.4051138633639635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 18535 + }, + { + "epoch": 7.405513383939273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18536 + }, + { + "epoch": 7.405912904514582, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18537 + }, + { + "epoch": 7.406312425089892, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18538 + }, + { + "epoch": 7.406711945665202, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18539 + }, + { + "epoch": 7.4071114662405115, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18540 + }, + { + "epoch": 7.407510986815821, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18541 + }, + { + "epoch": 7.407910507391131, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18542 + }, + { + "epoch": 7.40831002796644, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18543 + }, + { + "epoch": 7.40870954854175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18544 + }, + { + "epoch": 7.40910906911706, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18545 + }, + { + "epoch": 7.409508589692369, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18546 + }, + { + "epoch": 7.409908110267679, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18547 + }, + { + "epoch": 7.410307630842988, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18548 + }, + { + "epoch": 7.410707151418298, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18549 + }, + { + "epoch": 7.411106671993608, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18550 + }, + { + "epoch": 7.411506192568917, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18551 + }, + { + "epoch": 7.411905713144227, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18552 + }, + { + "epoch": 7.412305233719536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18553 + }, + { + "epoch": 7.412704754294846, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18554 + }, + { + "epoch": 7.413104274870156, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18555 + }, + { + "epoch": 7.4135037954454654, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18556 + }, + { + "epoch": 7.413903316020775, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18557 + }, + { + "epoch": 7.414302836596085, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18558 + }, + { + "epoch": 7.414702357171394, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18559 + }, + { + "epoch": 7.415101877746704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18560 + }, + { + "epoch": 7.4155013983220135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18561 + }, + { + "epoch": 7.415900918897323, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18562 + }, + { + "epoch": 7.416300439472633, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18563 + }, + { + "epoch": 7.416699960047943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18564 + }, + { + "epoch": 7.417099480623252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18565 + }, + { + "epoch": 7.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.2993, + "step": 18566 + }, + { + "epoch": 7.417898521773871, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18567 + }, + { + "epoch": 7.418298042349181, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18568 + }, + { + "epoch": 7.418697562924491, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18569 + }, + { + "epoch": 7.4190970834998, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18570 + }, + { + "epoch": 7.41949660407511, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18571 + }, + { + "epoch": 7.419896124650419, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18572 + }, + { + "epoch": 7.420295645225729, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18573 + }, + { + "epoch": 7.420695165801039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18574 + }, + { + "epoch": 7.421094686376349, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18575 + }, + { + "epoch": 7.421494206951658, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18576 + }, + { + "epoch": 7.421893727526967, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18577 + }, + { + "epoch": 7.422293248102277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18578 + }, + { + "epoch": 7.422692768677587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18579 + }, + { + "epoch": 7.423092289252897, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18580 + }, + { + "epoch": 7.4234918098282066, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 18581 + }, + { + "epoch": 7.4238913304035155, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18582 + }, + { + "epoch": 7.424290850978825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18583 + }, + { + "epoch": 7.424690371554135, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18584 + }, + { + "epoch": 7.425089892129445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18585 + }, + { + "epoch": 7.425489412704755, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18586 + }, + { + "epoch": 7.4258889332800635, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18587 + }, + { + "epoch": 7.426288453855373, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18588 + }, + { + "epoch": 7.426687974430683, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18589 + }, + { + "epoch": 7.427087495005993, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18590 + }, + { + "epoch": 7.427487015581303, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18591 + }, + { + "epoch": 7.427886536156612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18592 + }, + { + "epoch": 7.428286056731921, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18593 + }, + { + "epoch": 7.428685577307231, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18594 + }, + { + "epoch": 7.429085097882541, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18595 + }, + { + "epoch": 7.429484618457851, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18596 + }, + { + "epoch": 7.4298841390331605, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18597 + }, + { + "epoch": 7.43028365960847, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18598 + }, + { + "epoch": 7.430683180183779, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18599 + }, + { + "epoch": 7.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18600 + }, + { + "epoch": 7.431482221334399, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18601 + }, + { + "epoch": 7.4318817419097085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18602 + }, + { + "epoch": 7.432281262485018, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18603 + }, + { + "epoch": 7.432680783060327, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18604 + }, + { + "epoch": 7.433080303635637, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18605 + }, + { + "epoch": 7.433479824210947, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18606 + }, + { + "epoch": 7.433879344786257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18607 + }, + { + "epoch": 7.434278865361566, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18608 + }, + { + "epoch": 7.434678385936876, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18609 + }, + { + "epoch": 7.435077906512185, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 18610 + }, + { + "epoch": 7.435477427087495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18611 + }, + { + "epoch": 7.435876947662805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18612 + }, + { + "epoch": 7.436276468238114, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 18613 + }, + { + "epoch": 7.436675988813424, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18614 + }, + { + "epoch": 7.437075509388734, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18615 + }, + { + "epoch": 7.437475029964043, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18616 + }, + { + "epoch": 7.437874550539353, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18617 + }, + { + "epoch": 7.438274071114662, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18618 + }, + { + "epoch": 7.438673591689972, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18619 + }, + { + "epoch": 7.439073112265282, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18620 + }, + { + "epoch": 7.439472632840591, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18621 + }, + { + "epoch": 7.439872153415901, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18622 + }, + { + "epoch": 7.4402716739912105, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18623 + }, + { + "epoch": 7.44067119456652, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18624 + }, + { + "epoch": 7.44107071514183, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3015, + "step": 18625 + }, + { + "epoch": 7.44147023571714, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18626 + }, + { + "epoch": 7.441869756292449, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18627 + }, + { + "epoch": 7.4422692768677585, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18628 + }, + { + "epoch": 7.442668797443068, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18629 + }, + { + "epoch": 7.443068318018378, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18630 + }, + { + "epoch": 7.443467838593688, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18631 + }, + { + "epoch": 7.443867359168998, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 18632 + }, + { + "epoch": 7.444266879744307, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18633 + }, + { + "epoch": 7.444666400319616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 18634 + }, + { + "epoch": 7.445065920894926, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 18635 + }, + { + "epoch": 7.445465441470236, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18636 + }, + { + "epoch": 7.445864962045546, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18637 + }, + { + "epoch": 7.446264482620855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18638 + }, + { + "epoch": 7.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18639 + }, + { + "epoch": 7.447063523771474, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 18640 + }, + { + "epoch": 7.447463044346784, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18641 + }, + { + "epoch": 7.447862564922094, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18642 + }, + { + "epoch": 7.448262085497403, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18643 + }, + { + "epoch": 7.448661606072712, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18644 + }, + { + "epoch": 7.449061126648022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18645 + }, + { + "epoch": 7.449460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18646 + }, + { + "epoch": 7.449860167798642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18647 + }, + { + "epoch": 7.450259688373952, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18648 + }, + { + "epoch": 7.4506592089492605, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18649 + }, + { + "epoch": 7.45105872952457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18650 + }, + { + "epoch": 7.45145825009988, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18651 + }, + { + "epoch": 7.45185777067519, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18652 + }, + { + "epoch": 7.4522572912505, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18653 + }, + { + "epoch": 7.452656811825809, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18654 + }, + { + "epoch": 7.453056332401118, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18655 + }, + { + "epoch": 7.453455852976428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18656 + }, + { + "epoch": 7.453855373551738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 18657 + }, + { + "epoch": 7.454254894127048, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18658 + }, + { + "epoch": 7.4546544147023575, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18659 + }, + { + "epoch": 7.455053935277666, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18660 + }, + { + "epoch": 7.455453455852976, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18661 + }, + { + "epoch": 7.455852976428286, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18662 + }, + { + "epoch": 7.456252497003596, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18663 + }, + { + "epoch": 7.4566520175789055, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18664 + }, + { + "epoch": 7.457051538154215, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18665 + }, + { + "epoch": 7.457451058729524, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18666 + }, + { + "epoch": 7.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18667 + }, + { + "epoch": 7.458250099880144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18668 + }, + { + "epoch": 7.4586496204554535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18669 + }, + { + "epoch": 7.459049141030763, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18670 + }, + { + "epoch": 7.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18671 + }, + { + "epoch": 7.459848182181382, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18672 + }, + { + "epoch": 7.460247702756692, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18673 + }, + { + "epoch": 7.460647223332002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18674 + }, + { + "epoch": 7.461046743907311, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 18675 + }, + { + "epoch": 7.461446264482621, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18676 + }, + { + "epoch": 7.46184578505793, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18677 + }, + { + "epoch": 7.46224530563324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18678 + }, + { + "epoch": 7.46264482620855, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18679 + }, + { + "epoch": 7.463044346783859, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18680 + }, + { + "epoch": 7.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18681 + }, + { + "epoch": 7.463843387934479, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18682 + }, + { + "epoch": 7.464242908509788, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18683 + }, + { + "epoch": 7.464642429085098, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18684 + }, + { + "epoch": 7.4650419496604075, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18685 + }, + { + "epoch": 7.465441470235717, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18686 + }, + { + "epoch": 7.465840990811027, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18687 + }, + { + "epoch": 7.466240511386337, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18688 + }, + { + "epoch": 7.466640031961646, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18689 + }, + { + "epoch": 7.4670395525369555, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18690 + }, + { + "epoch": 7.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18691 + }, + { + "epoch": 7.467838593687575, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18692 + }, + { + "epoch": 7.468238114262885, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18693 + }, + { + "epoch": 7.468637634838194, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18694 + }, + { + "epoch": 7.4690371554135035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18695 + }, + { + "epoch": 7.469436675988813, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18696 + }, + { + "epoch": 7.469836196564123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18697 + }, + { + "epoch": 7.470235717139433, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18698 + }, + { + "epoch": 7.470635237714743, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18699 + }, + { + "epoch": 7.471034758290052, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18700 + }, + { + "epoch": 7.471434278865361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 18701 + }, + { + "epoch": 7.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18702 + }, + { + "epoch": 7.472233320015981, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.2967, + "step": 18703 + }, + { + "epoch": 7.472632840591291, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18704 + }, + { + "epoch": 7.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18705 + }, + { + "epoch": 7.473431881741909, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18706 + }, + { + "epoch": 7.473831402317219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18707 + }, + { + "epoch": 7.474230922892529, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18708 + }, + { + "epoch": 7.474630443467839, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 18709 + }, + { + "epoch": 7.475029964043149, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 18710 + }, + { + "epoch": 7.4754294846184575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18711 + }, + { + "epoch": 7.475829005193767, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18712 + }, + { + "epoch": 7.476228525769077, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18713 + }, + { + "epoch": 7.476628046344387, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18714 + }, + { + "epoch": 7.477027566919697, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18715 + }, + { + "epoch": 7.477427087495006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 18716 + }, + { + "epoch": 7.477826608070315, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18717 + }, + { + "epoch": 7.478226128645625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18718 + }, + { + "epoch": 7.478625649220935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 18719 + }, + { + "epoch": 7.479025169796245, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18720 + }, + { + "epoch": 7.479424690371554, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18721 + }, + { + "epoch": 7.479824210946864, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18722 + }, + { + "epoch": 7.480223731522173, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18723 + }, + { + "epoch": 7.480623252097483, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18724 + }, + { + "epoch": 7.481022772672793, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18725 + }, + { + "epoch": 7.4814222932481025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18726 + }, + { + "epoch": 7.481821813823412, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18727 + }, + { + "epoch": 7.482221334398721, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18728 + }, + { + "epoch": 7.482620854974031, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18729 + }, + { + "epoch": 7.483020375549341, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18730 + }, + { + "epoch": 7.4834198961246505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18731 + }, + { + "epoch": 7.48381941669996, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18732 + }, + { + "epoch": 7.484218937275269, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18733 + }, + { + "epoch": 7.484618457850579, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18734 + }, + { + "epoch": 7.485017978425889, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18735 + }, + { + "epoch": 7.485417499001199, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18736 + }, + { + "epoch": 7.485817019576508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18737 + }, + { + "epoch": 7.486216540151818, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18738 + }, + { + "epoch": 7.486616060727127, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18739 + }, + { + "epoch": 7.487015581302437, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18740 + }, + { + "epoch": 7.487415101877747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18741 + }, + { + "epoch": 7.487814622453056, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18742 + }, + { + "epoch": 7.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18743 + }, + { + "epoch": 7.488613663603676, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18744 + }, + { + "epoch": 7.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18745 + }, + { + "epoch": 7.489412704754295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18746 + }, + { + "epoch": 7.4898122253296044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18747 + }, + { + "epoch": 7.490211745904914, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18748 + }, + { + "epoch": 7.490611266480224, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18749 + }, + { + "epoch": 7.491010787055533, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18750 + }, + { + "epoch": 7.491410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18751 + }, + { + "epoch": 7.4918098282061525, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18752 + }, + { + "epoch": 7.492209348781462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18753 + }, + { + "epoch": 7.492608869356772, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18754 + }, + { + "epoch": 7.493008389932082, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18755 + }, + { + "epoch": 7.493407910507391, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18756 + }, + { + "epoch": 7.4938074310827005, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18757 + }, + { + "epoch": 7.49420695165801, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18758 + }, + { + "epoch": 7.49460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18759 + }, + { + "epoch": 7.49500599280863, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18760 + }, + { + "epoch": 7.49540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18761 + }, + { + "epoch": 7.495805033959249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18762 + }, + { + "epoch": 7.496204554534558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18763 + }, + { + "epoch": 7.496604075109868, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18764 + }, + { + "epoch": 7.497003595685178, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18765 + }, + { + "epoch": 7.497403116260488, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18766 + }, + { + "epoch": 7.497802636835797, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18767 + }, + { + "epoch": 7.498202157411106, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18768 + }, + { + "epoch": 7.498601677986416, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18769 + }, + { + "epoch": 7.499001198561726, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18770 + }, + { + "epoch": 7.499400719137036, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18771 + }, + { + "epoch": 7.4998002397123456, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18772 + }, + { + "epoch": 7.5001997602876544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18773 + }, + { + "epoch": 7.500599280862964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18774 + }, + { + "epoch": 7.500998801438274, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18775 + }, + { + "epoch": 7.501398322013584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18776 + }, + { + "epoch": 7.501797842588894, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18777 + }, + { + "epoch": 7.502197363164203, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18778 + }, + { + "epoch": 7.502596883739512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18779 + }, + { + "epoch": 7.502996404314822, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18780 + }, + { + "epoch": 7.503395924890132, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18781 + }, + { + "epoch": 7.503795445465442, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18782 + }, + { + "epoch": 7.504194966040751, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18783 + }, + { + "epoch": 7.50459448661606, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18784 + }, + { + "epoch": 7.50499400719137, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18785 + }, + { + "epoch": 7.50539352776668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18786 + }, + { + "epoch": 7.50579304834199, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18787 + }, + { + "epoch": 7.5061925689172995, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18788 + }, + { + "epoch": 7.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18789 + }, + { + "epoch": 7.506991610067918, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18790 + }, + { + "epoch": 7.507391130643228, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18791 + }, + { + "epoch": 7.507790651218538, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18792 + }, + { + "epoch": 7.5081901717938475, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18793 + }, + { + "epoch": 7.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18794 + }, + { + "epoch": 7.508989212944467, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18795 + }, + { + "epoch": 7.509388733519776, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18796 + }, + { + "epoch": 7.509788254095086, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18797 + }, + { + "epoch": 7.5101877746703956, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18798 + }, + { + "epoch": 7.510587295245705, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18799 + }, + { + "epoch": 7.510986815821015, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18800 + }, + { + "epoch": 7.511386336396324, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18801 + }, + { + "epoch": 7.511785856971634, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18802 + }, + { + "epoch": 7.512185377546944, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18803 + }, + { + "epoch": 7.512584898122253, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18804 + }, + { + "epoch": 7.512984418697563, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 18805 + }, + { + "epoch": 7.513383939272872, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18806 + }, + { + "epoch": 7.513783459848182, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18807 + }, + { + "epoch": 7.514182980423492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18808 + }, + { + "epoch": 7.514582500998801, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18809 + }, + { + "epoch": 7.514982021574111, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18810 + }, + { + "epoch": 7.515381542149421, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18811 + }, + { + "epoch": 7.515781062724731, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18812 + }, + { + "epoch": 7.51618058330004, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18813 + }, + { + "epoch": 7.5165801038753495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18814 + }, + { + "epoch": 7.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18815 + }, + { + "epoch": 7.517379145025969, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18816 + }, + { + "epoch": 7.517778665601279, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18817 + }, + { + "epoch": 7.518178186176588, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18818 + }, + { + "epoch": 7.5185777067518975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18819 + }, + { + "epoch": 7.518977227327207, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18820 + }, + { + "epoch": 7.519376747902517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18821 + }, + { + "epoch": 7.519776268477827, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18822 + }, + { + "epoch": 7.520175789053136, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18823 + }, + { + "epoch": 7.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18824 + }, + { + "epoch": 7.520974830203755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18825 + }, + { + "epoch": 7.521374350779065, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18826 + }, + { + "epoch": 7.521773871354375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18827 + }, + { + "epoch": 7.522173391929685, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18828 + }, + { + "epoch": 7.5225729125049945, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18829 + }, + { + "epoch": 7.522972433080303, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18830 + }, + { + "epoch": 7.523371953655613, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18831 + }, + { + "epoch": 7.523771474230923, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18832 + }, + { + "epoch": 7.524170994806233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18833 + }, + { + "epoch": 7.5245705153815425, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18834 + }, + { + "epoch": 7.524970035956851, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18835 + }, + { + "epoch": 7.525369556532161, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18836 + }, + { + "epoch": 7.525769077107471, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18837 + }, + { + "epoch": 7.526168597682781, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18838 + }, + { + "epoch": 7.526568118258091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18839 + }, + { + "epoch": 7.5269676388333995, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18840 + }, + { + "epoch": 7.527367159408709, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18841 + }, + { + "epoch": 7.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18842 + }, + { + "epoch": 7.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18843 + }, + { + "epoch": 7.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18844 + }, + { + "epoch": 7.528965241709948, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18845 + }, + { + "epoch": 7.529364762285257, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18846 + }, + { + "epoch": 7.529764282860567, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 18847 + }, + { + "epoch": 7.530163803435877, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18848 + }, + { + "epoch": 7.530563324011187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18849 + }, + { + "epoch": 7.5309628445864965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18850 + }, + { + "epoch": 7.531362365161806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18851 + }, + { + "epoch": 7.531761885737115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18852 + }, + { + "epoch": 7.532161406312425, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18853 + }, + { + "epoch": 7.532560926887735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18854 + }, + { + "epoch": 7.5329604474630445, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18855 + }, + { + "epoch": 7.533359968038354, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18856 + }, + { + "epoch": 7.533759488613663, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18857 + }, + { + "epoch": 7.534159009188973, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18858 + }, + { + "epoch": 7.534558529764283, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18859 + }, + { + "epoch": 7.5349580503395925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18860 + }, + { + "epoch": 7.535357570914902, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18861 + }, + { + "epoch": 7.535757091490212, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18862 + }, + { + "epoch": 7.536156612065521, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18863 + }, + { + "epoch": 7.536556132640831, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18864 + }, + { + "epoch": 7.536955653216141, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18865 + }, + { + "epoch": 7.53735517379145, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18866 + }, + { + "epoch": 7.53775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18867 + }, + { + "epoch": 7.53815421494207, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18868 + }, + { + "epoch": 7.538553735517379, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18869 + }, + { + "epoch": 7.538953256092689, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18870 + }, + { + "epoch": 7.539352776667998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18871 + }, + { + "epoch": 7.539752297243308, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 18872 + }, + { + "epoch": 7.540151817818618, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18873 + }, + { + "epoch": 7.540551338393927, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18874 + }, + { + "epoch": 7.540950858969237, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18875 + }, + { + "epoch": 7.5413503795445465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18876 + }, + { + "epoch": 7.541749900119856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18877 + }, + { + "epoch": 7.542149420695166, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18878 + }, + { + "epoch": 7.542548941270476, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18879 + }, + { + "epoch": 7.542948461845785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18880 + }, + { + "epoch": 7.5433479824210945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18881 + }, + { + "epoch": 7.543747502996404, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18882 + }, + { + "epoch": 7.544147023571714, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18883 + }, + { + "epoch": 7.544546544147024, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18884 + }, + { + "epoch": 7.544946064722334, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18885 + }, + { + "epoch": 7.5453455852976425, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18886 + }, + { + "epoch": 7.545745105872952, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18887 + }, + { + "epoch": 7.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18888 + }, + { + "epoch": 7.546544147023572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18889 + }, + { + "epoch": 7.546943667598882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18890 + }, + { + "epoch": 7.547343188174191, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18891 + }, + { + "epoch": 7.5477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18892 + }, + { + "epoch": 7.54814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18893 + }, + { + "epoch": 7.54854174990012, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18894 + }, + { + "epoch": 7.54894127047543, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18895 + }, + { + "epoch": 7.549340791050739, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18896 + }, + { + "epoch": 7.549740311626048, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18897 + }, + { + "epoch": 7.550139832201358, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18898 + }, + { + "epoch": 7.550539352776668, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18899 + }, + { + "epoch": 7.550938873351978, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18900 + }, + { + "epoch": 7.551338393927288, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18901 + }, + { + "epoch": 7.551737914502597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18902 + }, + { + "epoch": 7.552137435077906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18903 + }, + { + "epoch": 7.552536955653216, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18904 + }, + { + "epoch": 7.552936476228526, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18905 + }, + { + "epoch": 7.553335996803836, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18906 + }, + { + "epoch": 7.553735517379145, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18907 + }, + { + "epoch": 7.554135037954454, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18908 + }, + { + "epoch": 7.554534558529764, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18909 + }, + { + "epoch": 7.554934079105074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18910 + }, + { + "epoch": 7.555333599680384, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18911 + }, + { + "epoch": 7.555733120255693, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18912 + }, + { + "epoch": 7.556132640831002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18913 + }, + { + "epoch": 7.556532161406312, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18914 + }, + { + "epoch": 7.556931681981622, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18915 + }, + { + "epoch": 7.557331202556932, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18916 + }, + { + "epoch": 7.5577307231322415, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18917 + }, + { + "epoch": 7.558130243707551, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18918 + }, + { + "epoch": 7.558529764282861, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18919 + }, + { + "epoch": 7.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18920 + }, + { + "epoch": 7.55932880543348, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18921 + }, + { + "epoch": 7.5597283260087895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18922 + }, + { + "epoch": 7.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18923 + }, + { + "epoch": 7.560527367159409, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18924 + }, + { + "epoch": 7.560926887734718, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18925 + }, + { + "epoch": 7.561326408310028, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18926 + }, + { + "epoch": 7.561725928885338, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18927 + }, + { + "epoch": 7.562125449460647, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18928 + }, + { + "epoch": 7.562524970035957, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18929 + }, + { + "epoch": 7.562924490611266, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 18930 + }, + { + "epoch": 7.563324011186576, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18931 + }, + { + "epoch": 7.563723531761886, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18932 + }, + { + "epoch": 7.564123052337195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18933 + }, + { + "epoch": 7.564522572912505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18934 + }, + { + "epoch": 7.564922093487815, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18935 + }, + { + "epoch": 7.565321614063124, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18936 + }, + { + "epoch": 7.565721134638434, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18937 + }, + { + "epoch": 7.566120655213743, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18938 + }, + { + "epoch": 7.566520175789053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18939 + }, + { + "epoch": 7.566919696364363, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18940 + }, + { + "epoch": 7.567319216939673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18941 + }, + { + "epoch": 7.567718737514982, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 18942 + }, + { + "epoch": 7.5681182580902915, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18943 + }, + { + "epoch": 7.568517778665601, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18944 + }, + { + "epoch": 7.568917299240911, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18945 + }, + { + "epoch": 7.569316819816221, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18946 + }, + { + "epoch": 7.56971634039153, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18947 + }, + { + "epoch": 7.5701158609668395, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18948 + }, + { + "epoch": 7.570515381542149, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18949 + }, + { + "epoch": 7.570914902117459, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 18950 + }, + { + "epoch": 7.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18951 + }, + { + "epoch": 7.571713943268079, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18952 + }, + { + "epoch": 7.572113463843388, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18953 + }, + { + "epoch": 7.572512984418697, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18954 + }, + { + "epoch": 7.572912504994007, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18955 + }, + { + "epoch": 7.573312025569317, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18956 + }, + { + "epoch": 7.573711546144627, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18957 + }, + { + "epoch": 7.5741110667199365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18958 + }, + { + "epoch": 7.574510587295245, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18959 + }, + { + "epoch": 7.574910107870555, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 18960 + }, + { + "epoch": 7.575309628445865, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18961 + }, + { + "epoch": 7.575709149021175, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18962 + }, + { + "epoch": 7.5761086695964845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18963 + }, + { + "epoch": 7.5765081901717934, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18964 + }, + { + "epoch": 7.576907710747103, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18965 + }, + { + "epoch": 7.577307231322413, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18966 + }, + { + "epoch": 7.577706751897723, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 18967 + }, + { + "epoch": 7.578106272473033, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18968 + }, + { + "epoch": 7.578505793048342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18969 + }, + { + "epoch": 7.578905313623651, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18970 + }, + { + "epoch": 7.579304834198961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18971 + }, + { + "epoch": 7.579704354774271, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 18972 + }, + { + "epoch": 7.580103875349581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18973 + }, + { + "epoch": 7.58050339592489, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18974 + }, + { + "epoch": 7.5809029165002, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18975 + }, + { + "epoch": 7.581302437075509, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18976 + }, + { + "epoch": 7.581701957650819, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18977 + }, + { + "epoch": 7.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18978 + }, + { + "epoch": 7.5825009988014385, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18979 + }, + { + "epoch": 7.582900519376748, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18980 + }, + { + "epoch": 7.583300039952057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 18981 + }, + { + "epoch": 7.583699560527367, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18982 + }, + { + "epoch": 7.584099081102677, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18983 + }, + { + "epoch": 7.5844986016779865, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18984 + }, + { + "epoch": 7.584898122253296, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18985 + }, + { + "epoch": 7.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18986 + }, + { + "epoch": 7.585697163403915, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18987 + }, + { + "epoch": 7.586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18988 + }, + { + "epoch": 7.5864962045545346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18989 + }, + { + "epoch": 7.586895725129844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18990 + }, + { + "epoch": 7.587295245705154, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18991 + }, + { + "epoch": 7.587694766280464, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18992 + }, + { + "epoch": 7.588094286855773, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18993 + }, + { + "epoch": 7.588493807431083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18994 + }, + { + "epoch": 7.588893328006392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18995 + }, + { + "epoch": 7.589292848581702, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18996 + }, + { + "epoch": 7.589692369157012, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18997 + }, + { + "epoch": 7.590091889732321, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18998 + }, + { + "epoch": 7.590491410307631, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18999 + }, + { + "epoch": 7.59089093088294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19000 + }, + { + "epoch": 7.59129045145825, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19001 + }, + { + "epoch": 7.59168997203356, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19002 + }, + { + "epoch": 7.592089492608869, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 19003 + }, + { + "epoch": 7.592489013184179, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19004 + }, + { + "epoch": 7.5928885337594885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19005 + }, + { + "epoch": 7.593288054334798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19006 + }, + { + "epoch": 7.593687574910108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19007 + }, + { + "epoch": 7.594087095485418, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19008 + }, + { + "epoch": 7.594486616060728, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19009 + }, + { + "epoch": 7.5948861366360365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19010 + }, + { + "epoch": 7.595285657211346, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19011 + }, + { + "epoch": 7.595685177786656, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19012 + }, + { + "epoch": 7.596084698361966, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19013 + }, + { + "epoch": 7.596484218937276, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 19014 + }, + { + "epoch": 7.5968837395125846, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19015 + }, + { + "epoch": 7.597283260087894, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19016 + }, + { + "epoch": 7.597682780663204, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19017 + }, + { + "epoch": 7.598082301238514, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19018 + }, + { + "epoch": 7.598481821813824, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19019 + }, + { + "epoch": 7.598881342389133, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19020 + }, + { + "epoch": 7.599280862964442, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19021 + }, + { + "epoch": 7.599680383539752, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19022 + }, + { + "epoch": 7.600079904115062, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19023 + }, + { + "epoch": 7.600479424690372, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19024 + }, + { + "epoch": 7.6008789452656815, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19025 + }, + { + "epoch": 7.60127846584099, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19026 + }, + { + "epoch": 7.6016779864163, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19027 + }, + { + "epoch": 7.60207750699161, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19028 + }, + { + "epoch": 7.60247702756692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19029 + }, + { + "epoch": 7.60287654814223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19030 + }, + { + "epoch": 7.603276068717539, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19031 + }, + { + "epoch": 7.603675589292848, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19032 + }, + { + "epoch": 7.604075109868158, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19033 + }, + { + "epoch": 7.604474630443468, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19034 + }, + { + "epoch": 7.604874151018778, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19035 + }, + { + "epoch": 7.605273671594087, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19036 + }, + { + "epoch": 7.605673192169396, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19037 + }, + { + "epoch": 7.606072712744706, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19038 + }, + { + "epoch": 7.606472233320016, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19039 + }, + { + "epoch": 7.606871753895326, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19040 + }, + { + "epoch": 7.6072712744706354, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19041 + }, + { + "epoch": 7.607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19042 + }, + { + "epoch": 7.608070315621254, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19043 + }, + { + "epoch": 7.608469836196564, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19044 + }, + { + "epoch": 7.608869356771874, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19045 + }, + { + "epoch": 7.6092688773471835, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19046 + }, + { + "epoch": 7.609668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19047 + }, + { + "epoch": 7.610067918497803, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19048 + }, + { + "epoch": 7.610467439073112, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19049 + }, + { + "epoch": 7.610866959648422, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19050 + }, + { + "epoch": 7.6112664802237315, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19051 + }, + { + "epoch": 7.611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 19052 + }, + { + "epoch": 7.612065521374351, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19053 + }, + { + "epoch": 7.61246504194966, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19054 + }, + { + "epoch": 7.61286456252497, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19055 + }, + { + "epoch": 7.61326408310028, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19056 + }, + { + "epoch": 7.613663603675589, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19057 + }, + { + "epoch": 7.614063124250899, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19058 + }, + { + "epoch": 7.614462644826209, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19059 + }, + { + "epoch": 7.614862165401518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19060 + }, + { + "epoch": 7.615261685976828, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19061 + }, + { + "epoch": 7.615661206552137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19062 + }, + { + "epoch": 7.616060727127447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19063 + }, + { + "epoch": 7.616460247702757, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19064 + }, + { + "epoch": 7.616859768278067, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19065 + }, + { + "epoch": 7.617259288853376, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19066 + }, + { + "epoch": 7.6176588094286855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19067 + }, + { + "epoch": 7.618058330003995, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19068 + }, + { + "epoch": 7.618457850579305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19069 + }, + { + "epoch": 7.618857371154615, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19070 + }, + { + "epoch": 7.619256891729924, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19071 + }, + { + "epoch": 7.6196564123052335, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19072 + }, + { + "epoch": 7.620055932880543, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19073 + }, + { + "epoch": 7.620455453455853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19074 + }, + { + "epoch": 7.620854974031163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19075 + }, + { + "epoch": 7.621254494606472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19076 + }, + { + "epoch": 7.6216540151817815, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19077 + }, + { + "epoch": 7.622053535757091, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19078 + }, + { + "epoch": 7.622453056332401, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19079 + }, + { + "epoch": 7.622852576907711, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19080 + }, + { + "epoch": 7.623252097483021, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19081 + }, + { + "epoch": 7.6236516180583305, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19082 + }, + { + "epoch": 7.624051138633639, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19083 + }, + { + "epoch": 7.624450659208949, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19084 + }, + { + "epoch": 7.624850179784259, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19085 + }, + { + "epoch": 7.625249700359569, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19086 + }, + { + "epoch": 7.6256492209348785, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19087 + }, + { + "epoch": 7.626048741510187, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19088 + }, + { + "epoch": 7.626448262085497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19089 + }, + { + "epoch": 7.626847782660807, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19090 + }, + { + "epoch": 7.627247303236117, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19091 + }, + { + "epoch": 7.627646823811427, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19092 + }, + { + "epoch": 7.6280463443867355, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19093 + }, + { + "epoch": 7.628445864962045, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19094 + }, + { + "epoch": 7.628845385537355, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19095 + }, + { + "epoch": 7.629244906112665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19096 + }, + { + "epoch": 7.629644426687975, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19097 + }, + { + "epoch": 7.630043947263284, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19098 + }, + { + "epoch": 7.630443467838594, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19099 + }, + { + "epoch": 7.630842988413903, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19100 + }, + { + "epoch": 7.631242508989213, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19101 + }, + { + "epoch": 7.631642029564523, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19102 + }, + { + "epoch": 7.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19103 + }, + { + "epoch": 7.632441070715142, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19104 + }, + { + "epoch": 7.632840591290451, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19105 + }, + { + "epoch": 7.633240111865761, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19106 + }, + { + "epoch": 7.633639632441071, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19107 + }, + { + "epoch": 7.6340391530163805, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19108 + }, + { + "epoch": 7.63443867359169, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19109 + }, + { + "epoch": 7.634838194166999, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19110 + }, + { + "epoch": 7.635237714742309, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19111 + }, + { + "epoch": 7.635637235317619, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19112 + }, + { + "epoch": 7.6360367558929285, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19113 + }, + { + "epoch": 7.636436276468238, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19114 + }, + { + "epoch": 7.636835797043548, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19115 + }, + { + "epoch": 7.637235317618857, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 19116 + }, + { + "epoch": 7.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19117 + }, + { + "epoch": 7.638034358769477, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19118 + }, + { + "epoch": 7.638433879344786, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19119 + }, + { + "epoch": 7.638833399920096, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19120 + }, + { + "epoch": 7.639232920495406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19121 + }, + { + "epoch": 7.639632441070715, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19122 + }, + { + "epoch": 7.640031961646025, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19123 + }, + { + "epoch": 7.640431482221334, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19124 + }, + { + "epoch": 7.640831002796644, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19125 + }, + { + "epoch": 7.641230523371954, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19126 + }, + { + "epoch": 7.641630043947263, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 19127 + }, + { + "epoch": 7.642029564522573, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19128 + }, + { + "epoch": 7.642429085097882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19129 + }, + { + "epoch": 7.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19130 + }, + { + "epoch": 7.643228126248502, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19131 + }, + { + "epoch": 7.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19132 + }, + { + "epoch": 7.644027167399121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 19133 + }, + { + "epoch": 7.6444266879744305, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19134 + }, + { + "epoch": 7.64482620854974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19135 + }, + { + "epoch": 7.64522572912505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19136 + }, + { + "epoch": 7.64562524970036, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19137 + }, + { + "epoch": 7.64602477027567, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19138 + }, + { + "epoch": 7.6464242908509785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19139 + }, + { + "epoch": 7.646823811426288, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 19140 + }, + { + "epoch": 7.647223332001598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19141 + }, + { + "epoch": 7.647622852576908, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19142 + }, + { + "epoch": 7.648022373152218, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 19143 + }, + { + "epoch": 7.648421893727527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19144 + }, + { + "epoch": 7.648821414302836, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19145 + }, + { + "epoch": 7.649220934878146, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19146 + }, + { + "epoch": 7.649620455453456, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19147 + }, + { + "epoch": 7.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19148 + }, + { + "epoch": 7.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19149 + }, + { + "epoch": 7.650819017179384, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 19150 + }, + { + "epoch": 7.651218537754694, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19151 + }, + { + "epoch": 7.651618058330004, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19152 + }, + { + "epoch": 7.652017578905314, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19153 + }, + { + "epoch": 7.6524170994806235, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19154 + }, + { + "epoch": 7.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19155 + }, + { + "epoch": 7.653216140631242, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19156 + }, + { + "epoch": 7.653615661206552, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19157 + }, + { + "epoch": 7.654015181781862, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19158 + }, + { + "epoch": 7.654414702357172, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19159 + }, + { + "epoch": 7.654814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19160 + }, + { + "epoch": 7.65521374350779, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19161 + }, + { + "epoch": 7.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19162 + }, + { + "epoch": 7.65601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 19163 + }, + { + "epoch": 7.65641230523372, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19164 + }, + { + "epoch": 7.656811825809029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19165 + }, + { + "epoch": 7.657211346384338, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19166 + }, + { + "epoch": 7.657610866959648, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 19167 + }, + { + "epoch": 7.658010387534958, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19168 + }, + { + "epoch": 7.658409908110268, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19169 + }, + { + "epoch": 7.6588094286855775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19170 + }, + { + "epoch": 7.659208949260887, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19171 + }, + { + "epoch": 7.659608469836197, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19172 + }, + { + "epoch": 7.660007990411506, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19173 + }, + { + "epoch": 7.660407510986816, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19174 + }, + { + "epoch": 7.6608070315621255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19175 + }, + { + "epoch": 7.661206552137435, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19176 + }, + { + "epoch": 7.661606072712745, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19177 + }, + { + "epoch": 7.662005593288054, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19178 + }, + { + "epoch": 7.662405113863364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19179 + }, + { + "epoch": 7.6628046344386735, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19180 + }, + { + "epoch": 7.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19181 + }, + { + "epoch": 7.663603675589293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19182 + }, + { + "epoch": 7.664003196164602, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19183 + }, + { + "epoch": 7.664402716739912, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19184 + }, + { + "epoch": 7.664802237315222, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 19185 + }, + { + "epoch": 7.665201757890531, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19186 + }, + { + "epoch": 7.665601278465841, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 19187 + }, + { + "epoch": 7.666000799041151, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19188 + }, + { + "epoch": 7.666400319616461, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19189 + }, + { + "epoch": 7.66679984019177, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19190 + }, + { + "epoch": 7.667199360767079, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19191 + }, + { + "epoch": 7.667598881342389, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19192 + }, + { + "epoch": 7.667998401917699, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19193 + }, + { + "epoch": 7.668397922493009, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19194 + }, + { + "epoch": 7.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19195 + }, + { + "epoch": 7.6691969636436275, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19196 + }, + { + "epoch": 7.669596484218937, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19197 + }, + { + "epoch": 7.669996004794247, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19198 + }, + { + "epoch": 7.670395525369557, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19199 + }, + { + "epoch": 7.670795045944866, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19200 + }, + { + "epoch": 7.6711945665201755, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19201 + }, + { + "epoch": 7.671594087095485, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19202 + }, + { + "epoch": 7.671993607670795, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19203 + }, + { + "epoch": 7.672393128246105, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 19204 + }, + { + "epoch": 7.672792648821415, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 19205 + }, + { + "epoch": 7.6731921693967235, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19206 + }, + { + "epoch": 7.673591689972033, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19207 + }, + { + "epoch": 7.673991210547343, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19208 + }, + { + "epoch": 7.674390731122653, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19209 + }, + { + "epoch": 7.674790251697963, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19210 + }, + { + "epoch": 7.6751897722732725, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19211 + }, + { + "epoch": 7.675589292848581, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19212 + }, + { + "epoch": 7.675988813423891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19213 + }, + { + "epoch": 7.676388333999201, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19214 + }, + { + "epoch": 7.676787854574511, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19215 + }, + { + "epoch": 7.6771873751498205, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19216 + }, + { + "epoch": 7.677586895725129, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 19217 + }, + { + "epoch": 7.677986416300439, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 19218 + }, + { + "epoch": 7.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19219 + }, + { + "epoch": 7.678785457451059, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19220 + }, + { + "epoch": 7.679184978026369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19221 + }, + { + "epoch": 7.679584498601678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19222 + }, + { + "epoch": 7.679984019176987, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19223 + }, + { + "epoch": 7.680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19224 + }, + { + "epoch": 7.680783060327607, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19225 + }, + { + "epoch": 7.681182580902917, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19226 + }, + { + "epoch": 7.681582101478226, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19227 + }, + { + "epoch": 7.681981622053536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19228 + }, + { + "epoch": 7.682381142628845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19229 + }, + { + "epoch": 7.682780663204155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19230 + }, + { + "epoch": 7.683180183779465, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19231 + }, + { + "epoch": 7.683579704354774, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19232 + }, + { + "epoch": 7.683979224930084, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19233 + }, + { + "epoch": 7.684378745505393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19234 + }, + { + "epoch": 7.684778266080703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19235 + }, + { + "epoch": 7.685177786656013, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19236 + }, + { + "epoch": 7.6855773072313225, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19237 + }, + { + "epoch": 7.685976827806632, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19238 + }, + { + "epoch": 7.686376348381942, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19239 + }, + { + "epoch": 7.686775868957251, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19240 + }, + { + "epoch": 7.687175389532561, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19241 + }, + { + "epoch": 7.6875749101078705, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19242 + }, + { + "epoch": 7.68797443068318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19243 + }, + { + "epoch": 7.68837395125849, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19244 + }, + { + "epoch": 7.6887734718338, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19245 + }, + { + "epoch": 7.689172992409109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19246 + }, + { + "epoch": 7.689572512984419, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19247 + }, + { + "epoch": 7.689972033559728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 19248 + }, + { + "epoch": 7.690371554135038, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19249 + }, + { + "epoch": 7.690771074710348, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19250 + }, + { + "epoch": 7.691170595285657, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19251 + }, + { + "epoch": 7.691570115860967, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19252 + }, + { + "epoch": 7.691969636436276, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19253 + }, + { + "epoch": 7.692369157011586, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19254 + }, + { + "epoch": 7.692768677586896, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19255 + }, + { + "epoch": 7.693168198162205, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19256 + }, + { + "epoch": 7.693567718737515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19257 + }, + { + "epoch": 7.6939672393128244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19258 + }, + { + "epoch": 7.694366759888134, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19259 + }, + { + "epoch": 7.694766280463444, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19260 + }, + { + "epoch": 7.695165801038754, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19261 + }, + { + "epoch": 7.695565321614064, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19262 + }, + { + "epoch": 7.6959648421893725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19263 + }, + { + "epoch": 7.696364362764682, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 19264 + }, + { + "epoch": 7.696763883339992, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 19265 + }, + { + "epoch": 7.697163403915302, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19266 + }, + { + "epoch": 7.697562924490612, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19267 + }, + { + "epoch": 7.6979624450659205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 19268 + }, + { + "epoch": 7.69836196564123, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19269 + }, + { + "epoch": 7.69876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19270 + }, + { + "epoch": 7.69916100679185, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19271 + }, + { + "epoch": 7.69956052736716, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19272 + }, + { + "epoch": 7.699960047942469, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19273 + }, + { + "epoch": 7.700359568517778, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19274 + }, + { + "epoch": 7.700759089093088, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19275 + }, + { + "epoch": 7.701158609668398, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 19276 + }, + { + "epoch": 7.701558130243708, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19277 + }, + { + "epoch": 7.7019576508190175, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19278 + }, + { + "epoch": 7.702357171394327, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19279 + }, + { + "epoch": 7.702756691969636, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19280 + }, + { + "epoch": 7.703156212544946, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19281 + }, + { + "epoch": 7.703555733120256, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19282 + }, + { + "epoch": 7.7039552536955656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19283 + }, + { + "epoch": 7.704354774270875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19284 + }, + { + "epoch": 7.704754294846184, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19285 + }, + { + "epoch": 7.705153815421494, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19286 + }, + { + "epoch": 7.705553335996804, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19287 + }, + { + "epoch": 7.705952856572114, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19288 + }, + { + "epoch": 7.706352377147423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19289 + }, + { + "epoch": 7.706751897722732, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19290 + }, + { + "epoch": 7.707151418298042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19291 + }, + { + "epoch": 7.707550938873352, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19292 + }, + { + "epoch": 7.707950459448662, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19293 + }, + { + "epoch": 7.708349980023971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19294 + }, + { + "epoch": 7.708749500599281, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19295 + }, + { + "epoch": 7.70914902117459, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19296 + }, + { + "epoch": 7.7095485417499, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19297 + }, + { + "epoch": 7.70994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19298 + }, + { + "epoch": 7.7103475829005195, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19299 + }, + { + "epoch": 7.710747103475829, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19300 + }, + { + "epoch": 7.711146624051139, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19301 + }, + { + "epoch": 7.711546144626448, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19302 + }, + { + "epoch": 7.711945665201758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19303 + }, + { + "epoch": 7.7123451857770675, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19304 + }, + { + "epoch": 7.712744706352377, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19305 + }, + { + "epoch": 7.713144226927687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19306 + }, + { + "epoch": 7.713543747502996, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19307 + }, + { + "epoch": 7.713943268078306, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19308 + }, + { + "epoch": 7.714342788653616, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19309 + }, + { + "epoch": 7.714742309228925, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19310 + }, + { + "epoch": 7.715141829804235, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19311 + }, + { + "epoch": 7.715541350379545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19312 + }, + { + "epoch": 7.715940870954854, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19313 + }, + { + "epoch": 7.716340391530164, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19314 + }, + { + "epoch": 7.716739912105473, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 19315 + }, + { + "epoch": 7.717139432680783, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19316 + }, + { + "epoch": 7.717538953256093, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19317 + }, + { + "epoch": 7.717938473831403, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19318 + }, + { + "epoch": 7.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19319 + }, + { + "epoch": 7.718737514982021, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19320 + }, + { + "epoch": 7.719137035557331, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19321 + }, + { + "epoch": 7.719536556132641, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19322 + }, + { + "epoch": 7.719936076707951, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19323 + }, + { + "epoch": 7.72033559728326, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19324 + }, + { + "epoch": 7.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19325 + }, + { + "epoch": 7.721134638433879, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19326 + }, + { + "epoch": 7.721534159009189, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19327 + }, + { + "epoch": 7.721933679584499, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19328 + }, + { + "epoch": 7.722333200159808, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19329 + }, + { + "epoch": 7.7227327207351175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19330 + }, + { + "epoch": 7.723132241310427, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19331 + }, + { + "epoch": 7.723531761885737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 19332 + }, + { + "epoch": 7.723931282461047, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19333 + }, + { + "epoch": 7.724330803036357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19334 + }, + { + "epoch": 7.7247303236116664, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19335 + }, + { + "epoch": 7.725129844186975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19336 + }, + { + "epoch": 7.725529364762285, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19337 + }, + { + "epoch": 7.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19338 + }, + { + "epoch": 7.726328405912905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19339 + }, + { + "epoch": 7.7267279264882145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19340 + }, + { + "epoch": 7.727127447063523, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19341 + }, + { + "epoch": 7.727526967638833, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19342 + }, + { + "epoch": 7.727926488214143, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19343 + }, + { + "epoch": 7.728326008789453, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19344 + }, + { + "epoch": 7.7287255293647625, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 19345 + }, + { + "epoch": 7.729125049940071, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19346 + }, + { + "epoch": 7.729524570515381, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19347 + }, + { + "epoch": 7.729924091090691, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19348 + }, + { + "epoch": 7.730323611666001, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19349 + }, + { + "epoch": 7.730723132241311, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19350 + }, + { + "epoch": 7.73112265281662, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19351 + }, + { + "epoch": 7.73152217339193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19352 + }, + { + "epoch": 7.731921693967239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 19353 + }, + { + "epoch": 7.732321214542549, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19354 + }, + { + "epoch": 7.732720735117859, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 19355 + }, + { + "epoch": 7.733120255693168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19356 + }, + { + "epoch": 7.733519776268478, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19357 + }, + { + "epoch": 7.733919296843787, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 19358 + }, + { + "epoch": 7.734318817419097, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19359 + }, + { + "epoch": 7.734718337994407, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19360 + }, + { + "epoch": 7.7351178585697165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19361 + }, + { + "epoch": 7.735517379145026, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19362 + }, + { + "epoch": 7.735916899720335, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19363 + }, + { + "epoch": 7.736316420295645, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19364 + }, + { + "epoch": 7.736715940870955, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19365 + }, + { + "epoch": 7.7371154614462645, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 19366 + }, + { + "epoch": 7.737514982021574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19367 + }, + { + "epoch": 7.737914502596884, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19368 + }, + { + "epoch": 7.738314023172194, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19369 + }, + { + "epoch": 7.738713543747503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19370 + }, + { + "epoch": 7.7391130643228125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19371 + }, + { + "epoch": 7.739512584898122, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19372 + }, + { + "epoch": 7.739912105473432, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19373 + }, + { + "epoch": 7.740311626048742, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19374 + }, + { + "epoch": 7.740711146624051, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19375 + }, + { + "epoch": 7.741110667199361, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19376 + }, + { + "epoch": 7.74151018777467, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19377 + }, + { + "epoch": 7.74190970834998, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19378 + }, + { + "epoch": 7.74230922892529, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19379 + }, + { + "epoch": 7.742708749500599, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19380 + }, + { + "epoch": 7.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19381 + }, + { + "epoch": 7.743507790651218, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19382 + }, + { + "epoch": 7.743907311226528, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19383 + }, + { + "epoch": 7.744306831801838, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19384 + }, + { + "epoch": 7.744706352377148, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19385 + }, + { + "epoch": 7.745105872952457, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19386 + }, + { + "epoch": 7.7455053935277665, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19387 + }, + { + "epoch": 7.745904914103076, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19388 + }, + { + "epoch": 7.746304434678386, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 19389 + }, + { + "epoch": 7.746703955253696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 19390 + }, + { + "epoch": 7.747103475829006, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19391 + }, + { + "epoch": 7.7475029964043145, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19392 + }, + { + "epoch": 7.747902516979624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19393 + }, + { + "epoch": 7.748302037554934, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19394 + }, + { + "epoch": 7.748701558130244, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19395 + }, + { + "epoch": 7.749101078705554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19396 + }, + { + "epoch": 7.7495005992808625, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19397 + }, + { + "epoch": 7.749900119856172, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19398 + }, + { + "epoch": 7.750299640431482, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19399 + }, + { + "epoch": 7.750699161006792, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19400 + }, + { + "epoch": 7.751098681582102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 19401 + }, + { + "epoch": 7.7514982021574115, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 19402 + }, + { + "epoch": 7.75189772273272, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19403 + }, + { + "epoch": 7.75229724330803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19404 + }, + { + "epoch": 7.75269676388334, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19405 + }, + { + "epoch": 7.75309628445865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19406 + }, + { + "epoch": 7.7534958050339595, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19407 + }, + { + "epoch": 7.753895325609269, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 19408 + }, + { + "epoch": 7.754294846184578, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19409 + }, + { + "epoch": 7.754694366759888, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19410 + }, + { + "epoch": 7.755093887335198, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19411 + }, + { + "epoch": 7.755493407910508, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 19412 + }, + { + "epoch": 7.755892928485817, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19413 + }, + { + "epoch": 7.756292449061126, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19414 + }, + { + "epoch": 7.756691969636436, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19415 + }, + { + "epoch": 7.757091490211746, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19416 + }, + { + "epoch": 7.757491010787056, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19417 + }, + { + "epoch": 7.757890531362365, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19418 + }, + { + "epoch": 7.758290051937674, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19419 + }, + { + "epoch": 7.758689572512984, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 19420 + }, + { + "epoch": 7.759089093088294, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 19421 + }, + { + "epoch": 7.759488613663604, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19422 + }, + { + "epoch": 7.759888134238913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19423 + }, + { + "epoch": 7.760287654814223, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19424 + }, + { + "epoch": 7.760687175389533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 19425 + }, + { + "epoch": 7.761086695964842, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19426 + }, + { + "epoch": 7.761486216540152, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 19427 + }, + { + "epoch": 7.7618857371154615, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19428 + }, + { + "epoch": 7.762285257690771, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19429 + }, + { + "epoch": 7.762684778266081, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19430 + }, + { + "epoch": 7.76308429884139, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19431 + }, + { + "epoch": 7.7634838194167, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19432 + }, + { + "epoch": 7.7638833399920095, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19433 + }, + { + "epoch": 7.764282860567319, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19434 + }, + { + "epoch": 7.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19435 + }, + { + "epoch": 7.765081901717938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19436 + }, + { + "epoch": 7.765481422293248, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 19437 + }, + { + "epoch": 7.765880942868558, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19438 + }, + { + "epoch": 7.766280463443867, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19439 + }, + { + "epoch": 7.766679984019177, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19440 + }, + { + "epoch": 7.767079504594487, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19441 + }, + { + "epoch": 7.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19442 + }, + { + "epoch": 7.767878545745106, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19443 + }, + { + "epoch": 7.768278066320415, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19444 + }, + { + "epoch": 7.768677586895725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19445 + }, + { + "epoch": 7.769077107471035, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 19446 + }, + { + "epoch": 7.769476628046345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19447 + }, + { + "epoch": 7.769876148621654, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19448 + }, + { + "epoch": 7.770275669196963, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19449 + }, + { + "epoch": 7.770675189772273, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 19450 + }, + { + "epoch": 7.771074710347583, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19451 + }, + { + "epoch": 7.771474230922893, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19452 + }, + { + "epoch": 7.771873751498202, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19453 + }, + { + "epoch": 7.7722732720735115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19454 + }, + { + "epoch": 7.772672792648821, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19455 + }, + { + "epoch": 7.773072313224131, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19456 + }, + { + "epoch": 7.773471833799441, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19457 + }, + { + "epoch": 7.773871354374751, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19458 + }, + { + "epoch": 7.77427087495006, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19459 + }, + { + "epoch": 7.774670395525369, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19460 + }, + { + "epoch": 7.775069916100679, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19461 + }, + { + "epoch": 7.775469436675989, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19462 + }, + { + "epoch": 7.775868957251299, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19463 + }, + { + "epoch": 7.7762684778266085, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19464 + }, + { + "epoch": 7.776667998401917, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19465 + }, + { + "epoch": 7.777067518977227, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 19466 + }, + { + "epoch": 7.777467039552537, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19467 + }, + { + "epoch": 7.777866560127847, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19468 + }, + { + "epoch": 7.7782660807031565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19469 + }, + { + "epoch": 7.778665601278465, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19470 + }, + { + "epoch": 7.779065121853775, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 19471 + }, + { + "epoch": 7.779464642429085, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 19472 + }, + { + "epoch": 7.779864163004395, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19473 + }, + { + "epoch": 7.7802636835797045, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19474 + }, + { + "epoch": 7.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19475 + }, + { + "epoch": 7.781062724730323, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19476 + }, + { + "epoch": 7.781462245305633, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19477 + }, + { + "epoch": 7.781861765880943, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19478 + }, + { + "epoch": 7.782261286456253, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19479 + }, + { + "epoch": 7.782660807031562, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19480 + }, + { + "epoch": 7.783060327606872, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19481 + }, + { + "epoch": 7.783459848182181, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19482 + }, + { + "epoch": 7.783859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19483 + }, + { + "epoch": 7.784258889332801, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19484 + }, + { + "epoch": 7.78465840990811, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19485 + }, + { + "epoch": 7.78505793048342, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19486 + }, + { + "epoch": 7.785457451058729, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19487 + }, + { + "epoch": 7.785856971634039, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 19488 + }, + { + "epoch": 7.786256492209349, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19489 + }, + { + "epoch": 7.7866560127846585, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19490 + }, + { + "epoch": 7.787055533359968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19491 + }, + { + "epoch": 7.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19492 + }, + { + "epoch": 7.787854574510587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 19493 + }, + { + "epoch": 7.788254095085897, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19494 + }, + { + "epoch": 7.7886536156612065, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19495 + }, + { + "epoch": 7.789053136236516, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19496 + }, + { + "epoch": 7.789452656811826, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19497 + }, + { + "epoch": 7.789852177387136, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19498 + }, + { + "epoch": 7.790251697962445, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19499 + }, + { + "epoch": 7.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19500 + }, + { + "epoch": 7.791050739113064, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19501 + }, + { + "epoch": 7.791450259688374, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19502 + }, + { + "epoch": 7.791849780263684, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19503 + }, + { + "epoch": 7.792249300838993, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19504 + }, + { + "epoch": 7.792648821414303, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19505 + }, + { + "epoch": 7.793048341989612, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 19506 + }, + { + "epoch": 7.793447862564922, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19507 + }, + { + "epoch": 7.793847383140232, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 19508 + }, + { + "epoch": 7.794246903715541, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19509 + }, + { + "epoch": 7.794646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19510 + }, + { + "epoch": 7.79504594486616, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19511 + }, + { + "epoch": 7.79544546544147, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19512 + }, + { + "epoch": 7.79584498601678, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19513 + }, + { + "epoch": 7.79624450659209, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19514 + }, + { + "epoch": 7.7966440271674, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19515 + }, + { + "epoch": 7.7970435477427085, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19516 + }, + { + "epoch": 7.797443068318018, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 19517 + }, + { + "epoch": 7.797842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 19518 + }, + { + "epoch": 7.798242109468638, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19519 + }, + { + "epoch": 7.798641630043948, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19520 + }, + { + "epoch": 7.7990411506192565, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19521 + }, + { + "epoch": 7.799440671194566, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19522 + }, + { + "epoch": 7.799840191769876, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19523 + }, + { + "epoch": 7.800239712345186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19524 + }, + { + "epoch": 7.800639232920496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 19525 + }, + { + "epoch": 7.801038753495805, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19526 + }, + { + "epoch": 7.801438274071114, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19527 + }, + { + "epoch": 7.801837794646424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19528 + }, + { + "epoch": 7.802237315221734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19529 + }, + { + "epoch": 7.802636835797044, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19530 + }, + { + "epoch": 7.8030363563723535, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19531 + }, + { + "epoch": 7.803435876947663, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19532 + }, + { + "epoch": 7.803835397522972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19533 + }, + { + "epoch": 7.804234918098282, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19534 + }, + { + "epoch": 7.804634438673592, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19535 + }, + { + "epoch": 7.8050339592489015, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19536 + }, + { + "epoch": 7.805433479824211, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19537 + }, + { + "epoch": 7.80583300039952, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19538 + }, + { + "epoch": 7.80623252097483, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19539 + }, + { + "epoch": 7.80663204155014, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19540 + }, + { + "epoch": 7.80703156212545, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19541 + }, + { + "epoch": 7.807431082700759, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 19542 + }, + { + "epoch": 7.807830603276068, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19543 + }, + { + "epoch": 7.808230123851378, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 19544 + }, + { + "epoch": 7.808629644426688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19545 + }, + { + "epoch": 7.809029165001998, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19546 + }, + { + "epoch": 7.809428685577307, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19547 + }, + { + "epoch": 7.809828206152617, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19548 + }, + { + "epoch": 7.810227726727927, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19549 + }, + { + "epoch": 7.810627247303236, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 19550 + }, + { + "epoch": 7.811026767878546, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19551 + }, + { + "epoch": 7.8114262884538554, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 19552 + }, + { + "epoch": 7.811825809029165, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 19553 + }, + { + "epoch": 7.812225329604475, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 19554 + }, + { + "epoch": 7.812624850179784, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 19555 + }, + { + "epoch": 7.813024370755094, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19556 + }, + { + "epoch": 7.8134238913304035, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19557 + }, + { + "epoch": 7.813823411905713, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 19558 + }, + { + "epoch": 7.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19559 + }, + { + "epoch": 7.814622453056332, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19560 + }, + { + "epoch": 7.815021973631642, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19561 + }, + { + "epoch": 7.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19562 + }, + { + "epoch": 7.815821014782261, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19563 + }, + { + "epoch": 7.816220535357571, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19564 + }, + { + "epoch": 7.816620055932881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19565 + }, + { + "epoch": 7.81701957650819, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19566 + }, + { + "epoch": 7.8174190970835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19567 + }, + { + "epoch": 7.817818617658809, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19568 + }, + { + "epoch": 7.818218138234119, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19569 + }, + { + "epoch": 7.818617658809429, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19570 + }, + { + "epoch": 7.819017179384739, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19571 + }, + { + "epoch": 7.819416699960048, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19572 + }, + { + "epoch": 7.819816220535357, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19573 + }, + { + "epoch": 7.820215741110667, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19574 + }, + { + "epoch": 7.820615261685977, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19575 + }, + { + "epoch": 7.821014782261287, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19576 + }, + { + "epoch": 7.821414302836596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19577 + }, + { + "epoch": 7.8218138234119055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19578 + }, + { + "epoch": 7.822213343987215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19579 + }, + { + "epoch": 7.822612864562525, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19580 + }, + { + "epoch": 7.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19581 + }, + { + "epoch": 7.823411905713145, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 19582 + }, + { + "epoch": 7.8238114262884535, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19583 + }, + { + "epoch": 7.824210946863763, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 19584 + }, + { + "epoch": 7.824610467439073, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19585 + }, + { + "epoch": 7.825009988014383, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19586 + }, + { + "epoch": 7.825409508589693, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 19587 + }, + { + "epoch": 7.825809029165002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19588 + }, + { + "epoch": 7.826208549740311, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19589 + }, + { + "epoch": 7.826608070315621, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19590 + }, + { + "epoch": 7.827007590890931, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19591 + }, + { + "epoch": 7.827407111466241, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19592 + }, + { + "epoch": 7.8278066320415505, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19593 + }, + { + "epoch": 7.828206152616859, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19594 + }, + { + "epoch": 7.828605673192169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19595 + }, + { + "epoch": 7.829005193767479, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19596 + }, + { + "epoch": 7.829404714342789, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19597 + }, + { + "epoch": 7.8298042349180985, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19598 + }, + { + "epoch": 7.830203755493407, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19599 + }, + { + "epoch": 7.830603276068717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19600 + }, + { + "epoch": 7.831002796644027, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19601 + }, + { + "epoch": 7.831402317219337, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19602 + }, + { + "epoch": 7.831801837794647, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19603 + }, + { + "epoch": 7.832201358369956, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 19604 + }, + { + "epoch": 7.832600878945266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19605 + }, + { + "epoch": 7.833000399520575, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19606 + }, + { + "epoch": 7.833399920095885, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19607 + }, + { + "epoch": 7.833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19608 + }, + { + "epoch": 7.834198961246504, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 19609 + }, + { + "epoch": 7.834598481821814, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19610 + }, + { + "epoch": 7.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19611 + }, + { + "epoch": 7.835397522972433, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19612 + }, + { + "epoch": 7.835797043547743, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19613 + }, + { + "epoch": 7.836196564123052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19614 + }, + { + "epoch": 7.836596084698362, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19615 + }, + { + "epoch": 7.836995605273671, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19616 + }, + { + "epoch": 7.837395125848981, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19617 + }, + { + "epoch": 7.837794646424291, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19618 + }, + { + "epoch": 7.8381941669996005, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19619 + }, + { + "epoch": 7.83859368757491, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19620 + }, + { + "epoch": 7.83899320815022, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19621 + }, + { + "epoch": 7.83939272872553, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19622 + }, + { + "epoch": 7.839792249300839, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19623 + }, + { + "epoch": 7.8401917698761485, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19624 + }, + { + "epoch": 7.840591290451458, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19625 + }, + { + "epoch": 7.840990811026768, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19626 + }, + { + "epoch": 7.841390331602078, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19627 + }, + { + "epoch": 7.841789852177387, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19628 + }, + { + "epoch": 7.842189372752697, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19629 + }, + { + "epoch": 7.842588893328006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 19630 + }, + { + "epoch": 7.842988413903316, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19631 + }, + { + "epoch": 7.843387934478626, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 19632 + }, + { + "epoch": 7.843787455053935, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19633 + }, + { + "epoch": 7.844186975629245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 19634 + }, + { + "epoch": 7.844586496204554, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19635 + }, + { + "epoch": 7.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19636 + }, + { + "epoch": 7.845385537355174, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 19637 + }, + { + "epoch": 7.845785057930484, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19638 + }, + { + "epoch": 7.8461845785057935, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19639 + }, + { + "epoch": 7.846584099081102, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19640 + }, + { + "epoch": 7.846983619656412, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19641 + }, + { + "epoch": 7.847383140231722, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19642 + }, + { + "epoch": 7.847782660807032, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19643 + }, + { + "epoch": 7.848182181382342, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19644 + }, + { + "epoch": 7.8485817019576505, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19645 + }, + { + "epoch": 7.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19646 + }, + { + "epoch": 7.84938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19647 + }, + { + "epoch": 7.84978026368358, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19648 + }, + { + "epoch": 7.85017978425889, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19649 + }, + { + "epoch": 7.8505793048341985, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19650 + }, + { + "epoch": 7.850978825409508, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19651 + }, + { + "epoch": 7.851378345984818, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19652 + }, + { + "epoch": 7.851777866560128, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19653 + }, + { + "epoch": 7.852177387135438, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 19654 + }, + { + "epoch": 7.8525769077107475, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19655 + }, + { + "epoch": 7.852976428286056, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19656 + }, + { + "epoch": 7.853375948861366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19657 + }, + { + "epoch": 7.853775469436676, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19658 + }, + { + "epoch": 7.854174990011986, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19659 + }, + { + "epoch": 7.8545745105872955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19660 + }, + { + "epoch": 7.854974031162605, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19661 + }, + { + "epoch": 7.855373551737914, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19662 + }, + { + "epoch": 7.855773072313224, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19663 + }, + { + "epoch": 7.856172592888534, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 19664 + }, + { + "epoch": 7.8565721134638435, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19665 + }, + { + "epoch": 7.856971634039153, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19666 + }, + { + "epoch": 7.857371154614462, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 19667 + }, + { + "epoch": 7.857770675189772, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 19668 + }, + { + "epoch": 7.858170195765082, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19669 + }, + { + "epoch": 7.858569716340392, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 19670 + }, + { + "epoch": 7.858969236915701, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19671 + }, + { + "epoch": 7.859368757491011, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19672 + }, + { + "epoch": 7.85976827806632, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19673 + }, + { + "epoch": 7.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19674 + }, + { + "epoch": 7.86056731921694, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19675 + }, + { + "epoch": 7.860966839792249, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19676 + }, + { + "epoch": 7.861366360367559, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19677 + }, + { + "epoch": 7.861765880942869, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19678 + }, + { + "epoch": 7.862165401518178, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 19679 + }, + { + "epoch": 7.862564922093488, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19680 + }, + { + "epoch": 7.8629644426687975, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19681 + }, + { + "epoch": 7.863363963244107, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 19682 + }, + { + "epoch": 7.863763483819417, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19683 + }, + { + "epoch": 7.864163004394726, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19684 + }, + { + "epoch": 7.864562524970036, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19685 + }, + { + "epoch": 7.8649620455453455, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19686 + }, + { + "epoch": 7.865361566120655, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19687 + }, + { + "epoch": 7.865761086695965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19688 + }, + { + "epoch": 7.866160607271274, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19689 + }, + { + "epoch": 7.866560127846584, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19690 + }, + { + "epoch": 7.8669596484218935, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19691 + }, + { + "epoch": 7.867359168997203, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19692 + }, + { + "epoch": 7.867758689572513, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19693 + }, + { + "epoch": 7.868158210147823, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 19694 + }, + { + "epoch": 7.868557730723133, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19695 + }, + { + "epoch": 7.868957251298442, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19696 + }, + { + "epoch": 7.869356771873751, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 19697 + }, + { + "epoch": 7.869756292449061, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19698 + }, + { + "epoch": 7.870155813024371, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19699 + }, + { + "epoch": 7.870555333599681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19700 + }, + { + "epoch": 7.87095485417499, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19701 + }, + { + "epoch": 7.871354374750299, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19702 + }, + { + "epoch": 7.871753895325609, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19703 + }, + { + "epoch": 7.872153415900919, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19704 + }, + { + "epoch": 7.872552936476229, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19705 + }, + { + "epoch": 7.872952457051538, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19706 + }, + { + "epoch": 7.8733519776268475, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 19707 + }, + { + "epoch": 7.873751498202157, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19708 + }, + { + "epoch": 7.874151018777467, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19709 + }, + { + "epoch": 7.874550539352777, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19710 + }, + { + "epoch": 7.874950059928087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19711 + }, + { + "epoch": 7.875349580503396, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19712 + }, + { + "epoch": 7.875749101078705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 19713 + }, + { + "epoch": 7.876148621654015, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19714 + }, + { + "epoch": 7.876548142229325, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19715 + }, + { + "epoch": 7.876947662804635, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19716 + }, + { + "epoch": 7.877347183379944, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19717 + }, + { + "epoch": 7.877746703955253, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19718 + }, + { + "epoch": 7.878146224530563, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19719 + }, + { + "epoch": 7.878545745105873, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 19720 + }, + { + "epoch": 7.878945265681183, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19721 + }, + { + "epoch": 7.8793447862564925, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19722 + }, + { + "epoch": 7.879744306831801, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19723 + }, + { + "epoch": 7.880143827407111, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19724 + }, + { + "epoch": 7.880543347982421, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 19725 + }, + { + "epoch": 7.880942868557731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19726 + }, + { + "epoch": 7.8813423891330405, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19727 + }, + { + "epoch": 7.88174190970835, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19728 + }, + { + "epoch": 7.88214143028366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19729 + }, + { + "epoch": 7.882540950858969, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19730 + }, + { + "epoch": 7.882940471434279, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19731 + }, + { + "epoch": 7.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19732 + }, + { + "epoch": 7.883739512584898, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19733 + }, + { + "epoch": 7.884139033160208, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19734 + }, + { + "epoch": 7.884538553735517, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19735 + }, + { + "epoch": 7.884938074310827, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19736 + }, + { + "epoch": 7.885337594886137, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19737 + }, + { + "epoch": 7.885737115461446, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19738 + }, + { + "epoch": 7.886136636036756, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19739 + }, + { + "epoch": 7.886536156612065, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19740 + }, + { + "epoch": 7.886935677187375, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 19741 + }, + { + "epoch": 7.887335197762685, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 19742 + }, + { + "epoch": 7.887734718337994, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19743 + }, + { + "epoch": 7.888134238913304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19744 + }, + { + "epoch": 7.888533759488614, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19745 + }, + { + "epoch": 7.888933280063923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19746 + }, + { + "epoch": 7.889332800639233, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19747 + }, + { + "epoch": 7.8897323212145425, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19748 + }, + { + "epoch": 7.890131841789852, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19749 + }, + { + "epoch": 7.890531362365162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19750 + } + ], + "logging_steps": 1.0, + "max_steps": 250300, + "num_input_tokens_seen": 0, + "num_train_epochs": 100, + "save_steps": 250, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 3.9702998923115233e+18, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-19750/training_args.bin b/checkpoint-19750/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2bc6e8c6a7b45b79e23a7e4b282dcb7f0e746bc4 --- /dev/null +++ b/checkpoint-19750/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9307aa53030dab795f2b0f5373487f7b4270f37adff792b3db36f71bcb8dae52 +size 5816 diff --git a/checkpoint-20000/config.json b/checkpoint-20000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fe7dbf7d28e62328ecc2df097527bd51ec161f9d --- /dev/null +++ b/checkpoint-20000/config.json @@ -0,0 +1,73 @@ +{ + "ar_steps": 1, + "architectures": [ + "DiffVLMDiffusion" + ], + "attention_dropout": 0.0, + "bos_token_id": 151643, + "condition_layer": -1, + "eos_token_id": 151645, + "hidden_act": "silu", + "hidden_size": 1536, + "image_token_id": 151655, + "img_cross_attention_dim": 2048, + "img_diffuser_depth": 6, + "img_ffn_dim_multiplier": null, + "img_hidden_size": 1536, + "img_multiple_of": 256, + "img_norm_eps": 1e-05, + "img_num_attention_heads": 12, + "img_num_kv_heads": 12, + "img_qk_norm": true, + "in_channels": 512, + "initializer_range": 0.02, + "inject_img_diffuser": false, + "input_size": 64, + "intermediate_size": 8960, + "layerwise_start_idx": 21, + "lora_alpha": 128, + "lora_bias": "none", + "lora_dropout": 0.05, + "lora_enable": false, + "lora_r": 64, + "max_position_embeddings": 32768, + "max_window_layers": 28, + "model_type": "qwen2_vl", + "non_linearity": 1, + "norm_elementwise_affine": true, + "num_attention_heads": 12, + "num_hidden_layers": 28, + "num_key_value_heads": 2, + "patch_size": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": { + "mrope_section": [ + 16, + 24, + 24 + ], + "rope_type": "default", + "type": "default" + }, + "rope_theta": 1000000.0, + "sample_size": 128, + "sampling_steps": 7, + "sliding_window": null, + "tie_word_embeddings": true, + "torch_dtype": "bfloat16", + "transformers_version": "4.47.0", + "use_cache": true, + "use_sliding_window": false, + "vae_path": "mit-han-lab/dc-ae-f128c512-mix-1.0-diffusers", + "video_token_id": 151656, + "vision_config": { + "hidden_size": 1536, + "in_chans": 3, + "model_type": "qwen2_vl", + "spatial_patch_size": 14 + }, + "vision_end_token_id": 151653, + "vision_start_token_id": 151652, + "vision_token_id": 151654, + "vocab_size": 151936 +} diff --git a/checkpoint-20000/generation_config.json b/checkpoint-20000/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e3d465da90cb0ab59d8ba7babdefb0e88fbfa6b --- /dev/null +++ b/checkpoint-20000/generation_config.json @@ -0,0 +1,6 @@ +{ + "_from_model_config": true, + "bos_token_id": 151643, + "eos_token_id": 151645, + "transformers_version": "4.47.0" +} diff --git a/checkpoint-20000/model-00001-of-00002.safetensors b/checkpoint-20000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..24e98903f10540e5e015b23ec69b814153bfe44e --- /dev/null +++ b/checkpoint-20000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66bc74c852bff2719622a7e6018297f8f9b8a57a7413f7ff84062e8cf5f2d89d +size 4754815448 diff --git a/checkpoint-20000/model-00002-of-00002.safetensors b/checkpoint-20000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c76cf00e7a9c7ce6c6d8a7a347afcbf18b661e2d --- /dev/null +++ b/checkpoint-20000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7d096d5260908edb86a73ee04889ac23d52b4054c4adabbc4a0e6ec76854d6f +size 840447856 diff --git a/checkpoint-20000/model.safetensors.index.json b/checkpoint-20000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea3b7f98260439a7c863a0290604b077271f337 --- /dev/null +++ b/checkpoint-20000/model.safetensors.index.json @@ -0,0 +1,1076 @@ +{ + "metadata": { + "total_size": 5595140920 + }, + "weight_map": { + "embed_tokens.weight": "model-00001-of-00002.safetensors", + "img2text.bias": "model-00001-of-00002.safetensors", + "img2text.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_1.weight": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.bias": "model-00001-of-00002.safetensors", + "img_norm_out.linear_2.weight": "model-00001-of-00002.safetensors", + "layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.gate": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.21.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.21.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.21.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.22.gate": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.22.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.22.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.22.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.23.gate": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.23.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.23.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.23.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.24.gate": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.24.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.24.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.24.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.25.gate": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.25.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.25.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.25.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.25.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.26.gate": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.26.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.26.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.26.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.26.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.27.gate": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.bias": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.norm_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_k.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_q.weight": "model-00001-of-00002.safetensors", + "layers.27.img_attn1.to_v.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_ffn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_post_mixed_attn_norm.weight": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.bias": "model-00001-of-00002.safetensors", + "layers.27.img_scale_shift.linear.weight": "model-00001-of-00002.safetensors", + "layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.down_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.gate_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.mlp.up_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.k_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.o_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.q_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_A.weight": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.bias": "model-00001-of-00002.safetensors", + "layers.27.self_attn.v_proj_lora.lora_B.weight": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.bias": "model-00001-of-00002.safetensors", + "layers.27.to_img_diffuser.weight": "model-00001-of-00002.safetensors", + "layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "norm.weight": "model-00001-of-00002.safetensors", + "patch_embedder.proj.bias": "model-00001-of-00002.safetensors", + "patch_embedder.proj.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "t_embedder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.bias": "model-00001-of-00002.safetensors", + "vae.decoder.conv_in.weight": "model-00001-of-00002.safetensors", + "vae.decoder.conv_out.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.conv_out.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.1.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.10.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.8.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv1.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.conv2.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.num_batches_tracked": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_mean": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.running_var": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.2.9.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.3.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.4.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.0.conv.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.5.2.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.0.conv.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.6.2.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.0.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.norm_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_k.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_out.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_q.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.attn.to_v.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_depth.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_inverted.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.conv_point.weight": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.bias": "model-00002-of-00002.safetensors", + "vae.decoder.up_blocks.7.1.conv_out.norm.weight": "model-00002-of-00002.safetensors", + "vae.encoder.conv_in.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_in.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.conv_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.1.4.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.0.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.1.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.2.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.3.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.4.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.5.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.6.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv1.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.conv2.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.7.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.2.8.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.3.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.4.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.5.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.1.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.6.2.conv.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.0.conv_out.norm.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.norm_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_k.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_out.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_q.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.attn.to_v.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_depth.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_inverted.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.conv_point.weight": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.bias": "model-00001-of-00002.safetensors", + "vae.encoder.down_blocks.7.1.conv_out.norm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-20000/optimizer.pt b/checkpoint-20000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..9f7f3d32f3b2f0d3310d36c431214e32cea14ce6 --- /dev/null +++ b/checkpoint-20000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de204890721430a565a608348be42290f53cd9598ccd9249a62ec2c3067ea918 +size 5620161007 diff --git a/checkpoint-20000/rng_state_0.pth b/checkpoint-20000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..fbb079ff9553fbf9c874f8a3c5b0fb185af5562c --- /dev/null +++ b/checkpoint-20000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e9c95ccce3e7cda85e106c217820b9e13a20cf475996ac9ba3f9f5a6b79b670 +size 14960 diff --git a/checkpoint-20000/rng_state_1.pth b/checkpoint-20000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..3ec9fe9b2c93c2606f2939168290e99fa9e2c2f1 --- /dev/null +++ b/checkpoint-20000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:939fbc4030458dddc66d9014a34f9b2d0f7f817de1c177d8dc7a1304ab179454 +size 14960 diff --git a/checkpoint-20000/rng_state_2.pth b/checkpoint-20000/rng_state_2.pth new file mode 100644 index 0000000000000000000000000000000000000000..3f01e1bda9ef36704f7e22088856bf9467b0b324 --- /dev/null +++ b/checkpoint-20000/rng_state_2.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47aee5c53151bb439e2ac49de4d16fe759e7d14b4c55d76dbbdcb41aad005669 +size 14960 diff --git a/checkpoint-20000/rng_state_3.pth b/checkpoint-20000/rng_state_3.pth new file mode 100644 index 0000000000000000000000000000000000000000..5e4a1fab41c8ee001ca2553aa0c4195ba14a3afa --- /dev/null +++ b/checkpoint-20000/rng_state_3.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78eda574fa79f8de0cd6ccf00e9a38276ec91a85f5e4c88a0fa29ea2de36148b +size 14960 diff --git a/checkpoint-20000/scheduler.pt b/checkpoint-20000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..5745c92304307e6e3b27470e1526594c7a62aa52 --- /dev/null +++ b/checkpoint-20000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f14362361f7f2e0c1f7b0ffd27e3afe8874bec255213e18ad6a52ee2a8470bff +size 1064 diff --git a/checkpoint-20000/trainer_state.json b/checkpoint-20000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..b7b01ef84459b272090dc1411e154e1f54c025c7 --- /dev/null +++ b/checkpoint-20000/trainer_state.json @@ -0,0 +1,140033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.990411506192569, + "eval_steps": 500, + "global_step": 20000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00039952057530962844, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 3.1824, + "step": 1 + }, + { + "epoch": 0.0007990411506192569, + "grad_norm": 0.494140625, + "learning_rate": 0.0002, + "loss": 2.9778, + "step": 2 + }, + { + "epoch": 0.0011985617259288853, + "grad_norm": 0.451171875, + "learning_rate": 0.0002, + "loss": 2.7824, + "step": 3 + }, + { + "epoch": 0.0015980823012385138, + "grad_norm": 0.478515625, + "learning_rate": 0.0002, + "loss": 2.6074, + "step": 4 + }, + { + "epoch": 0.001997602876548142, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 2.4209, + "step": 5 + }, + { + "epoch": 0.0023971234518577705, + "grad_norm": 0.4921875, + "learning_rate": 0.0002, + "loss": 2.2482, + "step": 6 + }, + { + "epoch": 0.002796644027167399, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 2.1017, + "step": 7 + }, + { + "epoch": 0.0031961646024770275, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 2.0034, + "step": 8 + }, + { + "epoch": 0.003595685177786656, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.9567, + "step": 9 + }, + { + "epoch": 0.003995205753096284, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.9417, + "step": 10 + }, + { + "epoch": 0.0043947263284059125, + "grad_norm": 0.345703125, + "learning_rate": 0.0002, + "loss": 1.9526, + "step": 11 + }, + { + "epoch": 0.004794246903715541, + "grad_norm": 0.427734375, + "learning_rate": 0.0002, + "loss": 1.9613, + "step": 12 + }, + { + "epoch": 0.0051937674790251695, + "grad_norm": 0.32421875, + "learning_rate": 0.0002, + "loss": 1.9467, + "step": 13 + }, + { + "epoch": 0.005593288054334798, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.9215, + "step": 14 + }, + { + "epoch": 0.0059928086296444265, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.9154, + "step": 15 + }, + { + "epoch": 0.006392329204954055, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.9064, + "step": 16 + }, + { + "epoch": 0.0067918497802636835, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.888, + "step": 17 + }, + { + "epoch": 0.007191370355573312, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.8729, + "step": 18 + }, + { + "epoch": 0.0075908909308829405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8425, + "step": 19 + }, + { + "epoch": 0.007990411506192568, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.8468, + "step": 20 + }, + { + "epoch": 0.008389932081502197, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.8418, + "step": 21 + }, + { + "epoch": 0.008789452656811825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.825, + "step": 22 + }, + { + "epoch": 0.009188973232121454, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.8176, + "step": 23 + }, + { + "epoch": 0.009588493807431082, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.7827, + "step": 24 + }, + { + "epoch": 0.00998801438274071, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.7695, + "step": 25 + }, + { + "epoch": 0.010387534958050339, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.78, + "step": 26 + }, + { + "epoch": 0.010787055533359968, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.7588, + "step": 27 + }, + { + "epoch": 0.011186576108669596, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.743, + "step": 28 + }, + { + "epoch": 0.011586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.7357, + "step": 29 + }, + { + "epoch": 0.011985617259288853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.7271, + "step": 30 + }, + { + "epoch": 0.012385137834598482, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.7053, + "step": 31 + }, + { + "epoch": 0.01278465840990811, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.7033, + "step": 32 + }, + { + "epoch": 0.013184178985217739, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.6831, + "step": 33 + }, + { + "epoch": 0.013583699560527367, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.6882, + "step": 34 + }, + { + "epoch": 0.013983220135836995, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6682, + "step": 35 + }, + { + "epoch": 0.014382740711146624, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.6573, + "step": 36 + }, + { + "epoch": 0.014782261286456252, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.6507, + "step": 37 + }, + { + "epoch": 0.015181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.6424, + "step": 38 + }, + { + "epoch": 0.01558130243707551, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.6251, + "step": 39 + }, + { + "epoch": 0.015980823012385136, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.6215, + "step": 40 + }, + { + "epoch": 0.016380343587694766, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.6226, + "step": 41 + }, + { + "epoch": 0.016779864163004393, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.6005, + "step": 42 + }, + { + "epoch": 0.017179384738314023, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.6224, + "step": 43 + }, + { + "epoch": 0.01757890531362365, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5901, + "step": 44 + }, + { + "epoch": 0.01797842588893328, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5962, + "step": 45 + }, + { + "epoch": 0.018377946464242907, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.5895, + "step": 46 + }, + { + "epoch": 0.018777467039552537, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.5697, + "step": 47 + }, + { + "epoch": 0.019176987614862164, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.567, + "step": 48 + }, + { + "epoch": 0.019576508190171794, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.5531, + "step": 49 + }, + { + "epoch": 0.01997602876548142, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.554, + "step": 50 + }, + { + "epoch": 0.02037554934079105, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.5539, + "step": 51 + }, + { + "epoch": 0.020775069916100678, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.5591, + "step": 52 + }, + { + "epoch": 0.02117459049141031, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.545, + "step": 53 + }, + { + "epoch": 0.021574111066719935, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.5221, + "step": 54 + }, + { + "epoch": 0.021973631642029565, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.5234, + "step": 55 + }, + { + "epoch": 0.022373152217339192, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.5186, + "step": 56 + }, + { + "epoch": 0.022772672792648822, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.5273, + "step": 57 + }, + { + "epoch": 0.02317219336795845, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5201, + "step": 58 + }, + { + "epoch": 0.02357171394326808, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.5108, + "step": 59 + }, + { + "epoch": 0.023971234518577706, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.5078, + "step": 60 + }, + { + "epoch": 0.024370755093887336, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.5026, + "step": 61 + }, + { + "epoch": 0.024770275669196963, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.5076, + "step": 62 + }, + { + "epoch": 0.025169796244506593, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4956, + "step": 63 + }, + { + "epoch": 0.02556931681981622, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4925, + "step": 64 + }, + { + "epoch": 0.02596883739512585, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.5045, + "step": 65 + }, + { + "epoch": 0.026368357970435477, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.4843, + "step": 66 + }, + { + "epoch": 0.026767878545745107, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4807, + "step": 67 + }, + { + "epoch": 0.027167399121054734, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4763, + "step": 68 + }, + { + "epoch": 0.027566919696364364, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4784, + "step": 69 + }, + { + "epoch": 0.02796644027167399, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4758, + "step": 70 + }, + { + "epoch": 0.02836596084698362, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4871, + "step": 71 + }, + { + "epoch": 0.028765481422293248, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4753, + "step": 72 + }, + { + "epoch": 0.029165001997602878, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4922, + "step": 73 + }, + { + "epoch": 0.029564522572912505, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4759, + "step": 74 + }, + { + "epoch": 0.02996404314822213, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.4703, + "step": 75 + }, + { + "epoch": 0.030363563723531762, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4666, + "step": 76 + }, + { + "epoch": 0.03076308429884139, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4628, + "step": 77 + }, + { + "epoch": 0.03116260487415102, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.4656, + "step": 78 + }, + { + "epoch": 0.031562125449460646, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.4695, + "step": 79 + }, + { + "epoch": 0.03196164602477027, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4517, + "step": 80 + }, + { + "epoch": 0.032361166600079906, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4577, + "step": 81 + }, + { + "epoch": 0.03276068717538953, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4526, + "step": 82 + }, + { + "epoch": 0.03316020775069916, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4476, + "step": 83 + }, + { + "epoch": 0.033559728326008786, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.4499, + "step": 84 + }, + { + "epoch": 0.03395924890131842, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.4549, + "step": 85 + }, + { + "epoch": 0.03435876947662805, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4548, + "step": 86 + }, + { + "epoch": 0.034758290051937674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4525, + "step": 87 + }, + { + "epoch": 0.0351578106272473, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4514, + "step": 88 + }, + { + "epoch": 0.035557331202556934, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4343, + "step": 89 + }, + { + "epoch": 0.03595685177786656, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.4398, + "step": 90 + }, + { + "epoch": 0.03635637235317619, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4431, + "step": 91 + }, + { + "epoch": 0.036755892928485814, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.438, + "step": 92 + }, + { + "epoch": 0.03715541350379545, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4291, + "step": 93 + }, + { + "epoch": 0.037554934079105075, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.4373, + "step": 94 + }, + { + "epoch": 0.0379544546544147, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4409, + "step": 95 + }, + { + "epoch": 0.03835397522972433, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4305, + "step": 96 + }, + { + "epoch": 0.03875349580503396, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.4337, + "step": 97 + }, + { + "epoch": 0.03915301638034359, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4309, + "step": 98 + }, + { + "epoch": 0.039552536955653216, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.439, + "step": 99 + }, + { + "epoch": 0.03995205753096284, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 100 + }, + { + "epoch": 0.040351578106272476, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.42, + "step": 101 + }, + { + "epoch": 0.0407510986815821, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4211, + "step": 102 + }, + { + "epoch": 0.04115061925689173, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.4342, + "step": 103 + }, + { + "epoch": 0.041550139832201356, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4191, + "step": 104 + }, + { + "epoch": 0.04194966040751099, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.4234, + "step": 105 + }, + { + "epoch": 0.04234918098282062, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 106 + }, + { + "epoch": 0.042748701558130243, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.4308, + "step": 107 + }, + { + "epoch": 0.04314822213343987, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.4189, + "step": 108 + }, + { + "epoch": 0.043547742708749504, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.432, + "step": 109 + }, + { + "epoch": 0.04394726328405913, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4278, + "step": 110 + }, + { + "epoch": 0.04434678385936876, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 111 + }, + { + "epoch": 0.044746304434678384, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.4276, + "step": 112 + }, + { + "epoch": 0.04514582500998801, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.4203, + "step": 113 + }, + { + "epoch": 0.045545345585297645, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4313, + "step": 114 + }, + { + "epoch": 0.04594486616060727, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4186, + "step": 115 + }, + { + "epoch": 0.0463443867359169, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.4215, + "step": 116 + }, + { + "epoch": 0.046743907311226525, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.4272, + "step": 117 + }, + { + "epoch": 0.04714342788653616, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4262, + "step": 118 + }, + { + "epoch": 0.047542948461845785, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.416, + "step": 119 + }, + { + "epoch": 0.04794246903715541, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4071, + "step": 120 + }, + { + "epoch": 0.04834198961246504, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4241, + "step": 121 + }, + { + "epoch": 0.04874151018777467, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4244, + "step": 122 + }, + { + "epoch": 0.0491410307630843, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4252, + "step": 123 + }, + { + "epoch": 0.049540551338393926, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.424, + "step": 124 + }, + { + "epoch": 0.04994007191370355, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 125 + }, + { + "epoch": 0.05033959248901319, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.412, + "step": 126 + }, + { + "epoch": 0.05073911306432281, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4199, + "step": 127 + }, + { + "epoch": 0.05113863363963244, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4154, + "step": 128 + }, + { + "epoch": 0.05153815421494207, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 129 + }, + { + "epoch": 0.0519376747902517, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 130 + }, + { + "epoch": 0.05233719536556133, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4049, + "step": 131 + }, + { + "epoch": 0.052736715940870954, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4134, + "step": 132 + }, + { + "epoch": 0.05313623651618058, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 133 + }, + { + "epoch": 0.053535757091490214, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.4156, + "step": 134 + }, + { + "epoch": 0.05393527766679984, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 135 + }, + { + "epoch": 0.05433479824210947, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.415, + "step": 136 + }, + { + "epoch": 0.054734318817419095, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4155, + "step": 137 + }, + { + "epoch": 0.05513383939272873, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 138 + }, + { + "epoch": 0.055533359968038355, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.428, + "step": 139 + }, + { + "epoch": 0.05593288054334798, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 140 + }, + { + "epoch": 0.05633240111865761, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.4083, + "step": 141 + }, + { + "epoch": 0.05673192169396724, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.4147, + "step": 142 + }, + { + "epoch": 0.05713144226927687, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 143 + }, + { + "epoch": 0.057530962844586496, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4257, + "step": 144 + }, + { + "epoch": 0.05793048341989612, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 145 + }, + { + "epoch": 0.058330003995205756, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 146 + }, + { + "epoch": 0.05872952457051538, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.4168, + "step": 147 + }, + { + "epoch": 0.05912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 148 + }, + { + "epoch": 0.05952856572113464, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 149 + }, + { + "epoch": 0.05992808629644426, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4176, + "step": 150 + }, + { + "epoch": 0.0603276068717539, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4171, + "step": 151 + }, + { + "epoch": 0.060727127447063524, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.4151, + "step": 152 + }, + { + "epoch": 0.06112664802237315, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 153 + }, + { + "epoch": 0.06152616859768278, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3966, + "step": 154 + }, + { + "epoch": 0.06192568917299241, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.4121, + "step": 155 + }, + { + "epoch": 0.06232520974830204, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.4103, + "step": 156 + }, + { + "epoch": 0.06272473032361167, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 157 + }, + { + "epoch": 0.06312425089892129, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 158 + }, + { + "epoch": 0.06352377147423093, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.413, + "step": 159 + }, + { + "epoch": 0.06392329204954054, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4174, + "step": 160 + }, + { + "epoch": 0.06432281262485018, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4133, + "step": 161 + }, + { + "epoch": 0.06472233320015981, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 162 + }, + { + "epoch": 0.06512185377546943, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.4128, + "step": 163 + }, + { + "epoch": 0.06552137435077907, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3945, + "step": 164 + }, + { + "epoch": 0.0659208949260887, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 165 + }, + { + "epoch": 0.06632041550139832, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 166 + }, + { + "epoch": 0.06671993607670795, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.4104, + "step": 167 + }, + { + "epoch": 0.06711945665201757, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.4144, + "step": 168 + }, + { + "epoch": 0.0675189772273272, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4059, + "step": 169 + }, + { + "epoch": 0.06791849780263684, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.4091, + "step": 170 + }, + { + "epoch": 0.06831801837794646, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 171 + }, + { + "epoch": 0.0687175389532561, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3993, + "step": 172 + }, + { + "epoch": 0.06911705952856573, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 173 + }, + { + "epoch": 0.06951658010387535, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 174 + }, + { + "epoch": 0.06991610067918498, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4097, + "step": 175 + }, + { + "epoch": 0.0703156212544946, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.4227, + "step": 176 + }, + { + "epoch": 0.07071514182980423, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.4052, + "step": 177 + }, + { + "epoch": 0.07111466240511387, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.417, + "step": 178 + }, + { + "epoch": 0.07151418298042349, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 179 + }, + { + "epoch": 0.07191370355573312, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4105, + "step": 180 + }, + { + "epoch": 0.07231322413104276, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4047, + "step": 181 + }, + { + "epoch": 0.07271274470635238, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 182 + }, + { + "epoch": 0.07311226528166201, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 183 + }, + { + "epoch": 0.07351178585697163, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 184 + }, + { + "epoch": 0.07391130643228126, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.4, + "step": 185 + }, + { + "epoch": 0.0743108270075909, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 186 + }, + { + "epoch": 0.07471034758290052, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 187 + }, + { + "epoch": 0.07510986815821015, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 188 + }, + { + "epoch": 0.07550938873351977, + "grad_norm": 0.1591796875, + "learning_rate": 0.0002, + "loss": 1.4132, + "step": 189 + }, + { + "epoch": 0.0759089093088294, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.4067, + "step": 190 + }, + { + "epoch": 0.07630842988413904, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 191 + }, + { + "epoch": 0.07670795045944866, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 192 + }, + { + "epoch": 0.07710747103475829, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3964, + "step": 193 + }, + { + "epoch": 0.07750699161006792, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 194 + }, + { + "epoch": 0.07790651218537754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3979, + "step": 195 + }, + { + "epoch": 0.07830603276068718, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 196 + }, + { + "epoch": 0.0787055533359968, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.4111, + "step": 197 + }, + { + "epoch": 0.07910507391130643, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 198 + }, + { + "epoch": 0.07950459448661606, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 199 + }, + { + "epoch": 0.07990411506192568, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 200 + }, + { + "epoch": 0.08030363563723532, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 201 + }, + { + "epoch": 0.08070315621254495, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.399, + "step": 202 + }, + { + "epoch": 0.08110267678785457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4021, + "step": 203 + }, + { + "epoch": 0.0815021973631642, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 204 + }, + { + "epoch": 0.08190171793847383, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 205 + }, + { + "epoch": 0.08230123851378346, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 206 + }, + { + "epoch": 0.08270075908909309, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.4057, + "step": 207 + }, + { + "epoch": 0.08310027966440271, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4158, + "step": 208 + }, + { + "epoch": 0.08349980023971235, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3898, + "step": 209 + }, + { + "epoch": 0.08389932081502198, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 210 + }, + { + "epoch": 0.0842988413903316, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.4193, + "step": 211 + }, + { + "epoch": 0.08469836196564123, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.4026, + "step": 212 + }, + { + "epoch": 0.08509788254095085, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 213 + }, + { + "epoch": 0.08549740311626049, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4039, + "step": 214 + }, + { + "epoch": 0.08589692369157012, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4098, + "step": 215 + }, + { + "epoch": 0.08629644426687974, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.4015, + "step": 216 + }, + { + "epoch": 0.08669596484218937, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.4113, + "step": 217 + }, + { + "epoch": 0.08709548541749901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 218 + }, + { + "epoch": 0.08749500599280863, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.4081, + "step": 219 + }, + { + "epoch": 0.08789452656811826, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.4138, + "step": 220 + }, + { + "epoch": 0.08829404714342788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 221 + }, + { + "epoch": 0.08869356771873751, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.409, + "step": 222 + }, + { + "epoch": 0.08909308829404715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 223 + }, + { + "epoch": 0.08949260886935677, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.4062, + "step": 224 + }, + { + "epoch": 0.0898921294446664, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 225 + }, + { + "epoch": 0.09029165001997602, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.4055, + "step": 226 + }, + { + "epoch": 0.09069117059528566, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3904, + "step": 227 + }, + { + "epoch": 0.09109069117059529, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3984, + "step": 228 + }, + { + "epoch": 0.09149021174590491, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.4161, + "step": 229 + }, + { + "epoch": 0.09188973232121454, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.4016, + "step": 230 + }, + { + "epoch": 0.09228925289652418, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.4088, + "step": 231 + }, + { + "epoch": 0.0926887734718338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.4076, + "step": 232 + }, + { + "epoch": 0.09308829404714343, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 233 + }, + { + "epoch": 0.09348781462245305, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 234 + }, + { + "epoch": 0.09388733519776268, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.4013, + "step": 235 + }, + { + "epoch": 0.09428685577307232, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 236 + }, + { + "epoch": 0.09468637634838194, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3975, + "step": 237 + }, + { + "epoch": 0.09508589692369157, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.4092, + "step": 238 + }, + { + "epoch": 0.0954854174990012, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 239 + }, + { + "epoch": 0.09588493807431082, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.4064, + "step": 240 + }, + { + "epoch": 0.09628445864962046, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 241 + }, + { + "epoch": 0.09668397922493008, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 242 + }, + { + "epoch": 0.09708349980023971, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 243 + }, + { + "epoch": 0.09748302037554935, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.403, + "step": 244 + }, + { + "epoch": 0.09788254095085897, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3961, + "step": 245 + }, + { + "epoch": 0.0982820615261686, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.4061, + "step": 246 + }, + { + "epoch": 0.09868158210147823, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.4025, + "step": 247 + }, + { + "epoch": 0.09908110267678785, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 248 + }, + { + "epoch": 0.09948062325209749, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 249 + }, + { + "epoch": 0.0998801438274071, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 250 + }, + { + "epoch": 0.10027966440271674, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 251 + }, + { + "epoch": 0.10067918497802637, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 252 + }, + { + "epoch": 0.10107870555333599, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.4024, + "step": 253 + }, + { + "epoch": 0.10147822612864563, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 254 + }, + { + "epoch": 0.10187774670395526, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 255 + }, + { + "epoch": 0.10227726727926488, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 256 + }, + { + "epoch": 0.10267678785457451, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 257 + }, + { + "epoch": 0.10307630842988413, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 258 + }, + { + "epoch": 0.10347582900519377, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.4065, + "step": 259 + }, + { + "epoch": 0.1038753495805034, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 260 + }, + { + "epoch": 0.10427487015581302, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 261 + }, + { + "epoch": 0.10467439073112265, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 262 + }, + { + "epoch": 0.10507391130643227, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 263 + }, + { + "epoch": 0.10547343188174191, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 264 + }, + { + "epoch": 0.10587295245705154, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3988, + "step": 265 + }, + { + "epoch": 0.10627247303236116, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.394, + "step": 266 + }, + { + "epoch": 0.1066719936076708, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.4079, + "step": 267 + }, + { + "epoch": 0.10707151418298043, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 268 + }, + { + "epoch": 0.10747103475829005, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3954, + "step": 269 + }, + { + "epoch": 0.10787055533359968, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 270 + }, + { + "epoch": 0.1082700759089093, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 271 + }, + { + "epoch": 0.10866959648421894, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 272 + }, + { + "epoch": 0.10906911705952857, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.4035, + "step": 273 + }, + { + "epoch": 0.10946863763483819, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3899, + "step": 274 + }, + { + "epoch": 0.10986815821014782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 275 + }, + { + "epoch": 0.11026767878545746, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.4012, + "step": 276 + }, + { + "epoch": 0.11066719936076708, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 277 + }, + { + "epoch": 0.11106671993607671, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3948, + "step": 278 + }, + { + "epoch": 0.11146624051138633, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.4175, + "step": 279 + }, + { + "epoch": 0.11186576108669596, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 280 + }, + { + "epoch": 0.1122652816620056, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 281 + }, + { + "epoch": 0.11266480223731522, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3994, + "step": 282 + }, + { + "epoch": 0.11306432281262485, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 283 + }, + { + "epoch": 0.11346384338793448, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.404, + "step": 284 + }, + { + "epoch": 0.1138633639632441, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 285 + }, + { + "epoch": 0.11426288453855374, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 286 + }, + { + "epoch": 0.11466240511386336, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.4045, + "step": 287 + }, + { + "epoch": 0.11506192568917299, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 288 + }, + { + "epoch": 0.11546144626448263, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3869, + "step": 289 + }, + { + "epoch": 0.11586096683979225, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 290 + }, + { + "epoch": 0.11626048741510188, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3893, + "step": 291 + }, + { + "epoch": 0.11666000799041151, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.4123, + "step": 292 + }, + { + "epoch": 0.11705952856572113, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3971, + "step": 293 + }, + { + "epoch": 0.11745904914103077, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3991, + "step": 294 + }, + { + "epoch": 0.11785856971634039, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 295 + }, + { + "epoch": 0.11825809029165002, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 296 + }, + { + "epoch": 0.11865761086695965, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 297 + }, + { + "epoch": 0.11905713144226927, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 298 + }, + { + "epoch": 0.11945665201757891, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 299 + }, + { + "epoch": 0.11985617259288853, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 300 + }, + { + "epoch": 0.12025569316819816, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3998, + "step": 301 + }, + { + "epoch": 0.1206552137435078, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 302 + }, + { + "epoch": 0.12105473431881741, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 303 + }, + { + "epoch": 0.12145425489412705, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 304 + }, + { + "epoch": 0.12185377546943668, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3999, + "step": 305 + }, + { + "epoch": 0.1222532960447463, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3949, + "step": 306 + }, + { + "epoch": 0.12265281662005593, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 307 + }, + { + "epoch": 0.12305233719536555, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 308 + }, + { + "epoch": 0.12345185777067519, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 309 + }, + { + "epoch": 0.12385137834598482, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 310 + }, + { + "epoch": 0.12425089892129444, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 311 + }, + { + "epoch": 0.12465041949660408, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3922, + "step": 312 + }, + { + "epoch": 0.1250499400719137, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 313 + }, + { + "epoch": 0.12544946064722334, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 314 + }, + { + "epoch": 0.12584898122253296, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3962, + "step": 315 + }, + { + "epoch": 0.12624850179784258, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 316 + }, + { + "epoch": 0.12664802237315223, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 317 + }, + { + "epoch": 0.12704754294846185, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 318 + }, + { + "epoch": 0.12744706352377147, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 319 + }, + { + "epoch": 0.1278465840990811, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 320 + }, + { + "epoch": 0.12824610467439074, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 321 + }, + { + "epoch": 0.12864562524970036, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3987, + "step": 322 + }, + { + "epoch": 0.12904514582500998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 323 + }, + { + "epoch": 0.12944466640031962, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 324 + }, + { + "epoch": 0.12984418697562924, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 325 + }, + { + "epoch": 0.13024370755093886, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 326 + }, + { + "epoch": 0.1306432281262485, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3886, + "step": 327 + }, + { + "epoch": 0.13104274870155813, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 328 + }, + { + "epoch": 0.13144226927686775, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3974, + "step": 329 + }, + { + "epoch": 0.1318417898521774, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 330 + }, + { + "epoch": 0.13224131042748702, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 331 + }, + { + "epoch": 0.13264083100279664, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 332 + }, + { + "epoch": 0.1330403515781063, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3932, + "step": 333 + }, + { + "epoch": 0.1334398721534159, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 334 + }, + { + "epoch": 0.13383939272872553, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3926, + "step": 335 + }, + { + "epoch": 0.13423891330403515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3952, + "step": 336 + }, + { + "epoch": 0.1346384338793448, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3911, + "step": 337 + }, + { + "epoch": 0.1350379544546544, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3953, + "step": 338 + }, + { + "epoch": 0.13543747502996403, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.402, + "step": 339 + }, + { + "epoch": 0.13583699560527368, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 340 + }, + { + "epoch": 0.1362365161805833, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3925, + "step": 341 + }, + { + "epoch": 0.13663603675589292, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3894, + "step": 342 + }, + { + "epoch": 0.13703555733120257, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3889, + "step": 343 + }, + { + "epoch": 0.1374350779065122, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3995, + "step": 344 + }, + { + "epoch": 0.1378345984818218, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 345 + }, + { + "epoch": 0.13823411905713145, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 346 + }, + { + "epoch": 0.13863363963244107, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3983, + "step": 347 + }, + { + "epoch": 0.1390331602077507, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 348 + }, + { + "epoch": 0.13943268078306031, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3981, + "step": 349 + }, + { + "epoch": 0.13983220135836996, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 350 + }, + { + "epoch": 0.14023172193367958, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3938, + "step": 351 + }, + { + "epoch": 0.1406312425089892, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 352 + }, + { + "epoch": 0.14103076308429885, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 353 + }, + { + "epoch": 0.14143028365960847, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.391, + "step": 354 + }, + { + "epoch": 0.1418298042349181, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3937, + "step": 355 + }, + { + "epoch": 0.14222932481022774, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3858, + "step": 356 + }, + { + "epoch": 0.14262884538553736, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 357 + }, + { + "epoch": 0.14302836596084698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 358 + }, + { + "epoch": 0.14342788653615662, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 359 + }, + { + "epoch": 0.14382740711146624, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 360 + }, + { + "epoch": 0.14422692768677586, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 361 + }, + { + "epoch": 0.1446264482620855, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3963, + "step": 362 + }, + { + "epoch": 0.14502596883739513, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 363 + }, + { + "epoch": 0.14542548941270475, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 364 + }, + { + "epoch": 0.14582500998801437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3908, + "step": 365 + }, + { + "epoch": 0.14622453056332402, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3895, + "step": 366 + }, + { + "epoch": 0.14662405113863364, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3985, + "step": 367 + }, + { + "epoch": 0.14702357171394326, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 368 + }, + { + "epoch": 0.1474230922892529, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 369 + }, + { + "epoch": 0.14782261286456252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 370 + }, + { + "epoch": 0.14822213343987214, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 371 + }, + { + "epoch": 0.1486216540151818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 372 + }, + { + "epoch": 0.1490211745904914, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 373 + }, + { + "epoch": 0.14942069516580103, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3942, + "step": 374 + }, + { + "epoch": 0.14982021574111068, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 375 + }, + { + "epoch": 0.1502197363164203, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3882, + "step": 376 + }, + { + "epoch": 0.15061925689172992, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3903, + "step": 377 + }, + { + "epoch": 0.15101877746703954, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 378 + }, + { + "epoch": 0.1514182980423492, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 379 + }, + { + "epoch": 0.1518178186176588, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 380 + }, + { + "epoch": 0.15221733919296843, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3931, + "step": 381 + }, + { + "epoch": 0.15261685976827807, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.393, + "step": 382 + }, + { + "epoch": 0.1530163803435877, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 383 + }, + { + "epoch": 0.1534159009188973, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 384 + }, + { + "epoch": 0.15381542149420696, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 385 + }, + { + "epoch": 0.15421494206951658, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 386 + }, + { + "epoch": 0.1546144626448262, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 387 + }, + { + "epoch": 0.15501398322013585, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 388 + }, + { + "epoch": 0.15541350379544547, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 389 + }, + { + "epoch": 0.1558130243707551, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 390 + }, + { + "epoch": 0.15621254494606474, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 391 + }, + { + "epoch": 0.15661206552137436, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 392 + }, + { + "epoch": 0.15701158609668397, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 393 + }, + { + "epoch": 0.1574111066719936, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3844, + "step": 394 + }, + { + "epoch": 0.15781062724730324, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 395 + }, + { + "epoch": 0.15821014782261286, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 396 + }, + { + "epoch": 0.15860966839792248, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 397 + }, + { + "epoch": 0.15900918897323213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 398 + }, + { + "epoch": 0.15940870954854175, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3927, + "step": 399 + }, + { + "epoch": 0.15980823012385137, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3929, + "step": 400 + }, + { + "epoch": 0.16020775069916102, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3835, + "step": 401 + }, + { + "epoch": 0.16060727127447064, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3941, + "step": 402 + }, + { + "epoch": 0.16100679184978026, + "grad_norm": 0.359375, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 403 + }, + { + "epoch": 0.1614063124250899, + "grad_norm": 1.5703125, + "learning_rate": 0.0002, + "loss": 1.4032, + "step": 404 + }, + { + "epoch": 0.16180583300039952, + "grad_norm": 3.984375, + "learning_rate": 0.0002, + "loss": 1.48, + "step": 405 + }, + { + "epoch": 0.16220535357570914, + "grad_norm": 2.234375, + "learning_rate": 0.0002, + "loss": 1.4185, + "step": 406 + }, + { + "epoch": 0.1626048741510188, + "grad_norm": 3.921875, + "learning_rate": 0.0002, + "loss": 1.5046, + "step": 407 + }, + { + "epoch": 0.1630043947263284, + "grad_norm": 3.578125, + "learning_rate": 0.0002, + "loss": 1.4966, + "step": 408 + }, + { + "epoch": 0.16340391530163803, + "grad_norm": 0.57421875, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 409 + }, + { + "epoch": 0.16380343587694765, + "grad_norm": 2.671875, + "learning_rate": 0.0002, + "loss": 1.4878, + "step": 410 + }, + { + "epoch": 0.1642029564522573, + "grad_norm": 2.34375, + "learning_rate": 0.0002, + "loss": 1.491, + "step": 411 + }, + { + "epoch": 0.16460247702756692, + "grad_norm": 0.69140625, + "learning_rate": 0.0002, + "loss": 1.4038, + "step": 412 + }, + { + "epoch": 0.16500199760287654, + "grad_norm": 1.1015625, + "learning_rate": 0.0002, + "loss": 1.4232, + "step": 413 + }, + { + "epoch": 0.16540151817818619, + "grad_norm": 0.83984375, + "learning_rate": 0.0002, + "loss": 1.4109, + "step": 414 + }, + { + "epoch": 0.1658010387534958, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.4187, + "step": 415 + }, + { + "epoch": 0.16620055932880543, + "grad_norm": 0.408203125, + "learning_rate": 0.0002, + "loss": 1.4005, + "step": 416 + }, + { + "epoch": 0.16660007990411507, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.4107, + "step": 417 + }, + { + "epoch": 0.1669996004794247, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.4082, + "step": 418 + }, + { + "epoch": 0.1673991210547343, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3982, + "step": 419 + }, + { + "epoch": 0.16779864163004396, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.4033, + "step": 420 + }, + { + "epoch": 0.16819816220535358, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3989, + "step": 421 + }, + { + "epoch": 0.1685976827806632, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.4054, + "step": 422 + }, + { + "epoch": 0.16899720335597282, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3933, + "step": 423 + }, + { + "epoch": 0.16939672393128247, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 424 + }, + { + "epoch": 0.1697962445065921, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3909, + "step": 425 + }, + { + "epoch": 0.1701957650819017, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3978, + "step": 426 + }, + { + "epoch": 0.17059528565721135, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 427 + }, + { + "epoch": 0.17099480623252097, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3955, + "step": 428 + }, + { + "epoch": 0.1713943268078306, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 429 + }, + { + "epoch": 0.17179384738314024, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3936, + "step": 430 + }, + { + "epoch": 0.17219336795844986, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3965, + "step": 431 + }, + { + "epoch": 0.17259288853375948, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3935, + "step": 432 + }, + { + "epoch": 0.17299240910906913, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3892, + "step": 433 + }, + { + "epoch": 0.17339192968437875, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3906, + "step": 434 + }, + { + "epoch": 0.17379145025968837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 435 + }, + { + "epoch": 0.17419097083499802, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.392, + "step": 436 + }, + { + "epoch": 0.17459049141030764, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3847, + "step": 437 + }, + { + "epoch": 0.17499001198561726, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3928, + "step": 438 + }, + { + "epoch": 0.17538953256092688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 439 + }, + { + "epoch": 0.17578905313623652, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 440 + }, + { + "epoch": 0.17618857371154614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3923, + "step": 441 + }, + { + "epoch": 0.17658809428685576, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3912, + "step": 442 + }, + { + "epoch": 0.1769876148621654, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 443 + }, + { + "epoch": 0.17738713543747503, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3956, + "step": 444 + }, + { + "epoch": 0.17778665601278465, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 445 + }, + { + "epoch": 0.1781861765880943, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 446 + }, + { + "epoch": 0.17858569716340392, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 447 + }, + { + "epoch": 0.17898521773871354, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3905, + "step": 448 + }, + { + "epoch": 0.17938473831402318, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 449 + }, + { + "epoch": 0.1797842588893328, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 450 + }, + { + "epoch": 0.18018377946464242, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 451 + }, + { + "epoch": 0.18058330003995204, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3881, + "step": 452 + }, + { + "epoch": 0.1809828206152617, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.4009, + "step": 453 + }, + { + "epoch": 0.1813823411905713, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 454 + }, + { + "epoch": 0.18178186176588093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.4001, + "step": 455 + }, + { + "epoch": 0.18218138234119058, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 456 + }, + { + "epoch": 0.1825809029165002, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 457 + }, + { + "epoch": 0.18298042349180982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 458 + }, + { + "epoch": 0.18337994406711947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 459 + }, + { + "epoch": 0.18377946464242909, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3897, + "step": 460 + }, + { + "epoch": 0.1841789852177387, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 461 + }, + { + "epoch": 0.18457850579304835, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 462 + }, + { + "epoch": 0.18497802636835797, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3921, + "step": 463 + }, + { + "epoch": 0.1853775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 464 + }, + { + "epoch": 0.18577706751897724, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 465 + }, + { + "epoch": 0.18617658809428686, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3796, + "step": 466 + }, + { + "epoch": 0.18657610866959648, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.4007, + "step": 467 + }, + { + "epoch": 0.1869756292449061, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 468 + }, + { + "epoch": 0.18737514982021575, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 469 + }, + { + "epoch": 0.18777467039552537, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 470 + }, + { + "epoch": 0.188174190970835, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3867, + "step": 471 + }, + { + "epoch": 0.18857371154614463, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3951, + "step": 472 + }, + { + "epoch": 0.18897323212145425, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 473 + }, + { + "epoch": 0.18937275269676387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3913, + "step": 474 + }, + { + "epoch": 0.18977227327207352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3863, + "step": 475 + }, + { + "epoch": 0.19017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 476 + }, + { + "epoch": 0.19057131442269276, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3891, + "step": 477 + }, + { + "epoch": 0.1909708349980024, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 478 + }, + { + "epoch": 0.19137035557331203, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 479 + }, + { + "epoch": 0.19176987614862165, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3807, + "step": 480 + }, + { + "epoch": 0.1921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 481 + }, + { + "epoch": 0.19256891729924092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.4011, + "step": 482 + }, + { + "epoch": 0.19296843787455054, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3843, + "step": 483 + }, + { + "epoch": 0.19336795844986016, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 484 + }, + { + "epoch": 0.1937674790251698, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 485 + }, + { + "epoch": 0.19416699960047942, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3934, + "step": 486 + }, + { + "epoch": 0.19456652017578904, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 487 + }, + { + "epoch": 0.1949660407510987, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 488 + }, + { + "epoch": 0.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 489 + }, + { + "epoch": 0.19576508190171793, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 490 + }, + { + "epoch": 0.19616460247702758, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3816, + "step": 491 + }, + { + "epoch": 0.1965641230523372, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 492 + }, + { + "epoch": 0.19696364362764682, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3824, + "step": 493 + }, + { + "epoch": 0.19736316420295646, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.387, + "step": 494 + }, + { + "epoch": 0.19776268477826608, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3943, + "step": 495 + }, + { + "epoch": 0.1981622053535757, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 496 + }, + { + "epoch": 0.19856172592888532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 497 + }, + { + "epoch": 0.19896124650419497, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 498 + }, + { + "epoch": 0.1993607670795046, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 499 + }, + { + "epoch": 0.1997602876548142, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3874, + "step": 500 + }, + { + "epoch": 0.20015980823012386, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3896, + "step": 501 + }, + { + "epoch": 0.20055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 502 + }, + { + "epoch": 0.2009588493807431, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 503 + }, + { + "epoch": 0.20135836995605275, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 504 + }, + { + "epoch": 0.20175789053136237, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 505 + }, + { + "epoch": 0.20215741110667199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 506 + }, + { + "epoch": 0.20255693168198163, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 507 + }, + { + "epoch": 0.20295645225729125, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 508 + }, + { + "epoch": 0.20335597283260087, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 509 + }, + { + "epoch": 0.20375549340791052, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 510 + }, + { + "epoch": 0.20415501398322014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.384, + "step": 511 + }, + { + "epoch": 0.20455453455852976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 512 + }, + { + "epoch": 0.20495405513383938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 513 + }, + { + "epoch": 0.20535357570914903, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 514 + }, + { + "epoch": 0.20575309628445865, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 515 + }, + { + "epoch": 0.20615261685976827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 516 + }, + { + "epoch": 0.20655213743507791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.38, + "step": 517 + }, + { + "epoch": 0.20695165801038753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 518 + }, + { + "epoch": 0.20735117858569715, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 519 + }, + { + "epoch": 0.2077506991610068, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3857, + "step": 520 + }, + { + "epoch": 0.20815021973631642, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 521 + }, + { + "epoch": 0.20854974031162604, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 522 + }, + { + "epoch": 0.2089492608869357, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 523 + }, + { + "epoch": 0.2093487814622453, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 524 + }, + { + "epoch": 0.20974830203755493, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3845, + "step": 525 + }, + { + "epoch": 0.21014782261286455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 526 + }, + { + "epoch": 0.2105473431881742, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 527 + }, + { + "epoch": 0.21094686376348382, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3958, + "step": 528 + }, + { + "epoch": 0.21134638433879344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3852, + "step": 529 + }, + { + "epoch": 0.21174590491410308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3864, + "step": 530 + }, + { + "epoch": 0.2121454254894127, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 531 + }, + { + "epoch": 0.21254494606472232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 532 + }, + { + "epoch": 0.21294446664003197, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3837, + "step": 533 + }, + { + "epoch": 0.2133439872153416, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 534 + }, + { + "epoch": 0.2137435077906512, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.389, + "step": 535 + }, + { + "epoch": 0.21414302836596086, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3884, + "step": 536 + }, + { + "epoch": 0.21454254894127048, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 537 + }, + { + "epoch": 0.2149420695165801, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 538 + }, + { + "epoch": 0.21534159009188975, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 539 + }, + { + "epoch": 0.21574111066719937, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 540 + }, + { + "epoch": 0.21614063124250898, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 541 + }, + { + "epoch": 0.2165401518178186, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 542 + }, + { + "epoch": 0.21693967239312825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3754, + "step": 543 + }, + { + "epoch": 0.21733919296843787, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 544 + }, + { + "epoch": 0.2177387135437475, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 545 + }, + { + "epoch": 0.21813823411905714, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3831, + "step": 546 + }, + { + "epoch": 0.21853775469436676, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3887, + "step": 547 + }, + { + "epoch": 0.21893727526967638, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 548 + }, + { + "epoch": 0.21933679584498603, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 549 + }, + { + "epoch": 0.21973631642029565, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 550 + }, + { + "epoch": 0.22013583699560527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 551 + }, + { + "epoch": 0.2205353575709149, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 552 + }, + { + "epoch": 0.22093487814622453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 553 + }, + { + "epoch": 0.22133439872153415, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 554 + }, + { + "epoch": 0.2217339192968438, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3992, + "step": 555 + }, + { + "epoch": 0.22213343987215342, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 556 + }, + { + "epoch": 0.22253296044746304, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 557 + }, + { + "epoch": 0.22293248102277266, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 558 + }, + { + "epoch": 0.2233320015980823, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 559 + }, + { + "epoch": 0.22373152217339193, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3812, + "step": 560 + }, + { + "epoch": 0.22413104274870155, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3856, + "step": 561 + }, + { + "epoch": 0.2245305633240112, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 562 + }, + { + "epoch": 0.22493008389932082, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 563 + }, + { + "epoch": 0.22532960447463043, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 564 + }, + { + "epoch": 0.22572912504994008, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3815, + "step": 565 + }, + { + "epoch": 0.2261286456252497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3808, + "step": 566 + }, + { + "epoch": 0.22652816620055932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 567 + }, + { + "epoch": 0.22692768677586897, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 568 + }, + { + "epoch": 0.2273272073511786, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 569 + }, + { + "epoch": 0.2277267279264882, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3871, + "step": 570 + }, + { + "epoch": 0.22812624850179783, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3851, + "step": 571 + }, + { + "epoch": 0.22852576907710748, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 572 + }, + { + "epoch": 0.2289252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3841, + "step": 573 + }, + { + "epoch": 0.22932481022772672, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 574 + }, + { + "epoch": 0.22972433080303636, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3876, + "step": 575 + }, + { + "epoch": 0.23012385137834598, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 576 + }, + { + "epoch": 0.2305233719536556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3786, + "step": 577 + }, + { + "epoch": 0.23092289252896525, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3849, + "step": 578 + }, + { + "epoch": 0.23132241310427487, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 579 + }, + { + "epoch": 0.2317219336795845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3878, + "step": 580 + }, + { + "epoch": 0.23212145425489414, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 581 + }, + { + "epoch": 0.23252097483020376, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3781, + "step": 582 + }, + { + "epoch": 0.23292049540551338, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 583 + }, + { + "epoch": 0.23332001598082303, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3853, + "step": 584 + }, + { + "epoch": 0.23371953655613265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.388, + "step": 585 + }, + { + "epoch": 0.23411905713144227, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 586 + }, + { + "epoch": 0.23451857770675189, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 587 + }, + { + "epoch": 0.23491809828206153, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 588 + }, + { + "epoch": 0.23531761885737115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3794, + "step": 589 + }, + { + "epoch": 0.23571713943268077, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 590 + }, + { + "epoch": 0.23611666000799042, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 591 + }, + { + "epoch": 0.23651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 592 + }, + { + "epoch": 0.23691570115860966, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 593 + }, + { + "epoch": 0.2373152217339193, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 594 + }, + { + "epoch": 0.23771474230922893, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 595 + }, + { + "epoch": 0.23811426288453855, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3846, + "step": 596 + }, + { + "epoch": 0.2385137834598482, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3914, + "step": 597 + }, + { + "epoch": 0.23891330403515781, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 598 + }, + { + "epoch": 0.23931282461046743, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 599 + }, + { + "epoch": 0.23971234518577705, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 600 + }, + { + "epoch": 0.2401118657610867, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 601 + }, + { + "epoch": 0.24051138633639632, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 602 + }, + { + "epoch": 0.24091090691170594, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3907, + "step": 603 + }, + { + "epoch": 0.2413104274870156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 604 + }, + { + "epoch": 0.2417099480623252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 605 + }, + { + "epoch": 0.24210946863763483, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3879, + "step": 606 + }, + { + "epoch": 0.24250898921294448, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 607 + }, + { + "epoch": 0.2429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 608 + }, + { + "epoch": 0.24330803036356372, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 609 + }, + { + "epoch": 0.24370755093887336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 610 + }, + { + "epoch": 0.24410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 611 + }, + { + "epoch": 0.2445065920894926, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 612 + }, + { + "epoch": 0.24490611266480225, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 613 + }, + { + "epoch": 0.24530563324011187, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3885, + "step": 614 + }, + { + "epoch": 0.2457051538154215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 615 + }, + { + "epoch": 0.2461046743907311, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3783, + "step": 616 + }, + { + "epoch": 0.24650419496604076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3918, + "step": 617 + }, + { + "epoch": 0.24690371554135038, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 618 + }, + { + "epoch": 0.24730323611666, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.383, + "step": 619 + }, + { + "epoch": 0.24770275669196964, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3883, + "step": 620 + }, + { + "epoch": 0.24810227726727926, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 621 + }, + { + "epoch": 0.24850179784258888, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 622 + }, + { + "epoch": 0.24890131841789853, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3811, + "step": 623 + }, + { + "epoch": 0.24930083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 624 + }, + { + "epoch": 0.24970035956851777, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 625 + }, + { + "epoch": 0.2500998801438274, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3777, + "step": 626 + }, + { + "epoch": 0.250499400719137, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3748, + "step": 627 + }, + { + "epoch": 0.2508989212944467, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 628 + }, + { + "epoch": 0.2512984418697563, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.377, + "step": 629 + }, + { + "epoch": 0.2516979624450659, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 630 + }, + { + "epoch": 0.25209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3859, + "step": 631 + }, + { + "epoch": 0.25249700359568517, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 632 + }, + { + "epoch": 0.2528965241709948, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 633 + }, + { + "epoch": 0.25329604474630446, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 634 + }, + { + "epoch": 0.2536955653216141, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 635 + }, + { + "epoch": 0.2540950858969237, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 636 + }, + { + "epoch": 0.2544946064722333, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 637 + }, + { + "epoch": 0.25489412704754294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 638 + }, + { + "epoch": 0.25529364762285256, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3833, + "step": 639 + }, + { + "epoch": 0.2556931681981622, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3972, + "step": 640 + }, + { + "epoch": 0.25609268877347185, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 641 + }, + { + "epoch": 0.2564922093487815, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 642 + }, + { + "epoch": 0.2568917299240911, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3825, + "step": 643 + }, + { + "epoch": 0.2572912504994007, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 644 + }, + { + "epoch": 0.25769077107471033, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3862, + "step": 645 + }, + { + "epoch": 0.25809029165001995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 646 + }, + { + "epoch": 0.25848981222532963, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 647 + }, + { + "epoch": 0.25888933280063925, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 648 + }, + { + "epoch": 0.25928885337594887, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3704, + "step": 649 + }, + { + "epoch": 0.2596883739512585, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 650 + }, + { + "epoch": 0.2600878945265681, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.381, + "step": 651 + }, + { + "epoch": 0.26048741510187773, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 652 + }, + { + "epoch": 0.26088693567718735, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3866, + "step": 653 + }, + { + "epoch": 0.261286456252497, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 654 + }, + { + "epoch": 0.26168597682780664, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 655 + }, + { + "epoch": 0.26208549740311626, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3838, + "step": 656 + }, + { + "epoch": 0.2624850179784259, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 657 + }, + { + "epoch": 0.2628845385537355, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3802, + "step": 658 + }, + { + "epoch": 0.2632840591290451, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 659 + }, + { + "epoch": 0.2636835797043548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3865, + "step": 660 + }, + { + "epoch": 0.2640831002796644, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 661 + }, + { + "epoch": 0.26448262085497404, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3877, + "step": 662 + }, + { + "epoch": 0.26488214143028366, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 663 + }, + { + "epoch": 0.2652816620055933, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 664 + }, + { + "epoch": 0.2656811825809029, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3795, + "step": 665 + }, + { + "epoch": 0.2660807031562126, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 666 + }, + { + "epoch": 0.2664802237315222, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 667 + }, + { + "epoch": 0.2668797443068318, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 668 + }, + { + "epoch": 0.26727926488214143, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 669 + }, + { + "epoch": 0.26767878545745105, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 670 + }, + { + "epoch": 0.26807830603276067, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3916, + "step": 671 + }, + { + "epoch": 0.2684778266080703, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 672 + }, + { + "epoch": 0.26887734718337997, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.386, + "step": 673 + }, + { + "epoch": 0.2692768677586896, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 674 + }, + { + "epoch": 0.2696763883339992, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 675 + }, + { + "epoch": 0.2700759089093088, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 676 + }, + { + "epoch": 0.27047542948461845, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 677 + }, + { + "epoch": 0.27087495005992807, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 678 + }, + { + "epoch": 0.27127447063523774, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 679 + }, + { + "epoch": 0.27167399121054736, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 680 + }, + { + "epoch": 0.272073511785857, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.378, + "step": 681 + }, + { + "epoch": 0.2724730323611666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3888, + "step": 682 + }, + { + "epoch": 0.2728725529364762, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 683 + }, + { + "epoch": 0.27327207351178584, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 684 + }, + { + "epoch": 0.27367159408709546, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 685 + }, + { + "epoch": 0.27407111466240514, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3747, + "step": 686 + }, + { + "epoch": 0.27447063523771476, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 687 + }, + { + "epoch": 0.2748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3917, + "step": 688 + }, + { + "epoch": 0.275269676388334, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3809, + "step": 689 + }, + { + "epoch": 0.2756691969636436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 690 + }, + { + "epoch": 0.27606871753895323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 691 + }, + { + "epoch": 0.2764682381142629, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 692 + }, + { + "epoch": 0.27686775868957253, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 693 + }, + { + "epoch": 0.27726727926488215, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 694 + }, + { + "epoch": 0.27766679984019177, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3842, + "step": 695 + }, + { + "epoch": 0.2780663204155014, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 696 + }, + { + "epoch": 0.278465840990811, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 697 + }, + { + "epoch": 0.27886536156612063, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 698 + }, + { + "epoch": 0.2792648821414303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 699 + }, + { + "epoch": 0.2796644027167399, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 700 + }, + { + "epoch": 0.28006392329204954, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3742, + "step": 701 + }, + { + "epoch": 0.28046344386735916, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3755, + "step": 702 + }, + { + "epoch": 0.2808629644426688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 703 + }, + { + "epoch": 0.2812624850179784, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3805, + "step": 704 + }, + { + "epoch": 0.2816620055932881, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 705 + }, + { + "epoch": 0.2820615261685977, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 706 + }, + { + "epoch": 0.2824610467439073, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 707 + }, + { + "epoch": 0.28286056731921694, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 708 + }, + { + "epoch": 0.28326008789452656, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 709 + }, + { + "epoch": 0.2836596084698362, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3767, + "step": 710 + }, + { + "epoch": 0.2840591290451458, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 711 + }, + { + "epoch": 0.2844586496204555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 712 + }, + { + "epoch": 0.2848581701957651, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 713 + }, + { + "epoch": 0.2852576907710747, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 714 + }, + { + "epoch": 0.28565721134638433, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 715 + }, + { + "epoch": 0.28605673192169395, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 716 + }, + { + "epoch": 0.28645625249700357, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3788, + "step": 717 + }, + { + "epoch": 0.28685577307231325, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 718 + }, + { + "epoch": 0.28725529364762287, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 719 + }, + { + "epoch": 0.2876548142229325, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 720 + }, + { + "epoch": 0.2880543347982421, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3826, + "step": 721 + }, + { + "epoch": 0.2884538553735517, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 722 + }, + { + "epoch": 0.28885337594886135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 723 + }, + { + "epoch": 0.289252896524171, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 724 + }, + { + "epoch": 0.28965241709948064, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 725 + }, + { + "epoch": 0.29005193767479026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 726 + }, + { + "epoch": 0.2904514582500999, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 727 + }, + { + "epoch": 0.2908509788254095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3791, + "step": 728 + }, + { + "epoch": 0.2912504994007191, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3854, + "step": 729 + }, + { + "epoch": 0.29165001997602874, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 730 + }, + { + "epoch": 0.2920495405513384, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3827, + "step": 731 + }, + { + "epoch": 0.29244906112664804, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.385, + "step": 732 + }, + { + "epoch": 0.29284858170195766, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 733 + }, + { + "epoch": 0.2932481022772673, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3723, + "step": 734 + }, + { + "epoch": 0.2936476228525769, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3823, + "step": 735 + }, + { + "epoch": 0.2940471434278865, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 736 + }, + { + "epoch": 0.2944466640031962, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 737 + }, + { + "epoch": 0.2948461845785058, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 738 + }, + { + "epoch": 0.29524570515381543, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 739 + }, + { + "epoch": 0.29564522572912505, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.376, + "step": 740 + }, + { + "epoch": 0.29604474630443467, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3822, + "step": 741 + }, + { + "epoch": 0.2964442668797443, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 742 + }, + { + "epoch": 0.2968437874550539, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 743 + }, + { + "epoch": 0.2972433080303636, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3804, + "step": 744 + }, + { + "epoch": 0.2976428286056732, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 745 + }, + { + "epoch": 0.2980423491809828, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 746 + }, + { + "epoch": 0.29844186975629244, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 747 + }, + { + "epoch": 0.29884139033160206, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 748 + }, + { + "epoch": 0.2992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 749 + }, + { + "epoch": 0.29964043148222136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.382, + "step": 750 + }, + { + "epoch": 0.300039952057531, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 751 + }, + { + "epoch": 0.3004394726328406, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3745, + "step": 752 + }, + { + "epoch": 0.3008389932081502, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3875, + "step": 753 + }, + { + "epoch": 0.30123851378345984, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 754 + }, + { + "epoch": 0.30163803435876946, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 755 + }, + { + "epoch": 0.3020375549340791, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3732, + "step": 756 + }, + { + "epoch": 0.30243707550938875, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 757 + }, + { + "epoch": 0.3028365960846984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3765, + "step": 758 + }, + { + "epoch": 0.303236116660008, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 759 + }, + { + "epoch": 0.3036356372353176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3698, + "step": 760 + }, + { + "epoch": 0.30403515781062723, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 761 + }, + { + "epoch": 0.30443467838593685, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3761, + "step": 762 + }, + { + "epoch": 0.3048341989612465, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 763 + }, + { + "epoch": 0.30523371953655615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3718, + "step": 764 + }, + { + "epoch": 0.30563324011186577, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 765 + }, + { + "epoch": 0.3060327606871754, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 766 + }, + { + "epoch": 0.306432281262485, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 767 + }, + { + "epoch": 0.3068318018377946, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 768 + }, + { + "epoch": 0.3072313224131043, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3694, + "step": 769 + }, + { + "epoch": 0.3076308429884139, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3731, + "step": 770 + }, + { + "epoch": 0.30803036356372354, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 771 + }, + { + "epoch": 0.30842988413903316, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 772 + }, + { + "epoch": 0.3088294047143428, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 773 + }, + { + "epoch": 0.3092289252896524, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.379, + "step": 774 + }, + { + "epoch": 0.309628445864962, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3775, + "step": 775 + }, + { + "epoch": 0.3100279664402717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 776 + }, + { + "epoch": 0.3104274870155813, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 777 + }, + { + "epoch": 0.31082700759089094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 778 + }, + { + "epoch": 0.31122652816620056, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 779 + }, + { + "epoch": 0.3116260487415102, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 780 + }, + { + "epoch": 0.3120255693168198, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3818, + "step": 781 + }, + { + "epoch": 0.31242508989212947, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 782 + }, + { + "epoch": 0.3128246104674391, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 783 + }, + { + "epoch": 0.3132241310427487, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 784 + }, + { + "epoch": 0.31362365161805833, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 785 + }, + { + "epoch": 0.31402317219336795, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 786 + }, + { + "epoch": 0.31442269276867757, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3789, + "step": 787 + }, + { + "epoch": 0.3148222133439872, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 788 + }, + { + "epoch": 0.31522173391929686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 789 + }, + { + "epoch": 0.3156212544946065, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3768, + "step": 790 + }, + { + "epoch": 0.3160207750699161, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 791 + }, + { + "epoch": 0.3164202956452257, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3728, + "step": 792 + }, + { + "epoch": 0.31681981622053534, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 793 + }, + { + "epoch": 0.31721933679584496, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3798, + "step": 794 + }, + { + "epoch": 0.31761885737115464, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 795 + }, + { + "epoch": 0.31801837794646426, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 796 + }, + { + "epoch": 0.3184178985217739, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 797 + }, + { + "epoch": 0.3188174190970835, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 798 + }, + { + "epoch": 0.3192169396723931, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 799 + }, + { + "epoch": 0.31961646024770274, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3763, + "step": 800 + }, + { + "epoch": 0.32001598082301236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 801 + }, + { + "epoch": 0.32041550139832203, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 802 + }, + { + "epoch": 0.32081502197363165, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 803 + }, + { + "epoch": 0.3212145425489413, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 804 + }, + { + "epoch": 0.3216140631242509, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 805 + }, + { + "epoch": 0.3220135836995605, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 806 + }, + { + "epoch": 0.32241310427487013, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 807 + }, + { + "epoch": 0.3228126248501798, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 808 + }, + { + "epoch": 0.3232121454254894, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 809 + }, + { + "epoch": 0.32361166600079905, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 810 + }, + { + "epoch": 0.32401118657610867, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 811 + }, + { + "epoch": 0.3244107071514183, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 812 + }, + { + "epoch": 0.3248102277267279, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3743, + "step": 813 + }, + { + "epoch": 0.3252097483020376, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.373, + "step": 814 + }, + { + "epoch": 0.3256092688773472, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 815 + }, + { + "epoch": 0.3260087894526568, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 816 + }, + { + "epoch": 0.32640831002796644, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 817 + }, + { + "epoch": 0.32680783060327606, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 818 + }, + { + "epoch": 0.3272073511785857, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 819 + }, + { + "epoch": 0.3276068717538953, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 820 + }, + { + "epoch": 0.328006392329205, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 821 + }, + { + "epoch": 0.3284059129045146, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 822 + }, + { + "epoch": 0.3288054334798242, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 823 + }, + { + "epoch": 0.32920495405513384, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 824 + }, + { + "epoch": 0.32960447463044346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 825 + }, + { + "epoch": 0.3300039952057531, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 826 + }, + { + "epoch": 0.33040351578106275, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 827 + }, + { + "epoch": 0.33080303635637237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 828 + }, + { + "epoch": 0.331202556931682, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 829 + }, + { + "epoch": 0.3316020775069916, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 830 + }, + { + "epoch": 0.33200159808230123, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 831 + }, + { + "epoch": 0.33240111865761085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 832 + }, + { + "epoch": 0.33280063923292047, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 833 + }, + { + "epoch": 0.33320015980823015, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 834 + }, + { + "epoch": 0.33359968038353977, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 835 + }, + { + "epoch": 0.3339992009588494, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3762, + "step": 836 + }, + { + "epoch": 0.334398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 837 + }, + { + "epoch": 0.3347982421094686, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 838 + }, + { + "epoch": 0.33519776268477824, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 839 + }, + { + "epoch": 0.3355972832600879, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 840 + }, + { + "epoch": 0.33599680383539754, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 841 + }, + { + "epoch": 0.33639632441070716, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 842 + }, + { + "epoch": 0.3367958449860168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3735, + "step": 843 + }, + { + "epoch": 0.3371953655613264, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 844 + }, + { + "epoch": 0.337594886136636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 845 + }, + { + "epoch": 0.33799440671194564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3793, + "step": 846 + }, + { + "epoch": 0.3383939272872553, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 847 + }, + { + "epoch": 0.33879344786256493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 848 + }, + { + "epoch": 0.33919296843787455, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 849 + }, + { + "epoch": 0.3395924890131842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 850 + }, + { + "epoch": 0.3399920095884938, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 851 + }, + { + "epoch": 0.3403915301638034, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 852 + }, + { + "epoch": 0.3407910507391131, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 853 + }, + { + "epoch": 0.3411905713144227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3733, + "step": 854 + }, + { + "epoch": 0.34159009188973233, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 855 + }, + { + "epoch": 0.34198961246504195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 856 + }, + { + "epoch": 0.34238913304035157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 857 + }, + { + "epoch": 0.3427886536156612, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 858 + }, + { + "epoch": 0.34318817419097086, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 859 + }, + { + "epoch": 0.3435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 860 + }, + { + "epoch": 0.3439872153415901, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 861 + }, + { + "epoch": 0.3443867359168997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 862 + }, + { + "epoch": 0.34478625649220934, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 863 + }, + { + "epoch": 0.34518577706751896, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 864 + }, + { + "epoch": 0.3455852976428286, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 865 + }, + { + "epoch": 0.34598481821813826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 866 + }, + { + "epoch": 0.3463843387934479, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3776, + "step": 867 + }, + { + "epoch": 0.3467838593687575, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 868 + }, + { + "epoch": 0.3471833799440671, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3784, + "step": 869 + }, + { + "epoch": 0.34758290051937674, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 870 + }, + { + "epoch": 0.34798242109468636, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 871 + }, + { + "epoch": 0.34838194166999603, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 872 + }, + { + "epoch": 0.34878146224530565, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 873 + }, + { + "epoch": 0.34918098282061527, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 874 + }, + { + "epoch": 0.3495805033959249, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3814, + "step": 875 + }, + { + "epoch": 0.3499800239712345, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 876 + }, + { + "epoch": 0.35037954454654413, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 877 + }, + { + "epoch": 0.35077906512185375, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3757, + "step": 878 + }, + { + "epoch": 0.3511785856971634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 879 + }, + { + "epoch": 0.35157810627247305, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 880 + }, + { + "epoch": 0.35197762684778267, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 881 + }, + { + "epoch": 0.3523771474230923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3697, + "step": 882 + }, + { + "epoch": 0.3527766679984019, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 883 + }, + { + "epoch": 0.3531761885737115, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 884 + }, + { + "epoch": 0.3535757091490212, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3778, + "step": 885 + }, + { + "epoch": 0.3539752297243308, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 886 + }, + { + "epoch": 0.35437475029964044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 887 + }, + { + "epoch": 0.35477427087495006, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3751, + "step": 888 + }, + { + "epoch": 0.3551737914502597, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 889 + }, + { + "epoch": 0.3555733120255693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 890 + }, + { + "epoch": 0.3559728326008789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 891 + }, + { + "epoch": 0.3563723531761886, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 892 + }, + { + "epoch": 0.3567718737514982, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 893 + }, + { + "epoch": 0.35717139432680783, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 894 + }, + { + "epoch": 0.35757091490211745, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 895 + }, + { + "epoch": 0.3579704354774271, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3758, + "step": 896 + }, + { + "epoch": 0.3583699560527367, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3799, + "step": 897 + }, + { + "epoch": 0.35876947662804637, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 898 + }, + { + "epoch": 0.359168997203356, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 899 + }, + { + "epoch": 0.3595685177786656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 900 + }, + { + "epoch": 0.35996803835397523, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.372, + "step": 901 + }, + { + "epoch": 0.36036755892928485, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3774, + "step": 902 + }, + { + "epoch": 0.36076707950459447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 903 + }, + { + "epoch": 0.3611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 904 + }, + { + "epoch": 0.36156612065521376, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 905 + }, + { + "epoch": 0.3619656412305234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3736, + "step": 906 + }, + { + "epoch": 0.362365161805833, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 907 + }, + { + "epoch": 0.3627646823811426, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 908 + }, + { + "epoch": 0.36316420295645224, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 909 + }, + { + "epoch": 0.36356372353176186, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3721, + "step": 910 + }, + { + "epoch": 0.36396324410707154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3719, + "step": 911 + }, + { + "epoch": 0.36436276468238116, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 912 + }, + { + "epoch": 0.3647622852576908, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 913 + }, + { + "epoch": 0.3651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3756, + "step": 914 + }, + { + "epoch": 0.36556132640831, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 915 + }, + { + "epoch": 0.36596084698361964, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 916 + }, + { + "epoch": 0.3663603675589293, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 917 + }, + { + "epoch": 0.36675988813423893, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 918 + }, + { + "epoch": 0.36715940870954855, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 919 + }, + { + "epoch": 0.36755892928485817, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 920 + }, + { + "epoch": 0.3679584498601678, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 921 + }, + { + "epoch": 0.3683579704354774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 922 + }, + { + "epoch": 0.36875749101078703, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 923 + }, + { + "epoch": 0.3691570115860967, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 924 + }, + { + "epoch": 0.3695565321614063, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 925 + }, + { + "epoch": 0.36995605273671595, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 926 + }, + { + "epoch": 0.37035557331202557, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 927 + }, + { + "epoch": 0.3707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 928 + }, + { + "epoch": 0.3711546144626448, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3801, + "step": 929 + }, + { + "epoch": 0.3715541350379545, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 930 + }, + { + "epoch": 0.3719536556132641, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 931 + }, + { + "epoch": 0.3723531761885737, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 932 + }, + { + "epoch": 0.37275269676388334, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 933 + }, + { + "epoch": 0.37315221733919296, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 934 + }, + { + "epoch": 0.3735517379145026, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.371, + "step": 935 + }, + { + "epoch": 0.3739512584898122, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 936 + }, + { + "epoch": 0.3743507790651219, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3752, + "step": 937 + }, + { + "epoch": 0.3747502996404315, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 938 + }, + { + "epoch": 0.3751498202157411, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 939 + }, + { + "epoch": 0.37554934079105073, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 940 + }, + { + "epoch": 0.37594886136636035, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 941 + }, + { + "epoch": 0.37634838194167, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3819, + "step": 942 + }, + { + "epoch": 0.37674790251697965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 943 + }, + { + "epoch": 0.37714742309228927, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 944 + }, + { + "epoch": 0.3775469436675989, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3708, + "step": 945 + }, + { + "epoch": 0.3779464642429085, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 946 + }, + { + "epoch": 0.37834598481821813, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 947 + }, + { + "epoch": 0.37874550539352775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 948 + }, + { + "epoch": 0.37914502596883737, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 949 + }, + { + "epoch": 0.37954454654414704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 950 + }, + { + "epoch": 0.37994406711945666, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 951 + }, + { + "epoch": 0.3803435876947663, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 952 + }, + { + "epoch": 0.3807431082700759, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 953 + }, + { + "epoch": 0.3811426288453855, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 954 + }, + { + "epoch": 0.38154214942069514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3829, + "step": 955 + }, + { + "epoch": 0.3819416699960048, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 956 + }, + { + "epoch": 0.38234119057131444, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 957 + }, + { + "epoch": 0.38274071114662406, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3734, + "step": 958 + }, + { + "epoch": 0.3831402317219337, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 959 + }, + { + "epoch": 0.3835397522972433, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 960 + }, + { + "epoch": 0.3839392728725529, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 961 + }, + { + "epoch": 0.3843387934478626, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3661, + "step": 962 + }, + { + "epoch": 0.3847383140231722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 963 + }, + { + "epoch": 0.38513783459848183, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 964 + }, + { + "epoch": 0.38553735517379145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 965 + }, + { + "epoch": 0.38593687574910107, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 966 + }, + { + "epoch": 0.3863363963244107, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 967 + }, + { + "epoch": 0.3867359168997203, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 968 + }, + { + "epoch": 0.38713543747503, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 969 + }, + { + "epoch": 0.3875349580503396, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 970 + }, + { + "epoch": 0.3879344786256492, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 971 + }, + { + "epoch": 0.38833399920095885, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 972 + }, + { + "epoch": 0.38873351977626847, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 973 + }, + { + "epoch": 0.3891330403515781, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 974 + }, + { + "epoch": 0.38953256092688776, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 975 + }, + { + "epoch": 0.3899320815021974, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 976 + }, + { + "epoch": 0.390331602077507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 977 + }, + { + "epoch": 0.3907311226528166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3727, + "step": 978 + }, + { + "epoch": 0.39113064322812624, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 979 + }, + { + "epoch": 0.39153016380343586, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 980 + }, + { + "epoch": 0.3919296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3773, + "step": 981 + }, + { + "epoch": 0.39232920495405516, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3806, + "step": 982 + }, + { + "epoch": 0.3927287255293648, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 983 + }, + { + "epoch": 0.3931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3738, + "step": 984 + }, + { + "epoch": 0.393527766679984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 985 + }, + { + "epoch": 0.39392728725529363, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3746, + "step": 986 + }, + { + "epoch": 0.39432680783060325, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 987 + }, + { + "epoch": 0.39472632840591293, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3687, + "step": 988 + }, + { + "epoch": 0.39512584898122255, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3726, + "step": 989 + }, + { + "epoch": 0.39552536955653217, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 990 + }, + { + "epoch": 0.3959248901318418, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 991 + }, + { + "epoch": 0.3963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 992 + }, + { + "epoch": 0.39672393128246103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 993 + }, + { + "epoch": 0.39712345185777065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 994 + }, + { + "epoch": 0.3975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 995 + }, + { + "epoch": 0.39792249300838994, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 996 + }, + { + "epoch": 0.39832201358369956, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 997 + }, + { + "epoch": 0.3987215341590092, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 998 + }, + { + "epoch": 0.3991210547343188, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3749, + "step": 999 + }, + { + "epoch": 0.3995205753096284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1000 + }, + { + "epoch": 0.3999200958849381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1001 + }, + { + "epoch": 0.4003196164602477, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1002 + }, + { + "epoch": 0.40071913703555734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1003 + }, + { + "epoch": 0.40111865761086696, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1004 + }, + { + "epoch": 0.4015181781861766, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3764, + "step": 1005 + }, + { + "epoch": 0.4019176987614862, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1006 + }, + { + "epoch": 0.4023172193367959, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1007 + }, + { + "epoch": 0.4027167399121055, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1008 + }, + { + "epoch": 0.4031162604874151, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1009 + }, + { + "epoch": 0.40351578106272473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1010 + }, + { + "epoch": 0.40391530163803435, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1011 + }, + { + "epoch": 0.40431482221334397, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1012 + }, + { + "epoch": 0.4047143427886536, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3663, + "step": 1013 + }, + { + "epoch": 0.40511386336396327, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1014 + }, + { + "epoch": 0.4055133839392729, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1015 + }, + { + "epoch": 0.4059129045145825, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1016 + }, + { + "epoch": 0.4063124250898921, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1017 + }, + { + "epoch": 0.40671194566520175, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3739, + "step": 1018 + }, + { + "epoch": 0.40711146624051137, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1019 + }, + { + "epoch": 0.40751098681582104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1020 + }, + { + "epoch": 0.40791050739113066, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1021 + }, + { + "epoch": 0.4083100279664403, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1022 + }, + { + "epoch": 0.4087095485417499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1023 + }, + { + "epoch": 0.4091090691170595, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3792, + "step": 1024 + }, + { + "epoch": 0.40950858969236914, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1025 + }, + { + "epoch": 0.40990811026767876, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1026 + }, + { + "epoch": 0.41030763084298844, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1027 + }, + { + "epoch": 0.41070715141829806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1028 + }, + { + "epoch": 0.4111066719936077, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1029 + }, + { + "epoch": 0.4115061925689173, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1030 + }, + { + "epoch": 0.4119057131442269, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1031 + }, + { + "epoch": 0.41230523371953653, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1032 + }, + { + "epoch": 0.4127047542948462, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1033 + }, + { + "epoch": 0.41310427487015583, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1034 + }, + { + "epoch": 0.41350379544546545, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1035 + }, + { + "epoch": 0.41390331602077507, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1036 + }, + { + "epoch": 0.4143028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1037 + }, + { + "epoch": 0.4147023571713943, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1038 + }, + { + "epoch": 0.41510187774670393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1039 + }, + { + "epoch": 0.4155013983220136, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1040 + }, + { + "epoch": 0.4159009188973232, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1041 + }, + { + "epoch": 0.41630043947263284, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1042 + }, + { + "epoch": 0.41669996004794246, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1043 + }, + { + "epoch": 0.4170994806232521, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3722, + "step": 1044 + }, + { + "epoch": 0.4174990011985617, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1045 + }, + { + "epoch": 0.4178985217738714, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3668, + "step": 1046 + }, + { + "epoch": 0.418298042349181, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1047 + }, + { + "epoch": 0.4186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1048 + }, + { + "epoch": 0.41909708349980024, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3711, + "step": 1049 + }, + { + "epoch": 0.41949660407510986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1050 + }, + { + "epoch": 0.4198961246504195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1051 + }, + { + "epoch": 0.4202956452257291, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1052 + }, + { + "epoch": 0.4206951658010388, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1053 + }, + { + "epoch": 0.4210946863763484, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1054 + }, + { + "epoch": 0.421494206951658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1055 + }, + { + "epoch": 0.42189372752696763, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1056 + }, + { + "epoch": 0.42229324810227725, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3662, + "step": 1057 + }, + { + "epoch": 0.42269276867758687, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1058 + }, + { + "epoch": 0.42309228925289655, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3679, + "step": 1059 + }, + { + "epoch": 0.42349180982820617, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1060 + }, + { + "epoch": 0.4238913304035158, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1061 + }, + { + "epoch": 0.4242908509788254, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1062 + }, + { + "epoch": 0.424690371554135, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1063 + }, + { + "epoch": 0.42508989212944465, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1064 + }, + { + "epoch": 0.4254894127047543, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3725, + "step": 1065 + }, + { + "epoch": 0.42588893328006394, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1066 + }, + { + "epoch": 0.42628845385537356, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1067 + }, + { + "epoch": 0.4266879744306832, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3787, + "step": 1068 + }, + { + "epoch": 0.4270874950059928, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1069 + }, + { + "epoch": 0.4274870155813024, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1070 + }, + { + "epoch": 0.42788653615661204, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1071 + }, + { + "epoch": 0.4282860567319217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1072 + }, + { + "epoch": 0.42868557730723134, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1073 + }, + { + "epoch": 0.42908509788254096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1074 + }, + { + "epoch": 0.4294846184578506, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1075 + }, + { + "epoch": 0.4298841390331602, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1076 + }, + { + "epoch": 0.4302836596084698, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1077 + }, + { + "epoch": 0.4306831801837795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1078 + }, + { + "epoch": 0.4310827007590891, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1079 + }, + { + "epoch": 0.43148222133439873, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1080 + }, + { + "epoch": 0.43188174190970835, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1081 + }, + { + "epoch": 0.43228126248501797, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1082 + }, + { + "epoch": 0.4326807830603276, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3779, + "step": 1083 + }, + { + "epoch": 0.4330803036356372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1084 + }, + { + "epoch": 0.4334798242109469, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1085 + }, + { + "epoch": 0.4338793447862565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1086 + }, + { + "epoch": 0.4342788653615661, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3766, + "step": 1087 + }, + { + "epoch": 0.43467838593687574, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3716, + "step": 1088 + }, + { + "epoch": 0.43507790651218536, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1089 + }, + { + "epoch": 0.435477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1090 + }, + { + "epoch": 0.43587694766280466, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1091 + }, + { + "epoch": 0.4362764682381143, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1092 + }, + { + "epoch": 0.4366759888134239, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1093 + }, + { + "epoch": 0.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1094 + }, + { + "epoch": 0.43747502996404314, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1095 + }, + { + "epoch": 0.43787455053935276, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 1096 + }, + { + "epoch": 0.4382740711146624, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1097 + }, + { + "epoch": 0.43867359168997205, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1098 + }, + { + "epoch": 0.4390731122652817, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1099 + }, + { + "epoch": 0.4394726328405913, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1100 + }, + { + "epoch": 0.4398721534159009, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1101 + }, + { + "epoch": 0.44027167399121053, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1102 + }, + { + "epoch": 0.44067119456652015, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3712, + "step": 1103 + }, + { + "epoch": 0.4410707151418298, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1104 + }, + { + "epoch": 0.44147023571713945, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1105 + }, + { + "epoch": 0.44186975629244907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1106 + }, + { + "epoch": 0.4422692768677587, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1107 + }, + { + "epoch": 0.4426687974430683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1108 + }, + { + "epoch": 0.4430683180183779, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1109 + }, + { + "epoch": 0.4434678385936876, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3695, + "step": 1110 + }, + { + "epoch": 0.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1111 + }, + { + "epoch": 0.44426687974430684, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1112 + }, + { + "epoch": 0.44466640031961646, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1113 + }, + { + "epoch": 0.4450659208949261, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1114 + }, + { + "epoch": 0.4454654414702357, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1115 + }, + { + "epoch": 0.4458649620455453, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1116 + }, + { + "epoch": 0.446264482620855, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3744, + "step": 1117 + }, + { + "epoch": 0.4466640031961646, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1118 + }, + { + "epoch": 0.44706352377147424, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1119 + }, + { + "epoch": 0.44746304434678386, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1120 + }, + { + "epoch": 0.4478625649220935, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1121 + }, + { + "epoch": 0.4482620854974031, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1122 + }, + { + "epoch": 0.44866160607271277, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1123 + }, + { + "epoch": 0.4490611266480224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1124 + }, + { + "epoch": 0.449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3714, + "step": 1125 + }, + { + "epoch": 0.44986016779864163, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.37, + "step": 1126 + }, + { + "epoch": 0.45025968837395125, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1127 + }, + { + "epoch": 0.45065920894926087, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1128 + }, + { + "epoch": 0.4510587295245705, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1129 + }, + { + "epoch": 0.45145825009988017, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1130 + }, + { + "epoch": 0.4518577706751898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3707, + "step": 1131 + }, + { + "epoch": 0.4522572912504994, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1132 + }, + { + "epoch": 0.452656811825809, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3701, + "step": 1133 + }, + { + "epoch": 0.45305633240111864, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1134 + }, + { + "epoch": 0.45345585297642826, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1135 + }, + { + "epoch": 0.45385537355173794, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1136 + }, + { + "epoch": 0.45425489412704756, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1137 + }, + { + "epoch": 0.4546544147023572, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1138 + }, + { + "epoch": 0.4550539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3681, + "step": 1139 + }, + { + "epoch": 0.4554534558529764, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1140 + }, + { + "epoch": 0.45585297642828604, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3699, + "step": 1141 + }, + { + "epoch": 0.45625249700359566, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1142 + }, + { + "epoch": 0.45665201757890533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1143 + }, + { + "epoch": 0.45705153815421495, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1144 + }, + { + "epoch": 0.4574510587295246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1145 + }, + { + "epoch": 0.4578505793048342, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1146 + }, + { + "epoch": 0.4582500998801438, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1147 + }, + { + "epoch": 0.45864962045545343, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1148 + }, + { + "epoch": 0.4590491410307631, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1149 + }, + { + "epoch": 0.45944866160607273, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1150 + }, + { + "epoch": 0.45984818218138235, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3724, + "step": 1151 + }, + { + "epoch": 0.46024770275669197, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1152 + }, + { + "epoch": 0.4606472233320016, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.369, + "step": 1153 + }, + { + "epoch": 0.4610467439073112, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1154 + }, + { + "epoch": 0.4614462644826209, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1155 + }, + { + "epoch": 0.4618457850579305, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.366, + "step": 1156 + }, + { + "epoch": 0.4622453056332401, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3632, + "step": 1157 + }, + { + "epoch": 0.46264482620854974, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1158 + }, + { + "epoch": 0.46304434678385936, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1159 + }, + { + "epoch": 0.463443867359169, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 1160 + }, + { + "epoch": 0.4638433879344786, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1161 + }, + { + "epoch": 0.4642429085097883, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1162 + }, + { + "epoch": 0.4646424290850979, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1163 + }, + { + "epoch": 0.4650419496604075, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3769, + "step": 1164 + }, + { + "epoch": 0.46544147023571714, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1165 + }, + { + "epoch": 0.46584099081102676, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1166 + }, + { + "epoch": 0.4662405113863364, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1167 + }, + { + "epoch": 0.46664003196164605, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1168 + }, + { + "epoch": 0.46703955253695567, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1169 + }, + { + "epoch": 0.4674390731122653, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1170 + }, + { + "epoch": 0.4678385936875749, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1171 + }, + { + "epoch": 0.46823811426288453, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1172 + }, + { + "epoch": 0.46863763483819415, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1173 + }, + { + "epoch": 0.46903715541350377, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3609, + "step": 1174 + }, + { + "epoch": 0.46943667598881345, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1175 + }, + { + "epoch": 0.46983619656412307, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1176 + }, + { + "epoch": 0.4702357171394327, + "grad_norm": 0.19921875, + "learning_rate": 0.0002, + "loss": 1.3705, + "step": 1177 + }, + { + "epoch": 0.4706352377147423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3753, + "step": 1178 + }, + { + "epoch": 0.4710347582900519, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3683, + "step": 1179 + }, + { + "epoch": 0.47143427886536154, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1180 + }, + { + "epoch": 0.4718337994406712, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3693, + "step": 1181 + }, + { + "epoch": 0.47223332001598084, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1182 + }, + { + "epoch": 0.47263284059129046, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1183 + }, + { + "epoch": 0.4730323611666001, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1184 + }, + { + "epoch": 0.4734318817419097, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1185 + }, + { + "epoch": 0.4738314023172193, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1186 + }, + { + "epoch": 0.47423092289252894, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1187 + }, + { + "epoch": 0.4746304434678386, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1188 + }, + { + "epoch": 0.47502996404314823, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3657, + "step": 1189 + }, + { + "epoch": 0.47542948461845785, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3607, + "step": 1190 + }, + { + "epoch": 0.4758290051937675, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1191 + }, + { + "epoch": 0.4762285257690771, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1192 + }, + { + "epoch": 0.4766280463443867, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3613, + "step": 1193 + }, + { + "epoch": 0.4770275669196964, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1194 + }, + { + "epoch": 0.477427087495006, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1195 + }, + { + "epoch": 0.47782660807031563, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3677, + "step": 1196 + }, + { + "epoch": 0.47822612864562525, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1197 + }, + { + "epoch": 0.47862564922093487, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1198 + }, + { + "epoch": 0.4790251697962445, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1199 + }, + { + "epoch": 0.4794246903715541, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1200 + }, + { + "epoch": 0.4798242109468638, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1201 + }, + { + "epoch": 0.4802237315221734, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1202 + }, + { + "epoch": 0.480623252097483, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1203 + }, + { + "epoch": 0.48102277267279264, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3655, + "step": 1204 + }, + { + "epoch": 0.48142229324810226, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3645, + "step": 1205 + }, + { + "epoch": 0.4818218138234119, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3654, + "step": 1206 + }, + { + "epoch": 0.48222133439872156, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1207 + }, + { + "epoch": 0.4826208549740312, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1208 + }, + { + "epoch": 0.4830203755493408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1209 + }, + { + "epoch": 0.4834198961246504, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1210 + }, + { + "epoch": 0.48381941669996004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1211 + }, + { + "epoch": 0.48421893727526966, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1212 + }, + { + "epoch": 0.48461845785057933, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1213 + }, + { + "epoch": 0.48501797842588895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1214 + }, + { + "epoch": 0.48541749900119857, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3664, + "step": 1215 + }, + { + "epoch": 0.4858170195765082, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1216 + }, + { + "epoch": 0.4862165401518178, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 1217 + }, + { + "epoch": 0.48661606072712743, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3688, + "step": 1218 + }, + { + "epoch": 0.48701558130243705, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1219 + }, + { + "epoch": 0.4874151018777467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1220 + }, + { + "epoch": 0.48781462245305635, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3639, + "step": 1221 + }, + { + "epoch": 0.48821414302836597, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1222 + }, + { + "epoch": 0.4886136636036756, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3692, + "step": 1223 + }, + { + "epoch": 0.4890131841789852, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1224 + }, + { + "epoch": 0.4894127047542948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3667, + "step": 1225 + }, + { + "epoch": 0.4898122253296045, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1226 + }, + { + "epoch": 0.4902117459049141, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1227 + }, + { + "epoch": 0.49061126648022374, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1228 + }, + { + "epoch": 0.49101078705553336, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1229 + }, + { + "epoch": 0.491410307630843, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3643, + "step": 1230 + }, + { + "epoch": 0.4918098282061526, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1231 + }, + { + "epoch": 0.4922093487814622, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3587, + "step": 1232 + }, + { + "epoch": 0.4926088693567719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1233 + }, + { + "epoch": 0.4930083899320815, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3685, + "step": 1234 + }, + { + "epoch": 0.49340791050739113, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1235 + }, + { + "epoch": 0.49380743108270075, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1236 + }, + { + "epoch": 0.4942069516580104, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3772, + "step": 1237 + }, + { + "epoch": 0.49460647223332, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1238 + }, + { + "epoch": 0.49500599280862967, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1239 + }, + { + "epoch": 0.4954055133839393, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1240 + }, + { + "epoch": 0.4958050339592489, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1241 + }, + { + "epoch": 0.49620455453455853, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1242 + }, + { + "epoch": 0.49660407510986815, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3631, + "step": 1243 + }, + { + "epoch": 0.49700359568517777, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1244 + }, + { + "epoch": 0.4974031162604874, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1245 + }, + { + "epoch": 0.49780263683579706, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1246 + }, + { + "epoch": 0.4982021574111067, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1247 + }, + { + "epoch": 0.4986016779864163, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1248 + }, + { + "epoch": 0.4990011985617259, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1249 + }, + { + "epoch": 0.49940071913703554, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1250 + }, + { + "epoch": 0.49980023971234516, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1251 + }, + { + "epoch": 0.5001997602876548, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 1252 + }, + { + "epoch": 0.5005992808629645, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1253 + }, + { + "epoch": 0.500998801438274, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1254 + }, + { + "epoch": 0.5013983220135837, + "grad_norm": 0.2060546875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1255 + }, + { + "epoch": 0.5017978425888934, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1256 + }, + { + "epoch": 0.5021973631642029, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1257 + }, + { + "epoch": 0.5025968837395126, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1258 + }, + { + "epoch": 0.5029964043148222, + "grad_norm": 0.275390625, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1259 + }, + { + "epoch": 0.5033959248901319, + "grad_norm": 0.2197265625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1260 + }, + { + "epoch": 0.5037954454654414, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.367, + "step": 1261 + }, + { + "epoch": 0.5041949660407511, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3636, + "step": 1262 + }, + { + "epoch": 0.5045944866160608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.374, + "step": 1263 + }, + { + "epoch": 0.5049940071913703, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1264 + }, + { + "epoch": 0.50539352776668, + "grad_norm": 0.169921875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1265 + }, + { + "epoch": 0.5057930483419896, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1266 + }, + { + "epoch": 0.5061925689172992, + "grad_norm": 0.26171875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1267 + }, + { + "epoch": 0.5065920894926089, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3678, + "step": 1268 + }, + { + "epoch": 0.5069916100679185, + "grad_norm": 0.2001953125, + "learning_rate": 0.0002, + "loss": 1.365, + "step": 1269 + }, + { + "epoch": 0.5073911306432282, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3673, + "step": 1270 + }, + { + "epoch": 0.5077906512185377, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3702, + "step": 1271 + }, + { + "epoch": 0.5081901717938474, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1272 + }, + { + "epoch": 0.508589692369157, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 1273 + }, + { + "epoch": 0.5089892129444666, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1274 + }, + { + "epoch": 0.5093887335197763, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1275 + }, + { + "epoch": 0.5097882540950859, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1276 + }, + { + "epoch": 0.5101877746703956, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1277 + }, + { + "epoch": 0.5105872952457051, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3737, + "step": 1278 + }, + { + "epoch": 0.5109868158210148, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1279 + }, + { + "epoch": 0.5113863363963244, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1280 + }, + { + "epoch": 0.511785856971634, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3703, + "step": 1281 + }, + { + "epoch": 0.5121853775469437, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1282 + }, + { + "epoch": 0.5125848981222533, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1283 + }, + { + "epoch": 0.512984418697563, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1284 + }, + { + "epoch": 0.5133839392728725, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1285 + }, + { + "epoch": 0.5137834598481822, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.368, + "step": 1286 + }, + { + "epoch": 0.5141829804234919, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1287 + }, + { + "epoch": 0.5145825009988014, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1288 + }, + { + "epoch": 0.5149820215741111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1289 + }, + { + "epoch": 0.5153815421494207, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1290 + }, + { + "epoch": 0.5157810627247303, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1291 + }, + { + "epoch": 0.5161805833000399, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3785, + "step": 1292 + }, + { + "epoch": 0.5165801038753496, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1293 + }, + { + "epoch": 0.5169796244506593, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1294 + }, + { + "epoch": 0.5173791450259688, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1295 + }, + { + "epoch": 0.5177786656012785, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 1296 + }, + { + "epoch": 0.5181781861765881, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1297 + }, + { + "epoch": 0.5185777067518977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1298 + }, + { + "epoch": 0.5189772273272073, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1299 + }, + { + "epoch": 0.519376747902517, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1300 + }, + { + "epoch": 0.5197762684778267, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1301 + }, + { + "epoch": 0.5201757890531362, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1302 + }, + { + "epoch": 0.5205753096284459, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1303 + }, + { + "epoch": 0.5209748302037555, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1304 + }, + { + "epoch": 0.5213743507790651, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1305 + }, + { + "epoch": 0.5217738713543747, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1306 + }, + { + "epoch": 0.5221733919296844, + "grad_norm": 0.51171875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1307 + }, + { + "epoch": 0.522572912504994, + "grad_norm": 0.5859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1308 + }, + { + "epoch": 0.5229724330803036, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1309 + }, + { + "epoch": 0.5233719536556133, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1310 + }, + { + "epoch": 0.5237714742309229, + "grad_norm": 0.388671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1311 + }, + { + "epoch": 0.5241709948062325, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1312 + }, + { + "epoch": 0.5245705153815422, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1313 + }, + { + "epoch": 0.5249700359568518, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1314 + }, + { + "epoch": 0.5253695565321614, + "grad_norm": 0.357421875, + "learning_rate": 0.0002, + "loss": 1.3676, + "step": 1315 + }, + { + "epoch": 0.525769077107471, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1316 + }, + { + "epoch": 0.5261685976827807, + "grad_norm": 0.298828125, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1317 + }, + { + "epoch": 0.5265681182580902, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1318 + }, + { + "epoch": 0.5269676388333999, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3832, + "step": 1319 + }, + { + "epoch": 0.5273671594087096, + "grad_norm": 0.166015625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1320 + }, + { + "epoch": 0.5277666799840192, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1321 + }, + { + "epoch": 0.5281662005593288, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1322 + }, + { + "epoch": 0.5285657211346384, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 1323 + }, + { + "epoch": 0.5289652417099481, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3696, + "step": 1324 + }, + { + "epoch": 0.5293647622852576, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1325 + }, + { + "epoch": 0.5297642828605673, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1326 + }, + { + "epoch": 0.530163803435877, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1327 + }, + { + "epoch": 0.5305633240111866, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3635, + "step": 1328 + }, + { + "epoch": 0.5309628445864962, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1329 + }, + { + "epoch": 0.5313623651618058, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1330 + }, + { + "epoch": 0.5317618857371155, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1331 + }, + { + "epoch": 0.5321614063124251, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1332 + }, + { + "epoch": 0.5325609268877347, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1333 + }, + { + "epoch": 0.5329604474630444, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1334 + }, + { + "epoch": 0.533359968038354, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1335 + }, + { + "epoch": 0.5337594886136636, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1336 + }, + { + "epoch": 0.5341590091889732, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1337 + }, + { + "epoch": 0.5345585297642829, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1338 + }, + { + "epoch": 0.5349580503395925, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1339 + }, + { + "epoch": 0.5353575709149021, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1340 + }, + { + "epoch": 0.5357570914902118, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1341 + }, + { + "epoch": 0.5361566120655213, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1342 + }, + { + "epoch": 0.536556132640831, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1343 + }, + { + "epoch": 0.5369556532161406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3659, + "step": 1344 + }, + { + "epoch": 0.5373551737914503, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1345 + }, + { + "epoch": 0.5377546943667599, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1346 + }, + { + "epoch": 0.5381542149420695, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3603, + "step": 1347 + }, + { + "epoch": 0.5385537355173792, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1348 + }, + { + "epoch": 0.5389532560926887, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 1349 + }, + { + "epoch": 0.5393527766679984, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1350 + }, + { + "epoch": 0.539752297243308, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1351 + }, + { + "epoch": 0.5401518178186177, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1352 + }, + { + "epoch": 0.5405513383939273, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1353 + }, + { + "epoch": 0.5409508589692369, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3759, + "step": 1354 + }, + { + "epoch": 0.5413503795445466, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1355 + }, + { + "epoch": 0.5417499001198561, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3598, + "step": 1356 + }, + { + "epoch": 0.5421494206951658, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1357 + }, + { + "epoch": 0.5425489412704755, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1358 + }, + { + "epoch": 0.542948461845785, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1359 + }, + { + "epoch": 0.5433479824210947, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1360 + }, + { + "epoch": 0.5437475029964043, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1361 + }, + { + "epoch": 0.544147023571714, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1362 + }, + { + "epoch": 0.5445465441470235, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1363 + }, + { + "epoch": 0.5449460647223332, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3608, + "step": 1364 + }, + { + "epoch": 0.5453455852976429, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1365 + }, + { + "epoch": 0.5457451058729524, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1366 + }, + { + "epoch": 0.5461446264482621, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1367 + }, + { + "epoch": 0.5465441470235717, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3633, + "step": 1368 + }, + { + "epoch": 0.5469436675988814, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1369 + }, + { + "epoch": 0.5473431881741909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1370 + }, + { + "epoch": 0.5477427087495006, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1371 + }, + { + "epoch": 0.5481422293248103, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 1372 + }, + { + "epoch": 0.5485417499001198, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1373 + }, + { + "epoch": 0.5489412704754295, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1374 + }, + { + "epoch": 0.5493407910507391, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1375 + }, + { + "epoch": 0.5497403116260487, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3648, + "step": 1376 + }, + { + "epoch": 0.5501398322013584, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1377 + }, + { + "epoch": 0.550539352776668, + "grad_norm": 0.431640625, + "learning_rate": 0.0002, + "loss": 1.3626, + "step": 1378 + }, + { + "epoch": 0.5509388733519777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1379 + }, + { + "epoch": 0.5513383939272872, + "grad_norm": 0.326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1380 + }, + { + "epoch": 0.5517379145025969, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1381 + }, + { + "epoch": 0.5521374350779065, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1382 + }, + { + "epoch": 0.5525369556532161, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1383 + }, + { + "epoch": 0.5529364762285258, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1384 + }, + { + "epoch": 0.5533359968038354, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1385 + }, + { + "epoch": 0.5537355173791451, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1386 + }, + { + "epoch": 0.5541350379544546, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3709, + "step": 1387 + }, + { + "epoch": 0.5545345585297643, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1388 + }, + { + "epoch": 0.5549340791050739, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1389 + }, + { + "epoch": 0.5553335996803835, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3604, + "step": 1390 + }, + { + "epoch": 0.5557331202556932, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1391 + }, + { + "epoch": 0.5561326408310028, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1392 + }, + { + "epoch": 0.5565321614063125, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3674, + "step": 1393 + }, + { + "epoch": 0.556931681981622, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3689, + "step": 1394 + }, + { + "epoch": 0.5573312025569317, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1395 + }, + { + "epoch": 0.5577307231322413, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1396 + }, + { + "epoch": 0.5581302437075509, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1397 + }, + { + "epoch": 0.5585297642828606, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3706, + "step": 1398 + }, + { + "epoch": 0.5589292848581702, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1399 + }, + { + "epoch": 0.5593288054334798, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1400 + }, + { + "epoch": 0.5597283260087894, + "grad_norm": 0.26953125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1401 + }, + { + "epoch": 0.5601278465840991, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1402 + }, + { + "epoch": 0.5605273671594088, + "grad_norm": 0.177734375, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1403 + }, + { + "epoch": 0.5609268877347183, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1404 + }, + { + "epoch": 0.561326408310028, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 1405 + }, + { + "epoch": 0.5617259288853376, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1406 + }, + { + "epoch": 0.5621254494606472, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 1407 + }, + { + "epoch": 0.5625249700359568, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1408 + }, + { + "epoch": 0.5629244906112665, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1409 + }, + { + "epoch": 0.5633240111865762, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1410 + }, + { + "epoch": 0.5637235317618857, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1411 + }, + { + "epoch": 0.5641230523371954, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1412 + }, + { + "epoch": 0.564522572912505, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1413 + }, + { + "epoch": 0.5649220934878146, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1414 + }, + { + "epoch": 0.5653216140631242, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1415 + }, + { + "epoch": 0.5657211346384339, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1416 + }, + { + "epoch": 0.5661206552137436, + "grad_norm": 0.2021484375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1417 + }, + { + "epoch": 0.5665201757890531, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1418 + }, + { + "epoch": 0.5669196963643628, + "grad_norm": 0.220703125, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1419 + }, + { + "epoch": 0.5673192169396724, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1420 + }, + { + "epoch": 0.567718737514982, + "grad_norm": 0.423828125, + "learning_rate": 0.0002, + "loss": 1.364, + "step": 1421 + }, + { + "epoch": 0.5681182580902916, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3715, + "step": 1422 + }, + { + "epoch": 0.5685177786656013, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1423 + }, + { + "epoch": 0.568917299240911, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 1424 + }, + { + "epoch": 0.5693168198162205, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1425 + }, + { + "epoch": 0.5697163403915302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3612, + "step": 1426 + }, + { + "epoch": 0.5701158609668397, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1427 + }, + { + "epoch": 0.5705153815421494, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1428 + }, + { + "epoch": 0.5709149021174591, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1429 + }, + { + "epoch": 0.5713144226927687, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 1430 + }, + { + "epoch": 0.5717139432680783, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1431 + }, + { + "epoch": 0.5721134638433879, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1432 + }, + { + "epoch": 0.5725129844186976, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1433 + }, + { + "epoch": 0.5729125049940071, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1434 + }, + { + "epoch": 0.5733120255693168, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1435 + }, + { + "epoch": 0.5737115461446265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1436 + }, + { + "epoch": 0.5741110667199361, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1437 + }, + { + "epoch": 0.5745105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1438 + }, + { + "epoch": 0.5749101078705553, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.362, + "step": 1439 + }, + { + "epoch": 0.575309628445865, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1440 + }, + { + "epoch": 0.5757091490211745, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1441 + }, + { + "epoch": 0.5761086695964842, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1442 + }, + { + "epoch": 0.5765081901717939, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3717, + "step": 1443 + }, + { + "epoch": 0.5769077107471035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1444 + }, + { + "epoch": 0.5773072313224131, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1445 + }, + { + "epoch": 0.5777067518977227, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 1446 + }, + { + "epoch": 0.5781062724730324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1447 + }, + { + "epoch": 0.578505793048342, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 1448 + }, + { + "epoch": 0.5789053136236516, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1449 + }, + { + "epoch": 0.5793048341989613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 1450 + }, + { + "epoch": 0.5797043547742708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1451 + }, + { + "epoch": 0.5801038753495805, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1452 + }, + { + "epoch": 0.5805033959248901, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1453 + }, + { + "epoch": 0.5809029165001998, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1454 + }, + { + "epoch": 0.5813024370755094, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1455 + }, + { + "epoch": 0.581701957650819, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1456 + }, + { + "epoch": 0.5821014782261287, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1457 + }, + { + "epoch": 0.5825009988014382, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 1458 + }, + { + "epoch": 0.5829005193767479, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1459 + }, + { + "epoch": 0.5833000399520575, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1460 + }, + { + "epoch": 0.5836995605273672, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1461 + }, + { + "epoch": 0.5840990811026768, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1462 + }, + { + "epoch": 0.5844986016779864, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1463 + }, + { + "epoch": 0.5848981222532961, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1464 + }, + { + "epoch": 0.5852976428286056, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1465 + }, + { + "epoch": 0.5856971634039153, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 1466 + }, + { + "epoch": 0.5860966839792249, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1467 + }, + { + "epoch": 0.5864962045545346, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1468 + }, + { + "epoch": 0.5868957251298442, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1469 + }, + { + "epoch": 0.5872952457051538, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1470 + }, + { + "epoch": 0.5876947662804635, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1471 + }, + { + "epoch": 0.588094286855773, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1472 + }, + { + "epoch": 0.5884938074310827, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1473 + }, + { + "epoch": 0.5888933280063924, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1474 + }, + { + "epoch": 0.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1475 + }, + { + "epoch": 0.5896923691570116, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1476 + }, + { + "epoch": 0.5900918897323212, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1477 + }, + { + "epoch": 0.5904914103076309, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1478 + }, + { + "epoch": 0.5908909308829404, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1479 + }, + { + "epoch": 0.5912904514582501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1480 + }, + { + "epoch": 0.5916899720335598, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 1481 + }, + { + "epoch": 0.5920894926088693, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 1482 + }, + { + "epoch": 0.592489013184179, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1483 + }, + { + "epoch": 0.5928885337594886, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1484 + }, + { + "epoch": 0.5932880543347983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1485 + }, + { + "epoch": 0.5936875749101078, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1486 + }, + { + "epoch": 0.5940870954854175, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1487 + }, + { + "epoch": 0.5944866160607272, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1488 + }, + { + "epoch": 0.5948861366360367, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1489 + }, + { + "epoch": 0.5952856572113464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1490 + }, + { + "epoch": 0.595685177786656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 1491 + }, + { + "epoch": 0.5960846983619656, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1492 + }, + { + "epoch": 0.5964842189372753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1493 + }, + { + "epoch": 0.5968837395125849, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1494 + }, + { + "epoch": 0.5972832600878946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3614, + "step": 1495 + }, + { + "epoch": 0.5976827806632041, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1496 + }, + { + "epoch": 0.5980823012385138, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1497 + }, + { + "epoch": 0.5984818218138234, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 1498 + }, + { + "epoch": 0.598881342389133, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1499 + }, + { + "epoch": 0.5992808629644427, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1500 + }, + { + "epoch": 0.5996803835397523, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1501 + }, + { + "epoch": 0.600079904115062, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 1502 + }, + { + "epoch": 0.6004794246903715, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1503 + }, + { + "epoch": 0.6008789452656812, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1504 + }, + { + "epoch": 0.6012784658409908, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1505 + }, + { + "epoch": 0.6016779864163004, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1506 + }, + { + "epoch": 0.6020775069916101, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1507 + }, + { + "epoch": 0.6024770275669197, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 1508 + }, + { + "epoch": 0.6028765481422294, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1509 + }, + { + "epoch": 0.6032760687175389, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 1510 + }, + { + "epoch": 0.6036755892928486, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1511 + }, + { + "epoch": 0.6040751098681582, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1512 + }, + { + "epoch": 0.6044746304434678, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3669, + "step": 1513 + }, + { + "epoch": 0.6048741510187775, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1514 + }, + { + "epoch": 0.6052736715940871, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3658, + "step": 1515 + }, + { + "epoch": 0.6056731921693967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1516 + }, + { + "epoch": 0.6060727127447063, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1517 + }, + { + "epoch": 0.606472233320016, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1518 + }, + { + "epoch": 0.6068717538953257, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3601, + "step": 1519 + }, + { + "epoch": 0.6072712744706352, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1520 + }, + { + "epoch": 0.6076707950459449, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1521 + }, + { + "epoch": 0.6080703156212545, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1522 + }, + { + "epoch": 0.6084698361965641, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1523 + }, + { + "epoch": 0.6088693567718737, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1524 + }, + { + "epoch": 0.6092688773471834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1525 + }, + { + "epoch": 0.609668397922493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1526 + }, + { + "epoch": 0.6100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1527 + }, + { + "epoch": 0.6104674390731123, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1528 + }, + { + "epoch": 0.6108669596484219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3675, + "step": 1529 + }, + { + "epoch": 0.6112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1530 + }, + { + "epoch": 0.6116660007990411, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1531 + }, + { + "epoch": 0.6120655213743508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1532 + }, + { + "epoch": 0.6124650419496604, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1533 + }, + { + "epoch": 0.61286456252497, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1534 + }, + { + "epoch": 0.6132640831002797, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1535 + }, + { + "epoch": 0.6136636036755893, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1536 + }, + { + "epoch": 0.6140631242508989, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1537 + }, + { + "epoch": 0.6144626448262086, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1538 + }, + { + "epoch": 0.6148621654015182, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3656, + "step": 1539 + }, + { + "epoch": 0.6152616859768278, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1540 + }, + { + "epoch": 0.6156612065521374, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1541 + }, + { + "epoch": 0.6160607271274471, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1542 + }, + { + "epoch": 0.6164602477027566, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1543 + }, + { + "epoch": 0.6168597682780663, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1544 + }, + { + "epoch": 0.617259288853376, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1545 + }, + { + "epoch": 0.6176588094286856, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1546 + }, + { + "epoch": 0.6180583300039952, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1547 + }, + { + "epoch": 0.6184578505793048, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3665, + "step": 1548 + }, + { + "epoch": 0.6188573711546145, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1549 + }, + { + "epoch": 0.619256891729924, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1550 + }, + { + "epoch": 0.6196564123052337, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1551 + }, + { + "epoch": 0.6200559328805434, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1552 + }, + { + "epoch": 0.620455453455853, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1553 + }, + { + "epoch": 0.6208549740311626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 1554 + }, + { + "epoch": 0.6212544946064722, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1555 + }, + { + "epoch": 0.6216540151817819, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 1556 + }, + { + "epoch": 0.6220535357570914, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1557 + }, + { + "epoch": 0.6224530563324011, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1558 + }, + { + "epoch": 0.6228525769077108, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1559 + }, + { + "epoch": 0.6232520974830204, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1560 + }, + { + "epoch": 0.62365161805833, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1561 + }, + { + "epoch": 0.6240511386336396, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1562 + }, + { + "epoch": 0.6244506592089493, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3686, + "step": 1563 + }, + { + "epoch": 0.6248501797842589, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1564 + }, + { + "epoch": 0.6252497003595685, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1565 + }, + { + "epoch": 0.6256492209348782, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1566 + }, + { + "epoch": 0.6260487415101877, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1567 + }, + { + "epoch": 0.6264482620854974, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3619, + "step": 1568 + }, + { + "epoch": 0.626847782660807, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1569 + }, + { + "epoch": 0.6272473032361167, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1570 + }, + { + "epoch": 0.6276468238114263, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3569, + "step": 1571 + }, + { + "epoch": 0.6280463443867359, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1572 + }, + { + "epoch": 0.6284458649620456, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1573 + }, + { + "epoch": 0.6288453855373551, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1574 + }, + { + "epoch": 0.6292449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3624, + "step": 1575 + }, + { + "epoch": 0.6296444266879744, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 1576 + }, + { + "epoch": 0.630043947263284, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1577 + }, + { + "epoch": 0.6304434678385937, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1578 + }, + { + "epoch": 0.6308429884139033, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1579 + }, + { + "epoch": 0.631242508989213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1580 + }, + { + "epoch": 0.6316420295645225, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1581 + }, + { + "epoch": 0.6320415501398322, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1582 + }, + { + "epoch": 0.6324410707151419, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 1583 + }, + { + "epoch": 0.6328405912904514, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1584 + }, + { + "epoch": 0.6332401118657611, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 1585 + }, + { + "epoch": 0.6336396324410707, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1586 + }, + { + "epoch": 0.6340391530163804, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1587 + }, + { + "epoch": 0.6344386735916899, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 1588 + }, + { + "epoch": 0.6348381941669996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3585, + "step": 1589 + }, + { + "epoch": 0.6352377147423093, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1590 + }, + { + "epoch": 0.6356372353176188, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3623, + "step": 1591 + }, + { + "epoch": 0.6360367558929285, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3629, + "step": 1592 + }, + { + "epoch": 0.6364362764682381, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1593 + }, + { + "epoch": 0.6368357970435478, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1594 + }, + { + "epoch": 0.6372353176188573, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1595 + }, + { + "epoch": 0.637634838194167, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1596 + }, + { + "epoch": 0.6380343587694767, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1597 + }, + { + "epoch": 0.6384338793447862, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1598 + }, + { + "epoch": 0.6388333999200959, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1599 + }, + { + "epoch": 0.6392329204954055, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1600 + }, + { + "epoch": 0.6396324410707152, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1601 + }, + { + "epoch": 0.6400319616460247, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1602 + }, + { + "epoch": 0.6404314822213344, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1603 + }, + { + "epoch": 0.6408310027966441, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1604 + }, + { + "epoch": 0.6412305233719536, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 1605 + }, + { + "epoch": 0.6416300439472633, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1606 + }, + { + "epoch": 0.6420295645225729, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1607 + }, + { + "epoch": 0.6424290850978825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3627, + "step": 1608 + }, + { + "epoch": 0.6428286056731922, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1609 + }, + { + "epoch": 0.6432281262485018, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1610 + }, + { + "epoch": 0.6436276468238115, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1611 + }, + { + "epoch": 0.644027167399121, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1612 + }, + { + "epoch": 0.6444266879744307, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1613 + }, + { + "epoch": 0.6448262085497403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1614 + }, + { + "epoch": 0.6452257291250499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1615 + }, + { + "epoch": 0.6456252497003596, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 1616 + }, + { + "epoch": 0.6460247702756692, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1617 + }, + { + "epoch": 0.6464242908509789, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3554, + "step": 1618 + }, + { + "epoch": 0.6468238114262884, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1619 + }, + { + "epoch": 0.6472233320015981, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1620 + }, + { + "epoch": 0.6476228525769077, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1621 + }, + { + "epoch": 0.6480223731522173, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1622 + }, + { + "epoch": 0.648421893727527, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1623 + }, + { + "epoch": 0.6488214143028366, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1624 + }, + { + "epoch": 0.6492209348781462, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1625 + }, + { + "epoch": 0.6496204554534558, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1626 + }, + { + "epoch": 0.6500199760287655, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1627 + }, + { + "epoch": 0.6504194966040752, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 1628 + }, + { + "epoch": 0.6508190171793847, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1629 + }, + { + "epoch": 0.6512185377546944, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1630 + }, + { + "epoch": 0.651618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1631 + }, + { + "epoch": 0.6520175789053136, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1632 + }, + { + "epoch": 0.6524170994806232, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3562, + "step": 1633 + }, + { + "epoch": 0.6528166200559329, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 1634 + }, + { + "epoch": 0.6532161406312426, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1635 + }, + { + "epoch": 0.6536156612065521, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 1636 + }, + { + "epoch": 0.6540151817818618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1637 + }, + { + "epoch": 0.6544147023571714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1638 + }, + { + "epoch": 0.654814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1639 + }, + { + "epoch": 0.6552137435077906, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1640 + }, + { + "epoch": 0.6556132640831003, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1641 + }, + { + "epoch": 0.65601278465841, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1642 + }, + { + "epoch": 0.6564123052337195, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1643 + }, + { + "epoch": 0.6568118258090292, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 1644 + }, + { + "epoch": 0.6572113463843388, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1645 + }, + { + "epoch": 0.6576108669596484, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3682, + "step": 1646 + }, + { + "epoch": 0.658010387534958, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1647 + }, + { + "epoch": 0.6584099081102677, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1648 + }, + { + "epoch": 0.6588094286855773, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1649 + }, + { + "epoch": 0.6592089492608869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 1650 + }, + { + "epoch": 0.6596084698361966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 1651 + }, + { + "epoch": 0.6600079904115062, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1652 + }, + { + "epoch": 0.6604075109868158, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1653 + }, + { + "epoch": 0.6608070315621255, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1654 + }, + { + "epoch": 0.6612065521374351, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 1655 + }, + { + "epoch": 0.6616060727127447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1656 + }, + { + "epoch": 0.6620055932880543, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 1657 + }, + { + "epoch": 0.662405113863364, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1658 + }, + { + "epoch": 0.6628046344386735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1659 + }, + { + "epoch": 0.6632041550139832, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3642, + "step": 1660 + }, + { + "epoch": 0.6636036755892929, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1661 + }, + { + "epoch": 0.6640031961646025, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1662 + }, + { + "epoch": 0.6644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1663 + }, + { + "epoch": 0.6648022373152217, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1664 + }, + { + "epoch": 0.6652017578905314, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3541, + "step": 1665 + }, + { + "epoch": 0.6656012784658409, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 1666 + }, + { + "epoch": 0.6660007990411506, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1667 + }, + { + "epoch": 0.6664003196164603, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1668 + }, + { + "epoch": 0.6667998401917699, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1669 + }, + { + "epoch": 0.6671993607670795, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1670 + }, + { + "epoch": 0.6675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1671 + }, + { + "epoch": 0.6679984019176988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1672 + }, + { + "epoch": 0.6683979224930084, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3611, + "step": 1673 + }, + { + "epoch": 0.668797443068318, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1674 + }, + { + "epoch": 0.6691969636436277, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3625, + "step": 1675 + }, + { + "epoch": 0.6695964842189372, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1676 + }, + { + "epoch": 0.6699960047942469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 1677 + }, + { + "epoch": 0.6703955253695565, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1678 + }, + { + "epoch": 0.6707950459448662, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 1679 + }, + { + "epoch": 0.6711945665201758, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1680 + }, + { + "epoch": 0.6715940870954854, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1681 + }, + { + "epoch": 0.6719936076707951, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1682 + }, + { + "epoch": 0.6723931282461046, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1683 + }, + { + "epoch": 0.6727926488214143, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1684 + }, + { + "epoch": 0.6731921693967239, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1685 + }, + { + "epoch": 0.6735916899720336, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 1686 + }, + { + "epoch": 0.6739912105473432, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 1687 + }, + { + "epoch": 0.6743907311226528, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1688 + }, + { + "epoch": 0.6747902516979625, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 1689 + }, + { + "epoch": 0.675189772273272, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1690 + }, + { + "epoch": 0.6755892928485817, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1691 + }, + { + "epoch": 0.6759888134238913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1692 + }, + { + "epoch": 0.676388333999201, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 1693 + }, + { + "epoch": 0.6767878545745106, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1694 + }, + { + "epoch": 0.6771873751498202, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 1695 + }, + { + "epoch": 0.6775868957251299, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1696 + }, + { + "epoch": 0.6779864163004394, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1697 + }, + { + "epoch": 0.6783859368757491, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1698 + }, + { + "epoch": 0.6787854574510588, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1699 + }, + { + "epoch": 0.6791849780263683, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 1700 + }, + { + "epoch": 0.679584498601678, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3684, + "step": 1701 + }, + { + "epoch": 0.6799840191769876, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 1702 + }, + { + "epoch": 0.6803835397522973, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1703 + }, + { + "epoch": 0.6807830603276068, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1704 + }, + { + "epoch": 0.6811825809029165, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 1705 + }, + { + "epoch": 0.6815821014782262, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3544, + "step": 1706 + }, + { + "epoch": 0.6819816220535357, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 1707 + }, + { + "epoch": 0.6823811426288454, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1708 + }, + { + "epoch": 0.682780663204155, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1709 + }, + { + "epoch": 0.6831801837794647, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1710 + }, + { + "epoch": 0.6835797043547742, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1711 + }, + { + "epoch": 0.6839792249300839, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1712 + }, + { + "epoch": 0.6843787455053936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1713 + }, + { + "epoch": 0.6847782660807031, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 1714 + }, + { + "epoch": 0.6851777866560128, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 1715 + }, + { + "epoch": 0.6855773072313224, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1716 + }, + { + "epoch": 0.685976827806632, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1717 + }, + { + "epoch": 0.6863763483819417, + "grad_norm": 0.181640625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1718 + }, + { + "epoch": 0.6867758689572513, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3644, + "step": 1719 + }, + { + "epoch": 0.687175389532561, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1720 + }, + { + "epoch": 0.6875749101078705, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1721 + }, + { + "epoch": 0.6879744306831802, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1722 + }, + { + "epoch": 0.6883739512584898, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1723 + }, + { + "epoch": 0.6887734718337994, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1724 + }, + { + "epoch": 0.6891729924091091, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 1725 + }, + { + "epoch": 0.6895725129844187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1726 + }, + { + "epoch": 0.6899720335597284, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 1727 + }, + { + "epoch": 0.6903715541350379, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1728 + }, + { + "epoch": 0.6907710747103476, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1729 + }, + { + "epoch": 0.6911705952856572, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3628, + "step": 1730 + }, + { + "epoch": 0.6915701158609668, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1731 + }, + { + "epoch": 0.6919696364362765, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3605, + "step": 1732 + }, + { + "epoch": 0.6923691570115861, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1733 + }, + { + "epoch": 0.6927686775868958, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1734 + }, + { + "epoch": 0.6931681981622053, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1735 + }, + { + "epoch": 0.693567718737515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1736 + }, + { + "epoch": 0.6939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1737 + }, + { + "epoch": 0.6943667598881342, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1738 + }, + { + "epoch": 0.6947662804634439, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 1739 + }, + { + "epoch": 0.6951658010387535, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 1740 + }, + { + "epoch": 0.6955653216140631, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1741 + }, + { + "epoch": 0.6959648421893727, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 1742 + }, + { + "epoch": 0.6963643627646824, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 1743 + }, + { + "epoch": 0.6967638833399921, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1744 + }, + { + "epoch": 0.6971634039153016, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1745 + }, + { + "epoch": 0.6975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 1746 + }, + { + "epoch": 0.6979624450659209, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 1747 + }, + { + "epoch": 0.6983619656412305, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1748 + }, + { + "epoch": 0.6987614862165401, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1749 + }, + { + "epoch": 0.6991610067918498, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1750 + }, + { + "epoch": 0.6995605273671595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 1751 + }, + { + "epoch": 0.699960047942469, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1752 + }, + { + "epoch": 0.7003595685177787, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1753 + }, + { + "epoch": 0.7007590890930883, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1754 + }, + { + "epoch": 0.7011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1755 + }, + { + "epoch": 0.7015581302437075, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 1756 + }, + { + "epoch": 0.7019576508190172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 1757 + }, + { + "epoch": 0.7023571713943269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1758 + }, + { + "epoch": 0.7027566919696364, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 1759 + }, + { + "epoch": 0.7031562125449461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.357, + "step": 1760 + }, + { + "epoch": 0.7035557331202557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1761 + }, + { + "epoch": 0.7039552536955653, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 1762 + }, + { + "epoch": 0.7043547742708749, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 1763 + }, + { + "epoch": 0.7047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3577, + "step": 1764 + }, + { + "epoch": 0.7051538154214942, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1765 + }, + { + "epoch": 0.7055533359968038, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3671, + "step": 1766 + }, + { + "epoch": 0.7059528565721135, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3591, + "step": 1767 + }, + { + "epoch": 0.706352377147423, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1768 + }, + { + "epoch": 0.7067518977227327, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3649, + "step": 1769 + }, + { + "epoch": 0.7071514182980424, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 1770 + }, + { + "epoch": 0.707550938873352, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1771 + }, + { + "epoch": 0.7079504594486616, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3596, + "step": 1772 + }, + { + "epoch": 0.7083499800239712, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 1773 + }, + { + "epoch": 0.7087495005992809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3556, + "step": 1774 + }, + { + "epoch": 0.7091490211745904, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 1775 + }, + { + "epoch": 0.7095485417499001, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 1776 + }, + { + "epoch": 0.7099480623252098, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 1777 + }, + { + "epoch": 0.7103475829005194, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 1778 + }, + { + "epoch": 0.710747103475829, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1779 + }, + { + "epoch": 0.7111466240511386, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1780 + }, + { + "epoch": 0.7115461446264483, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1781 + }, + { + "epoch": 0.7119456652017578, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1782 + }, + { + "epoch": 0.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1783 + }, + { + "epoch": 0.7127447063523772, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1784 + }, + { + "epoch": 0.7131442269276868, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1785 + }, + { + "epoch": 0.7135437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1786 + }, + { + "epoch": 0.713943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 1787 + }, + { + "epoch": 0.7143427886536157, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1788 + }, + { + "epoch": 0.7147423092289253, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 1789 + }, + { + "epoch": 0.7151418298042349, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3647, + "step": 1790 + }, + { + "epoch": 0.7155413503795446, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1791 + }, + { + "epoch": 0.7159408709548541, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1792 + }, + { + "epoch": 0.7163403915301638, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1793 + }, + { + "epoch": 0.7167399121054734, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 1794 + }, + { + "epoch": 0.7171394326807831, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1795 + }, + { + "epoch": 0.7175389532560927, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1796 + }, + { + "epoch": 0.7179384738314023, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 1797 + }, + { + "epoch": 0.718337994406712, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1798 + }, + { + "epoch": 0.7187375149820215, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1799 + }, + { + "epoch": 0.7191370355573312, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 1800 + }, + { + "epoch": 0.7195365561326408, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 1801 + }, + { + "epoch": 0.7199360767079505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 1802 + }, + { + "epoch": 0.7203355972832601, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 1803 + }, + { + "epoch": 0.7207351178585697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1804 + }, + { + "epoch": 0.7211346384338794, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 1805 + }, + { + "epoch": 0.7215341590091889, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1806 + }, + { + "epoch": 0.7219336795844986, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.363, + "step": 1807 + }, + { + "epoch": 0.7223332001598082, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1808 + }, + { + "epoch": 0.7227327207351179, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1809 + }, + { + "epoch": 0.7231322413104275, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.359, + "step": 1810 + }, + { + "epoch": 0.7235317618857371, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1811 + }, + { + "epoch": 0.7239312824610468, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1812 + }, + { + "epoch": 0.7243308030363563, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 1813 + }, + { + "epoch": 0.724730323611666, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1814 + }, + { + "epoch": 0.7251298441869757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1815 + }, + { + "epoch": 0.7255293647622852, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1816 + }, + { + "epoch": 0.7259288853375949, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3571, + "step": 1817 + }, + { + "epoch": 0.7263284059129045, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3638, + "step": 1818 + }, + { + "epoch": 0.7267279264882142, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1819 + }, + { + "epoch": 0.7271274470635237, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1820 + }, + { + "epoch": 0.7275269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 1821 + }, + { + "epoch": 0.7279264882141431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 1822 + }, + { + "epoch": 0.7283260087894526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1823 + }, + { + "epoch": 0.7287255293647623, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 1824 + }, + { + "epoch": 0.7291250499400719, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 1825 + }, + { + "epoch": 0.7295245705153816, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1826 + }, + { + "epoch": 0.7299240910906911, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 1827 + }, + { + "epoch": 0.7303236116660008, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1828 + }, + { + "epoch": 0.7307231322413105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3691, + "step": 1829 + }, + { + "epoch": 0.73112265281662, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1830 + }, + { + "epoch": 0.7315221733919297, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1831 + }, + { + "epoch": 0.7319216939672393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1832 + }, + { + "epoch": 0.732321214542549, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1833 + }, + { + "epoch": 0.7327207351178586, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 1834 + }, + { + "epoch": 0.7331202556931682, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 1835 + }, + { + "epoch": 0.7335197762684779, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 1836 + }, + { + "epoch": 0.7339192968437874, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 1837 + }, + { + "epoch": 0.7343188174190971, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1838 + }, + { + "epoch": 0.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 1839 + }, + { + "epoch": 0.7351178585697163, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 1840 + }, + { + "epoch": 0.735517379145026, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1841 + }, + { + "epoch": 0.7359168997203356, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 1842 + }, + { + "epoch": 0.7363164202956453, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 1843 + }, + { + "epoch": 0.7367159408709548, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 1844 + }, + { + "epoch": 0.7371154614462645, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 1845 + }, + { + "epoch": 0.7375149820215741, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1846 + }, + { + "epoch": 0.7379145025968837, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1847 + }, + { + "epoch": 0.7383140231721934, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 1848 + }, + { + "epoch": 0.738713543747503, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 1849 + }, + { + "epoch": 0.7391130643228127, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 1850 + }, + { + "epoch": 0.7395125848981222, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1851 + }, + { + "epoch": 0.7399121054734319, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1852 + }, + { + "epoch": 0.7403116260487415, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1853 + }, + { + "epoch": 0.7407111466240511, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 1854 + }, + { + "epoch": 0.7411106671993608, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1855 + }, + { + "epoch": 0.7415101877746704, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3621, + "step": 1856 + }, + { + "epoch": 0.74190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1857 + }, + { + "epoch": 0.7423092289252896, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 1858 + }, + { + "epoch": 0.7427087495005993, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 1859 + }, + { + "epoch": 0.743108270075909, + "grad_norm": 0.2138671875, + "learning_rate": 0.0002, + "loss": 1.361, + "step": 1860 + }, + { + "epoch": 0.7435077906512185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1861 + }, + { + "epoch": 0.7439073112265282, + "grad_norm": 0.265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1862 + }, + { + "epoch": 0.7443068318018378, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 1863 + }, + { + "epoch": 0.7447063523771474, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1864 + }, + { + "epoch": 0.745105872952457, + "grad_norm": 0.267578125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1865 + }, + { + "epoch": 0.7455053935277667, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1866 + }, + { + "epoch": 0.7459049141030764, + "grad_norm": 0.1630859375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 1867 + }, + { + "epoch": 0.7463044346783859, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3606, + "step": 1868 + }, + { + "epoch": 0.7467039552536956, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1869 + }, + { + "epoch": 0.7471034758290052, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 1870 + }, + { + "epoch": 0.7475029964043148, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 1871 + }, + { + "epoch": 0.7479025169796244, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1872 + }, + { + "epoch": 0.7483020375549341, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 1873 + }, + { + "epoch": 0.7487015581302437, + "grad_norm": 0.341796875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 1874 + }, + { + "epoch": 0.7491010787055533, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 1875 + }, + { + "epoch": 0.749500599280863, + "grad_norm": 0.248046875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 1876 + }, + { + "epoch": 0.7499001198561726, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1877 + }, + { + "epoch": 0.7502996404314822, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 1878 + }, + { + "epoch": 0.7506991610067919, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 1879 + }, + { + "epoch": 0.7510986815821015, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.356, + "step": 1880 + }, + { + "epoch": 0.7514982021574111, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 1881 + }, + { + "epoch": 0.7518977227327207, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3584, + "step": 1882 + }, + { + "epoch": 0.7522972433080304, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1883 + }, + { + "epoch": 0.75269676388334, + "grad_norm": 0.171875, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1884 + }, + { + "epoch": 0.7530962844586496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 1885 + }, + { + "epoch": 0.7534958050339593, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 1886 + }, + { + "epoch": 0.7538953256092689, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 1887 + }, + { + "epoch": 0.7542948461845785, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 1888 + }, + { + "epoch": 0.7546943667598881, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 1889 + }, + { + "epoch": 0.7550938873351978, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 1890 + }, + { + "epoch": 0.7554934079105073, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 1891 + }, + { + "epoch": 0.755892928485817, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 1892 + }, + { + "epoch": 0.7562924490611267, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3599, + "step": 1893 + }, + { + "epoch": 0.7566919696364363, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1894 + }, + { + "epoch": 0.7570914902117459, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1895 + }, + { + "epoch": 0.7574910107870555, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1896 + }, + { + "epoch": 0.7578905313623652, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 1897 + }, + { + "epoch": 0.7582900519376747, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 1898 + }, + { + "epoch": 0.7586895725129844, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 1899 + }, + { + "epoch": 0.7590890930882941, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 1900 + }, + { + "epoch": 0.7594886136636037, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1901 + }, + { + "epoch": 0.7598881342389133, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1902 + }, + { + "epoch": 0.7602876548142229, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 1903 + }, + { + "epoch": 0.7606871753895326, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 1904 + }, + { + "epoch": 0.7610866959648422, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 1905 + }, + { + "epoch": 0.7614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 1906 + }, + { + "epoch": 0.7618857371154615, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 1907 + }, + { + "epoch": 0.762285257690771, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1908 + }, + { + "epoch": 0.7626847782660807, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 1909 + }, + { + "epoch": 0.7630842988413903, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 1910 + }, + { + "epoch": 0.7634838194167, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1911 + }, + { + "epoch": 0.7638833399920096, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3589, + "step": 1912 + }, + { + "epoch": 0.7642828605673192, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1913 + }, + { + "epoch": 0.7646823811426289, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1914 + }, + { + "epoch": 0.7650819017179384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3592, + "step": 1915 + }, + { + "epoch": 0.7654814222932481, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 1916 + }, + { + "epoch": 0.7658809428685577, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1917 + }, + { + "epoch": 0.7662804634438674, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1918 + }, + { + "epoch": 0.766679984019177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 1919 + }, + { + "epoch": 0.7670795045944866, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1920 + }, + { + "epoch": 0.7674790251697963, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1921 + }, + { + "epoch": 0.7678785457451058, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1922 + }, + { + "epoch": 0.7682780663204155, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3617, + "step": 1923 + }, + { + "epoch": 0.7686775868957252, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1924 + }, + { + "epoch": 0.7690771074710347, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 1925 + }, + { + "epoch": 0.7694766280463444, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 1926 + }, + { + "epoch": 0.769876148621654, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1927 + }, + { + "epoch": 0.7702756691969637, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3582, + "step": 1928 + }, + { + "epoch": 0.7706751897722732, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1929 + }, + { + "epoch": 0.7710747103475829, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 1930 + }, + { + "epoch": 0.7714742309228926, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 1931 + }, + { + "epoch": 0.7718737514982021, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1932 + }, + { + "epoch": 0.7722732720735118, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1933 + }, + { + "epoch": 0.7726727926488214, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 1934 + }, + { + "epoch": 0.7730723132241311, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 1935 + }, + { + "epoch": 0.7734718337994406, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 1936 + }, + { + "epoch": 0.7738713543747503, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 1937 + }, + { + "epoch": 0.77427087495006, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3713, + "step": 1938 + }, + { + "epoch": 0.7746703955253695, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 1939 + }, + { + "epoch": 0.7750699161006792, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 1940 + }, + { + "epoch": 0.7754694366759888, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 1941 + }, + { + "epoch": 0.7758689572512985, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 1942 + }, + { + "epoch": 0.776268477826608, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 1943 + }, + { + "epoch": 0.7766679984019177, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 1944 + }, + { + "epoch": 0.7770675189772274, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 1945 + }, + { + "epoch": 0.7774670395525369, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 1946 + }, + { + "epoch": 0.7778665601278466, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 1947 + }, + { + "epoch": 0.7782660807031562, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 1948 + }, + { + "epoch": 0.7786656012784658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 1949 + }, + { + "epoch": 0.7790651218537755, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 1950 + }, + { + "epoch": 0.7794646424290851, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 1951 + }, + { + "epoch": 0.7798641630043948, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 1952 + }, + { + "epoch": 0.7802636835797043, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 1953 + }, + { + "epoch": 0.780663204155014, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 1954 + }, + { + "epoch": 0.7810627247303236, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1955 + }, + { + "epoch": 0.7814622453056332, + "grad_norm": 0.25390625, + "learning_rate": 0.0002, + "loss": 1.3579, + "step": 1956 + }, + { + "epoch": 0.7818617658809429, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 1957 + }, + { + "epoch": 0.7822612864562525, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3573, + "step": 1958 + }, + { + "epoch": 0.7826608070315622, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 1959 + }, + { + "epoch": 0.7830603276068717, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 1960 + }, + { + "epoch": 0.7834598481821814, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 1961 + }, + { + "epoch": 0.783859368757491, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 1962 + }, + { + "epoch": 0.7842588893328006, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3651, + "step": 1963 + }, + { + "epoch": 0.7846584099081103, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 1964 + }, + { + "epoch": 0.7850579304834199, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 1965 + }, + { + "epoch": 0.7854574510587295, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 1966 + }, + { + "epoch": 0.7858569716340391, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 1967 + }, + { + "epoch": 0.7862564922093488, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 1968 + }, + { + "epoch": 0.7866560127846585, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 1969 + }, + { + "epoch": 0.787055533359968, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 1970 + }, + { + "epoch": 0.7874550539352777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 1971 + }, + { + "epoch": 0.7878545745105873, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 1972 + }, + { + "epoch": 0.7882540950858969, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 1973 + }, + { + "epoch": 0.7886536156612065, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 1974 + }, + { + "epoch": 0.7890531362365162, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1975 + }, + { + "epoch": 0.7894526568118259, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 1976 + }, + { + "epoch": 0.7898521773871354, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 1977 + }, + { + "epoch": 0.7902516979624451, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 1978 + }, + { + "epoch": 0.7906512185377547, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 1979 + }, + { + "epoch": 0.7910507391130643, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1980 + }, + { + "epoch": 0.7914502596883739, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 1981 + }, + { + "epoch": 0.7918497802636836, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 1982 + }, + { + "epoch": 0.7922493008389933, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 1983 + }, + { + "epoch": 0.7926488214143028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 1984 + }, + { + "epoch": 0.7930483419896125, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 1985 + }, + { + "epoch": 0.7934478625649221, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 1986 + }, + { + "epoch": 0.7938473831402317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 1987 + }, + { + "epoch": 0.7942469037155413, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 1988 + }, + { + "epoch": 0.794646424290851, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 1989 + }, + { + "epoch": 0.7950459448661606, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 1990 + }, + { + "epoch": 0.7954454654414702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 1991 + }, + { + "epoch": 0.7958449860167799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1992 + }, + { + "epoch": 0.7962445065920895, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 1993 + }, + { + "epoch": 0.7966440271673991, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 1994 + }, + { + "epoch": 0.7970435477427088, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 1995 + }, + { + "epoch": 0.7974430683180184, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 1996 + }, + { + "epoch": 0.797842588893328, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 1997 + }, + { + "epoch": 0.7982421094686376, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 1998 + }, + { + "epoch": 0.7986416300439473, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 1999 + }, + { + "epoch": 0.7990411506192568, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2000 + }, + { + "epoch": 0.7994406711945665, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2001 + }, + { + "epoch": 0.7998401917698762, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2002 + }, + { + "epoch": 0.8002397123451858, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2003 + }, + { + "epoch": 0.8006392329204954, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2004 + }, + { + "epoch": 0.801038753495805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3641, + "step": 2005 + }, + { + "epoch": 0.8014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2006 + }, + { + "epoch": 0.8018377946464242, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2007 + }, + { + "epoch": 0.8022373152217339, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2008 + }, + { + "epoch": 0.8026368357970436, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2009 + }, + { + "epoch": 0.8030363563723532, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3564, + "step": 2010 + }, + { + "epoch": 0.8034358769476628, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2011 + }, + { + "epoch": 0.8038353975229724, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2012 + }, + { + "epoch": 0.8042349180982821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2013 + }, + { + "epoch": 0.8046344386735917, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 2014 + }, + { + "epoch": 0.8050339592489013, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3646, + "step": 2015 + }, + { + "epoch": 0.805433479824211, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2016 + }, + { + "epoch": 0.8058330003995205, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2017 + }, + { + "epoch": 0.8062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2018 + }, + { + "epoch": 0.8066320415501398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2019 + }, + { + "epoch": 0.8070315621254495, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2020 + }, + { + "epoch": 0.8074310827007591, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2021 + }, + { + "epoch": 0.8078306032760687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2022 + }, + { + "epoch": 0.8082301238513784, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2023 + }, + { + "epoch": 0.8086296444266879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2024 + }, + { + "epoch": 0.8090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2025 + }, + { + "epoch": 0.8094286855773072, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 2026 + }, + { + "epoch": 0.8098282061526169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2027 + }, + { + "epoch": 0.8102277267279265, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2028 + }, + { + "epoch": 0.8106272473032361, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2029 + }, + { + "epoch": 0.8110267678785458, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2030 + }, + { + "epoch": 0.8114262884538553, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2031 + }, + { + "epoch": 0.811825809029165, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.36, + "step": 2032 + }, + { + "epoch": 0.8122253296044746, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2033 + }, + { + "epoch": 0.8126248501797843, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2034 + }, + { + "epoch": 0.8130243707550939, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 2035 + }, + { + "epoch": 0.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2036 + }, + { + "epoch": 0.8138234119057132, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 2037 + }, + { + "epoch": 0.8142229324810227, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2038 + }, + { + "epoch": 0.8146224530563324, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2039 + }, + { + "epoch": 0.8150219736316421, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2040 + }, + { + "epoch": 0.8154214942069516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2041 + }, + { + "epoch": 0.8158210147822613, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2042 + }, + { + "epoch": 0.8162205353575709, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2043 + }, + { + "epoch": 0.8166200559328806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2044 + }, + { + "epoch": 0.8170195765081901, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2045 + }, + { + "epoch": 0.8174190970834998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2046 + }, + { + "epoch": 0.8178186176588095, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2047 + }, + { + "epoch": 0.818218138234119, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2048 + }, + { + "epoch": 0.8186176588094287, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3529, + "step": 2049 + }, + { + "epoch": 0.8190171793847383, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2050 + }, + { + "epoch": 0.819416699960048, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2051 + }, + { + "epoch": 0.8198162205353575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2052 + }, + { + "epoch": 0.8202157411106672, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2053 + }, + { + "epoch": 0.8206152616859769, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2054 + }, + { + "epoch": 0.8210147822612864, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2055 + }, + { + "epoch": 0.8214143028365961, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2056 + }, + { + "epoch": 0.8218138234119057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2057 + }, + { + "epoch": 0.8222133439872154, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2058 + }, + { + "epoch": 0.8226128645625249, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2059 + }, + { + "epoch": 0.8230123851378346, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2060 + }, + { + "epoch": 0.8234119057131443, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2061 + }, + { + "epoch": 0.8238114262884538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2062 + }, + { + "epoch": 0.8242109468637635, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2063 + }, + { + "epoch": 0.8246104674390731, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2064 + }, + { + "epoch": 0.8250099880143827, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2065 + }, + { + "epoch": 0.8254095085896924, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2066 + }, + { + "epoch": 0.825809029165002, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2067 + }, + { + "epoch": 0.8262085497403117, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2068 + }, + { + "epoch": 0.8266080703156212, + "grad_norm": 0.2265625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2069 + }, + { + "epoch": 0.8270075908909309, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2070 + }, + { + "epoch": 0.8274071114662405, + "grad_norm": 0.306640625, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2071 + }, + { + "epoch": 0.8278066320415501, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2072 + }, + { + "epoch": 0.8282061526168598, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 2073 + }, + { + "epoch": 0.8286056731921694, + "grad_norm": 0.39453125, + "learning_rate": 0.0002, + "loss": 1.3637, + "step": 2074 + }, + { + "epoch": 0.829005193767479, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2075 + }, + { + "epoch": 0.8294047143427886, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2076 + }, + { + "epoch": 0.8298042349180983, + "grad_norm": 0.38671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2077 + }, + { + "epoch": 0.8302037554934079, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2078 + }, + { + "epoch": 0.8306032760687175, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2079 + }, + { + "epoch": 0.8310027966440272, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2080 + }, + { + "epoch": 0.8314023172193368, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2081 + }, + { + "epoch": 0.8318018377946464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2082 + }, + { + "epoch": 0.832201358369956, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2083 + }, + { + "epoch": 0.8326008789452657, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2084 + }, + { + "epoch": 0.8330003995205754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2085 + }, + { + "epoch": 0.8333999200958849, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2086 + }, + { + "epoch": 0.8337994406711946, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2087 + }, + { + "epoch": 0.8341989612465042, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2088 + }, + { + "epoch": 0.8345984818218138, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2089 + }, + { + "epoch": 0.8349980023971234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.355, + "step": 2090 + }, + { + "epoch": 0.8353975229724331, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3586, + "step": 2091 + }, + { + "epoch": 0.8357970435477428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2092 + }, + { + "epoch": 0.8361965641230523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2093 + }, + { + "epoch": 0.836596084698362, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2094 + }, + { + "epoch": 0.8369956052736716, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2095 + }, + { + "epoch": 0.8373951258489812, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2096 + }, + { + "epoch": 0.8377946464242908, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2097 + }, + { + "epoch": 0.8381941669996005, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2098 + }, + { + "epoch": 0.8385936875749102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2099 + }, + { + "epoch": 0.8389932081502197, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2100 + }, + { + "epoch": 0.8393927287255294, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3597, + "step": 2101 + }, + { + "epoch": 0.839792249300839, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2102 + }, + { + "epoch": 0.8401917698761486, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2103 + }, + { + "epoch": 0.8405912904514582, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 2104 + }, + { + "epoch": 0.8409908110267679, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2105 + }, + { + "epoch": 0.8413903316020775, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2106 + }, + { + "epoch": 0.8417898521773871, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2107 + }, + { + "epoch": 0.8421893727526968, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2108 + }, + { + "epoch": 0.8425888933280063, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2109 + }, + { + "epoch": 0.842988413903316, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2110 + }, + { + "epoch": 0.8433879344786257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2111 + }, + { + "epoch": 0.8437874550539353, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2112 + }, + { + "epoch": 0.8441869756292449, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2113 + }, + { + "epoch": 0.8445864962045545, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2114 + }, + { + "epoch": 0.8449860167798642, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2115 + }, + { + "epoch": 0.8453855373551737, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2116 + }, + { + "epoch": 0.8457850579304834, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3551, + "step": 2117 + }, + { + "epoch": 0.8461845785057931, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2118 + }, + { + "epoch": 0.8465840990811027, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2119 + }, + { + "epoch": 0.8469836196564123, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2120 + }, + { + "epoch": 0.8473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2121 + }, + { + "epoch": 0.8477826608070316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2122 + }, + { + "epoch": 0.8481821813823411, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 2123 + }, + { + "epoch": 0.8485817019576508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2124 + }, + { + "epoch": 0.8489812225329605, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 2125 + }, + { + "epoch": 0.84938074310827, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2126 + }, + { + "epoch": 0.8497802636835797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2127 + }, + { + "epoch": 0.8501797842588893, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2128 + }, + { + "epoch": 0.850579304834199, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2129 + }, + { + "epoch": 0.8509788254095086, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2130 + }, + { + "epoch": 0.8513783459848182, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2131 + }, + { + "epoch": 0.8517778665601279, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2132 + }, + { + "epoch": 0.8521773871354374, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2133 + }, + { + "epoch": 0.8525769077107471, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2134 + }, + { + "epoch": 0.8529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2135 + }, + { + "epoch": 0.8533759488613664, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2136 + }, + { + "epoch": 0.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2137 + }, + { + "epoch": 0.8541749900119856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2138 + }, + { + "epoch": 0.8545745105872953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2139 + }, + { + "epoch": 0.8549740311626048, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2140 + }, + { + "epoch": 0.8553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2141 + }, + { + "epoch": 0.8557730723132241, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2142 + }, + { + "epoch": 0.8561725928885338, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3576, + "step": 2143 + }, + { + "epoch": 0.8565721134638434, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2144 + }, + { + "epoch": 0.856971634039153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2145 + }, + { + "epoch": 0.8573711546144627, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2146 + }, + { + "epoch": 0.8577706751897722, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2147 + }, + { + "epoch": 0.8581701957650819, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3559, + "step": 2148 + }, + { + "epoch": 0.8585697163403915, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2149 + }, + { + "epoch": 0.8589692369157012, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2150 + }, + { + "epoch": 0.8593687574910108, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2151 + }, + { + "epoch": 0.8597682780663204, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2152 + }, + { + "epoch": 0.8601677986416301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3622, + "step": 2153 + }, + { + "epoch": 0.8605673192169396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3602, + "step": 2154 + }, + { + "epoch": 0.8609668397922493, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2155 + }, + { + "epoch": 0.861366360367559, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2156 + }, + { + "epoch": 0.8617658809428685, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2157 + }, + { + "epoch": 0.8621654015181782, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2158 + }, + { + "epoch": 0.8625649220934878, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2159 + }, + { + "epoch": 0.8629644426687975, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2160 + }, + { + "epoch": 0.863363963244107, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2161 + }, + { + "epoch": 0.8637634838194167, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2162 + }, + { + "epoch": 0.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2163 + }, + { + "epoch": 0.8645625249700359, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2164 + }, + { + "epoch": 0.8649620455453456, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2165 + }, + { + "epoch": 0.8653615661206552, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2166 + }, + { + "epoch": 0.8657610866959649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2167 + }, + { + "epoch": 0.8661606072712744, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2168 + }, + { + "epoch": 0.8665601278465841, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2169 + }, + { + "epoch": 0.8669596484218938, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2170 + }, + { + "epoch": 0.8673591689972033, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2171 + }, + { + "epoch": 0.867758689572513, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2172 + }, + { + "epoch": 0.8681582101478226, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2173 + }, + { + "epoch": 0.8685577307231322, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2174 + }, + { + "epoch": 0.8689572512984419, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2175 + }, + { + "epoch": 0.8693567718737515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2176 + }, + { + "epoch": 0.8697562924490612, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2177 + }, + { + "epoch": 0.8701558130243707, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2178 + }, + { + "epoch": 0.8705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 2179 + }, + { + "epoch": 0.87095485417499, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2180 + }, + { + "epoch": 0.8713543747502996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2181 + }, + { + "epoch": 0.8717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2182 + }, + { + "epoch": 0.8721534159009189, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2183 + }, + { + "epoch": 0.8725529364762286, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2184 + }, + { + "epoch": 0.8729524570515381, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2185 + }, + { + "epoch": 0.8733519776268478, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2186 + }, + { + "epoch": 0.8737514982021574, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2187 + }, + { + "epoch": 0.874151018777467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2188 + }, + { + "epoch": 0.8745505393527767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2189 + }, + { + "epoch": 0.8749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2190 + }, + { + "epoch": 0.875349580503396, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2191 + }, + { + "epoch": 0.8757491010787055, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2192 + }, + { + "epoch": 0.8761486216540152, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2193 + }, + { + "epoch": 0.8765481422293248, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2194 + }, + { + "epoch": 0.8769476628046344, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2195 + }, + { + "epoch": 0.8773471833799441, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2196 + }, + { + "epoch": 0.8777467039552537, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2197 + }, + { + "epoch": 0.8781462245305633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3618, + "step": 2198 + }, + { + "epoch": 0.8785457451058729, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3532, + "step": 2199 + }, + { + "epoch": 0.8789452656811826, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2200 + }, + { + "epoch": 0.8793447862564923, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2201 + }, + { + "epoch": 0.8797443068318018, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3533, + "step": 2202 + }, + { + "epoch": 0.8801438274071115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2203 + }, + { + "epoch": 0.8805433479824211, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2204 + }, + { + "epoch": 0.8809428685577307, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2205 + }, + { + "epoch": 0.8813423891330403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 2206 + }, + { + "epoch": 0.88174190970835, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2207 + }, + { + "epoch": 0.8821414302836597, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2208 + }, + { + "epoch": 0.8825409508589692, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2209 + }, + { + "epoch": 0.8829404714342789, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 2210 + }, + { + "epoch": 0.8833399920095885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2211 + }, + { + "epoch": 0.8837395125848981, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3574, + "step": 2212 + }, + { + "epoch": 0.8841390331602077, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2213 + }, + { + "epoch": 0.8845385537355174, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2214 + }, + { + "epoch": 0.884938074310827, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2215 + }, + { + "epoch": 0.8853375948861366, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2216 + }, + { + "epoch": 0.8857371154614463, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2217 + }, + { + "epoch": 0.8861366360367559, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2218 + }, + { + "epoch": 0.8865361566120655, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2219 + }, + { + "epoch": 0.8869356771873752, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2220 + }, + { + "epoch": 0.8873351977626848, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2221 + }, + { + "epoch": 0.8877347183379944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2222 + }, + { + "epoch": 0.888134238913304, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2223 + }, + { + "epoch": 0.8885337594886137, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2224 + }, + { + "epoch": 0.8889332800639232, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2225 + }, + { + "epoch": 0.8893328006392329, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2226 + }, + { + "epoch": 0.8897323212145426, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2227 + }, + { + "epoch": 0.8901318417898522, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 2228 + }, + { + "epoch": 0.8905313623651618, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2229 + }, + { + "epoch": 0.8909308829404714, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2230 + }, + { + "epoch": 0.8913304035157811, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2231 + }, + { + "epoch": 0.8917299240910906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2232 + }, + { + "epoch": 0.8921294446664003, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2233 + }, + { + "epoch": 0.89252896524171, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2234 + }, + { + "epoch": 0.8929284858170196, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2235 + }, + { + "epoch": 0.8933280063923292, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2236 + }, + { + "epoch": 0.8937275269676388, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2237 + }, + { + "epoch": 0.8941270475429485, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2238 + }, + { + "epoch": 0.894526568118258, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2239 + }, + { + "epoch": 0.8949260886935677, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2240 + }, + { + "epoch": 0.8953256092688774, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2241 + }, + { + "epoch": 0.895725129844187, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2242 + }, + { + "epoch": 0.8961246504194966, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2243 + }, + { + "epoch": 0.8965241709948062, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2244 + }, + { + "epoch": 0.8969236915701159, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2245 + }, + { + "epoch": 0.8973232121454255, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2246 + }, + { + "epoch": 0.8977227327207351, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2247 + }, + { + "epoch": 0.8981222532960448, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2248 + }, + { + "epoch": 0.8985217738713543, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2249 + }, + { + "epoch": 0.898921294446664, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2250 + }, + { + "epoch": 0.8993208150219736, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2251 + }, + { + "epoch": 0.8997203355972833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2252 + }, + { + "epoch": 0.9001198561725929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2253 + }, + { + "epoch": 0.9005193767479025, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2254 + }, + { + "epoch": 0.9009188973232122, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2255 + }, + { + "epoch": 0.9013184178985217, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2256 + }, + { + "epoch": 0.9017179384738314, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2257 + }, + { + "epoch": 0.902117459049141, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3595, + "step": 2258 + }, + { + "epoch": 0.9025169796244507, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2259 + }, + { + "epoch": 0.9029165001997603, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2260 + }, + { + "epoch": 0.9033160207750699, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2261 + }, + { + "epoch": 0.9037155413503796, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2262 + }, + { + "epoch": 0.9041150619256891, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2263 + }, + { + "epoch": 0.9045145825009988, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2264 + }, + { + "epoch": 0.9049141030763085, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2265 + }, + { + "epoch": 0.905313623651618, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2266 + }, + { + "epoch": 0.9057131442269277, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2267 + }, + { + "epoch": 0.9061126648022373, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2268 + }, + { + "epoch": 0.906512185377547, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3666, + "step": 2269 + }, + { + "epoch": 0.9069117059528565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2270 + }, + { + "epoch": 0.9073112265281662, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2271 + }, + { + "epoch": 0.9077107471034759, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2272 + }, + { + "epoch": 0.9081102676787854, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2273 + }, + { + "epoch": 0.9085097882540951, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2274 + }, + { + "epoch": 0.9089093088294047, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2275 + }, + { + "epoch": 0.9093088294047144, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2276 + }, + { + "epoch": 0.9097083499800239, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2277 + }, + { + "epoch": 0.9101078705553336, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2278 + }, + { + "epoch": 0.9105073911306433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2279 + }, + { + "epoch": 0.9109069117059528, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2280 + }, + { + "epoch": 0.9113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 2281 + }, + { + "epoch": 0.9117059528565721, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2282 + }, + { + "epoch": 0.9121054734318818, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 2283 + }, + { + "epoch": 0.9125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2284 + }, + { + "epoch": 0.912904514582501, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2285 + }, + { + "epoch": 0.9133040351578107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2286 + }, + { + "epoch": 0.9137035557331202, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2287 + }, + { + "epoch": 0.9141030763084299, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2288 + }, + { + "epoch": 0.9145025968837395, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2289 + }, + { + "epoch": 0.9149021174590491, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2290 + }, + { + "epoch": 0.9153016380343588, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 2291 + }, + { + "epoch": 0.9157011586096684, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2292 + }, + { + "epoch": 0.9161006791849781, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2293 + }, + { + "epoch": 0.9165001997602876, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2294 + }, + { + "epoch": 0.9168997203355973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2295 + }, + { + "epoch": 0.9172992409109069, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2296 + }, + { + "epoch": 0.9176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2297 + }, + { + "epoch": 0.9180982820615262, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2298 + }, + { + "epoch": 0.9184978026368358, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2299 + }, + { + "epoch": 0.9188973232121455, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2300 + }, + { + "epoch": 0.919296843787455, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2301 + }, + { + "epoch": 0.9196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2302 + }, + { + "epoch": 0.9200958849380743, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3566, + "step": 2303 + }, + { + "epoch": 0.9204954055133839, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2304 + }, + { + "epoch": 0.9208949260886936, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2305 + }, + { + "epoch": 0.9212944466640032, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2306 + }, + { + "epoch": 0.9216939672393129, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2307 + }, + { + "epoch": 0.9220934878146224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2308 + }, + { + "epoch": 0.9224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2309 + }, + { + "epoch": 0.9228925289652418, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2310 + }, + { + "epoch": 0.9232920495405513, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2311 + }, + { + "epoch": 0.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2312 + }, + { + "epoch": 0.9240910906911706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2313 + }, + { + "epoch": 0.9244906112664802, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3534, + "step": 2314 + }, + { + "epoch": 0.9248901318417898, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2315 + }, + { + "epoch": 0.9252896524170995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3653, + "step": 2316 + }, + { + "epoch": 0.9256891729924092, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2317 + }, + { + "epoch": 0.9260886935677187, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2318 + }, + { + "epoch": 0.9264882141430284, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2319 + }, + { + "epoch": 0.926887734718338, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2320 + }, + { + "epoch": 0.9272872552936476, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2321 + }, + { + "epoch": 0.9276867758689572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2322 + }, + { + "epoch": 0.9280862964442669, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2323 + }, + { + "epoch": 0.9284858170195766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2324 + }, + { + "epoch": 0.9288853375948861, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 2325 + }, + { + "epoch": 0.9292848581701958, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2326 + }, + { + "epoch": 0.9296843787455054, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2327 + }, + { + "epoch": 0.930083899320815, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2328 + }, + { + "epoch": 0.9304834198961246, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3547, + "step": 2329 + }, + { + "epoch": 0.9308829404714343, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2330 + }, + { + "epoch": 0.931282461046744, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2331 + }, + { + "epoch": 0.9316819816220535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3531, + "step": 2332 + }, + { + "epoch": 0.9320815021973632, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2333 + }, + { + "epoch": 0.9324810227726728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2334 + }, + { + "epoch": 0.9328805433479824, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2335 + }, + { + "epoch": 0.9332800639232921, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2336 + }, + { + "epoch": 0.9336795844986017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2337 + }, + { + "epoch": 0.9340791050739113, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2338 + }, + { + "epoch": 0.9344786256492209, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2339 + }, + { + "epoch": 0.9348781462245306, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2340 + }, + { + "epoch": 0.9352776667998401, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2341 + }, + { + "epoch": 0.9356771873751498, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2342 + }, + { + "epoch": 0.9360767079504595, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2343 + }, + { + "epoch": 0.9364762285257691, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 2344 + }, + { + "epoch": 0.9368757491010787, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2345 + }, + { + "epoch": 0.9372752696763883, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2346 + }, + { + "epoch": 0.937674790251698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2347 + }, + { + "epoch": 0.9380743108270075, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2348 + }, + { + "epoch": 0.9384738314023172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2349 + }, + { + "epoch": 0.9388733519776269, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2350 + }, + { + "epoch": 0.9392728725529365, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2351 + }, + { + "epoch": 0.9396723931282461, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3535, + "step": 2352 + }, + { + "epoch": 0.9400719137035557, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2353 + }, + { + "epoch": 0.9404714342788654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2354 + }, + { + "epoch": 0.9408709548541749, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2355 + }, + { + "epoch": 0.9412704754294846, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2356 + }, + { + "epoch": 0.9416699960047943, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2357 + }, + { + "epoch": 0.9420695165801038, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2358 + }, + { + "epoch": 0.9424690371554135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3558, + "step": 2359 + }, + { + "epoch": 0.9428685577307231, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2360 + }, + { + "epoch": 0.9432680783060328, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 2361 + }, + { + "epoch": 0.9436675988813424, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3583, + "step": 2362 + }, + { + "epoch": 0.944067119456652, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 2363 + }, + { + "epoch": 0.9444666400319617, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2364 + }, + { + "epoch": 0.9448661606072712, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3672, + "step": 2365 + }, + { + "epoch": 0.9452656811825809, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2366 + }, + { + "epoch": 0.9456652017578905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2367 + }, + { + "epoch": 0.9460647223332002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2368 + }, + { + "epoch": 0.9464642429085098, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2369 + }, + { + "epoch": 0.9468637634838194, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2370 + }, + { + "epoch": 0.9472632840591291, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2371 + }, + { + "epoch": 0.9476628046344386, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2372 + }, + { + "epoch": 0.9480623252097483, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2373 + }, + { + "epoch": 0.9484618457850579, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2374 + }, + { + "epoch": 0.9488613663603676, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2375 + }, + { + "epoch": 0.9492608869356772, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2376 + }, + { + "epoch": 0.9496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2377 + }, + { + "epoch": 0.9500599280862965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2378 + }, + { + "epoch": 0.950459448661606, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2379 + }, + { + "epoch": 0.9508589692369157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2380 + }, + { + "epoch": 0.9512584898122254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2381 + }, + { + "epoch": 0.951658010387535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 2382 + }, + { + "epoch": 0.9520575309628446, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2383 + }, + { + "epoch": 0.9524570515381542, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2384 + }, + { + "epoch": 0.9528565721134639, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 2385 + }, + { + "epoch": 0.9532560926887734, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2386 + }, + { + "epoch": 0.9536556132640831, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2387 + }, + { + "epoch": 0.9540551338393928, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2388 + }, + { + "epoch": 0.9544546544147023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2389 + }, + { + "epoch": 0.954854174990012, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3568, + "step": 2390 + }, + { + "epoch": 0.9552536955653216, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2391 + }, + { + "epoch": 0.9556532161406313, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 2392 + }, + { + "epoch": 0.9560527367159408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2393 + }, + { + "epoch": 0.9564522572912505, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2394 + }, + { + "epoch": 0.9568517778665602, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2395 + }, + { + "epoch": 0.9572512984418697, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2396 + }, + { + "epoch": 0.9576508190171794, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2397 + }, + { + "epoch": 0.958050339592489, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3536, + "step": 2398 + }, + { + "epoch": 0.9584498601677987, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 2399 + }, + { + "epoch": 0.9588493807431082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2400 + }, + { + "epoch": 0.9592489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2401 + }, + { + "epoch": 0.9596484218937276, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 2402 + }, + { + "epoch": 0.9600479424690371, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2403 + }, + { + "epoch": 0.9604474630443468, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2404 + }, + { + "epoch": 0.9608469836196564, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3514, + "step": 2405 + }, + { + "epoch": 0.961246504194966, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2406 + }, + { + "epoch": 0.9616460247702757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2407 + }, + { + "epoch": 0.9620455453455853, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 2408 + }, + { + "epoch": 0.962445065920895, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2409 + }, + { + "epoch": 0.9628445864962045, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2410 + }, + { + "epoch": 0.9632441070715142, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2411 + }, + { + "epoch": 0.9636436276468238, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2412 + }, + { + "epoch": 0.9640431482221334, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2413 + }, + { + "epoch": 0.9644426687974431, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2414 + }, + { + "epoch": 0.9648421893727527, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2415 + }, + { + "epoch": 0.9652417099480624, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 2416 + }, + { + "epoch": 0.9656412305233719, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2417 + }, + { + "epoch": 0.9660407510986816, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2418 + }, + { + "epoch": 0.9664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2419 + }, + { + "epoch": 0.9668397922493008, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2420 + }, + { + "epoch": 0.9672393128246105, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3505, + "step": 2421 + }, + { + "epoch": 0.9676388333999201, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2422 + }, + { + "epoch": 0.9680383539752297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2423 + }, + { + "epoch": 0.9684378745505393, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2424 + }, + { + "epoch": 0.968837395125849, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3652, + "step": 2425 + }, + { + "epoch": 0.9692369157011587, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2426 + }, + { + "epoch": 0.9696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3553, + "step": 2427 + }, + { + "epoch": 0.9700359568517779, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2428 + }, + { + "epoch": 0.9704354774270875, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2429 + }, + { + "epoch": 0.9708349980023971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3523, + "step": 2430 + }, + { + "epoch": 0.9712345185777067, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2431 + }, + { + "epoch": 0.9716340391530164, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2432 + }, + { + "epoch": 0.9720335597283261, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2433 + }, + { + "epoch": 0.9724330803036356, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2434 + }, + { + "epoch": 0.9728326008789453, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2435 + }, + { + "epoch": 0.9732321214542549, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2436 + }, + { + "epoch": 0.9736316420295645, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2437 + }, + { + "epoch": 0.9740311626048741, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2438 + }, + { + "epoch": 0.9744306831801838, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2439 + }, + { + "epoch": 0.9748302037554935, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2440 + }, + { + "epoch": 0.975229724330803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2441 + }, + { + "epoch": 0.9756292449061127, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2442 + }, + { + "epoch": 0.9760287654814223, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 2443 + }, + { + "epoch": 0.9764282860567319, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 2444 + }, + { + "epoch": 0.9768278066320415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2445 + }, + { + "epoch": 0.9772273272073512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2446 + }, + { + "epoch": 0.9776268477826608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2447 + }, + { + "epoch": 0.9780263683579704, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2448 + }, + { + "epoch": 0.9784258889332801, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2449 + }, + { + "epoch": 0.9788254095085897, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2450 + }, + { + "epoch": 0.9792249300838993, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2451 + }, + { + "epoch": 0.979624450659209, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2452 + }, + { + "epoch": 0.9800239712345186, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2453 + }, + { + "epoch": 0.9804234918098282, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2454 + }, + { + "epoch": 0.9808230123851378, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2455 + }, + { + "epoch": 0.9812225329604475, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2456 + }, + { + "epoch": 0.981622053535757, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2457 + }, + { + "epoch": 0.9820215741110667, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2458 + }, + { + "epoch": 0.9824210946863764, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2459 + }, + { + "epoch": 0.982820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2460 + }, + { + "epoch": 0.9832201358369956, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2461 + }, + { + "epoch": 0.9836196564123052, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2462 + }, + { + "epoch": 0.9840191769876149, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2463 + }, + { + "epoch": 0.9844186975629244, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2464 + }, + { + "epoch": 0.9848182181382341, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2465 + }, + { + "epoch": 0.9852177387135438, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2466 + }, + { + "epoch": 0.9856172592888534, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2467 + }, + { + "epoch": 0.986016779864163, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 2468 + }, + { + "epoch": 0.9864163004394726, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2469 + }, + { + "epoch": 0.9868158210147823, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2470 + }, + { + "epoch": 0.9872153415900919, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2471 + }, + { + "epoch": 0.9876148621654015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2472 + }, + { + "epoch": 0.9880143827407112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2473 + }, + { + "epoch": 0.9884139033160207, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 2474 + }, + { + "epoch": 0.9888134238913304, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 2475 + }, + { + "epoch": 0.98921294446664, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 2476 + }, + { + "epoch": 0.9896124650419497, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2477 + }, + { + "epoch": 0.9900119856172593, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2478 + }, + { + "epoch": 0.9904115061925689, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2479 + }, + { + "epoch": 0.9908110267678786, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2480 + }, + { + "epoch": 0.9912105473431881, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2481 + }, + { + "epoch": 0.9916100679184978, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2482 + }, + { + "epoch": 0.9920095884938074, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2483 + }, + { + "epoch": 0.9924091090691171, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2484 + }, + { + "epoch": 0.9928086296444267, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2485 + }, + { + "epoch": 0.9932081502197363, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2486 + }, + { + "epoch": 0.993607670795046, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2487 + }, + { + "epoch": 0.9940071913703555, + "grad_norm": 0.337890625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 2488 + }, + { + "epoch": 0.9944067119456652, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2489 + }, + { + "epoch": 0.9948062325209748, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2490 + }, + { + "epoch": 0.9952057530962845, + "grad_norm": 0.37109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2491 + }, + { + "epoch": 0.9956052736715941, + "grad_norm": 0.2236328125, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2492 + }, + { + "epoch": 0.9960047942469037, + "grad_norm": 0.2451171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2493 + }, + { + "epoch": 0.9964043148222134, + "grad_norm": 0.6328125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 2494 + }, + { + "epoch": 0.9968038353975229, + "grad_norm": 0.671875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2495 + }, + { + "epoch": 0.9972033559728326, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2496 + }, + { + "epoch": 0.9976028765481423, + "grad_norm": 0.419921875, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2497 + }, + { + "epoch": 0.9980023971234518, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2498 + }, + { + "epoch": 0.9984019176987615, + "grad_norm": 0.2470703125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2499 + }, + { + "epoch": 0.9988014382740711, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2500 + }, + { + "epoch": 0.9992009588493808, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 2501 + }, + { + "epoch": 0.9996004794246903, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2502 + }, + { + "epoch": 1.0, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2503 + }, + { + "epoch": 1.0003995205753096, + "grad_norm": 0.189453125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2504 + }, + { + "epoch": 1.0007990411506194, + "grad_norm": 0.283203125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2505 + }, + { + "epoch": 1.001198561725929, + "grad_norm": 0.2177734375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2506 + }, + { + "epoch": 1.0015980823012385, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2507 + }, + { + "epoch": 1.001997602876548, + "grad_norm": 0.228515625, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2508 + }, + { + "epoch": 1.0023971234518578, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2509 + }, + { + "epoch": 1.0027966440271674, + "grad_norm": 0.2216796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2510 + }, + { + "epoch": 1.003196164602477, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2511 + }, + { + "epoch": 1.0035956851777867, + "grad_norm": 0.1787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 2512 + }, + { + "epoch": 1.0039952057530963, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2513 + }, + { + "epoch": 1.0043947263284059, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2514 + }, + { + "epoch": 1.0047942469037154, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2515 + }, + { + "epoch": 1.0051937674790252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2516 + }, + { + "epoch": 1.0055932880543348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2517 + }, + { + "epoch": 1.0059928086296444, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2518 + }, + { + "epoch": 1.0063923292049541, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2519 + }, + { + "epoch": 1.0067918497802637, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2520 + }, + { + "epoch": 1.0071913703555733, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2521 + }, + { + "epoch": 1.0075908909308828, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2522 + }, + { + "epoch": 1.0079904115061926, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2523 + }, + { + "epoch": 1.0083899320815022, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2524 + }, + { + "epoch": 1.0087894526568117, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 2525 + }, + { + "epoch": 1.0091889732321215, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 2526 + }, + { + "epoch": 1.009588493807431, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2527 + }, + { + "epoch": 1.0099880143827407, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2528 + }, + { + "epoch": 1.0103875349580504, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2529 + }, + { + "epoch": 1.01078705553336, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2530 + }, + { + "epoch": 1.0111865761086696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2531 + }, + { + "epoch": 1.0115860966839791, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2532 + }, + { + "epoch": 1.011985617259289, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2533 + }, + { + "epoch": 1.0123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3567, + "step": 2534 + }, + { + "epoch": 1.012784658409908, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2535 + }, + { + "epoch": 1.0131841789852178, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3542, + "step": 2536 + }, + { + "epoch": 1.0135836995605274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2537 + }, + { + "epoch": 1.013983220135837, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2538 + }, + { + "epoch": 1.0143827407111465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2539 + }, + { + "epoch": 1.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2540 + }, + { + "epoch": 1.0151817818617659, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 2541 + }, + { + "epoch": 1.0155813024370755, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2542 + }, + { + "epoch": 1.0159808230123852, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2543 + }, + { + "epoch": 1.0163803435876948, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 2544 + }, + { + "epoch": 1.0167798641630044, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2545 + }, + { + "epoch": 1.017179384738314, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2546 + }, + { + "epoch": 1.0175789053136237, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2547 + }, + { + "epoch": 1.0179784258889333, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2548 + }, + { + "epoch": 1.0183779464642428, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2549 + }, + { + "epoch": 1.0187774670395526, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2550 + }, + { + "epoch": 1.0191769876148622, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2551 + }, + { + "epoch": 1.0195765081901718, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 2552 + }, + { + "epoch": 1.0199760287654813, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2553 + }, + { + "epoch": 1.020375549340791, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2554 + }, + { + "epoch": 1.0207750699161007, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 2555 + }, + { + "epoch": 1.0211745904914102, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2556 + }, + { + "epoch": 1.02157411106672, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 2557 + }, + { + "epoch": 1.0219736316420296, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2558 + }, + { + "epoch": 1.0223731522173392, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 2559 + }, + { + "epoch": 1.0227726727926487, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2560 + }, + { + "epoch": 1.0231721933679585, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2561 + }, + { + "epoch": 1.023571713943268, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 2562 + }, + { + "epoch": 1.0239712345185776, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2563 + }, + { + "epoch": 1.0243707550938874, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 2564 + }, + { + "epoch": 1.024770275669197, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2565 + }, + { + "epoch": 1.0251697962445065, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2566 + }, + { + "epoch": 1.0255693168198161, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 2567 + }, + { + "epoch": 1.025968837395126, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2568 + }, + { + "epoch": 1.0263683579704355, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2569 + }, + { + "epoch": 1.026767878545745, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2570 + }, + { + "epoch": 1.0271673991210548, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2571 + }, + { + "epoch": 1.0275669196963644, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 2572 + }, + { + "epoch": 1.027966440271674, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3575, + "step": 2573 + }, + { + "epoch": 1.0283659608469837, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2574 + }, + { + "epoch": 1.0287654814222933, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2575 + }, + { + "epoch": 1.0291650019976029, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 2576 + }, + { + "epoch": 1.0295645225729124, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3588, + "step": 2577 + }, + { + "epoch": 1.0299640431482222, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2578 + }, + { + "epoch": 1.0303635637235318, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 2579 + }, + { + "epoch": 1.0307630842988413, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 2580 + }, + { + "epoch": 1.0311626048741511, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2581 + }, + { + "epoch": 1.0315621254494607, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2582 + }, + { + "epoch": 1.0319616460247703, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2583 + }, + { + "epoch": 1.0323611666000798, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2584 + }, + { + "epoch": 1.0327606871753896, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2585 + }, + { + "epoch": 1.0331602077506992, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2586 + }, + { + "epoch": 1.0335597283260087, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 2587 + }, + { + "epoch": 1.0339592489013185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2588 + }, + { + "epoch": 1.034358769476628, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2589 + }, + { + "epoch": 1.0347582900519376, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2590 + }, + { + "epoch": 1.0351578106272472, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2591 + }, + { + "epoch": 1.035557331202557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 2592 + }, + { + "epoch": 1.0359568517778666, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2593 + }, + { + "epoch": 1.0363563723531761, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2594 + }, + { + "epoch": 1.036755892928486, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2595 + }, + { + "epoch": 1.0371554135037955, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2596 + }, + { + "epoch": 1.037554934079105, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 2597 + }, + { + "epoch": 1.0379544546544146, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2598 + }, + { + "epoch": 1.0383539752297244, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2599 + }, + { + "epoch": 1.038753495805034, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2600 + }, + { + "epoch": 1.0391530163803435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2601 + }, + { + "epoch": 1.0395525369556533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2602 + }, + { + "epoch": 1.0399520575309629, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2603 + }, + { + "epoch": 1.0403515781062724, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 2604 + }, + { + "epoch": 1.040751098681582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 2605 + }, + { + "epoch": 1.0411506192568918, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 2606 + }, + { + "epoch": 1.0415501398322013, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 2607 + }, + { + "epoch": 1.041949660407511, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2608 + }, + { + "epoch": 1.0423491809828207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 2609 + }, + { + "epoch": 1.0427487015581303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2610 + }, + { + "epoch": 1.0431482221334398, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2611 + }, + { + "epoch": 1.0435477427087494, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2612 + }, + { + "epoch": 1.0439472632840592, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2613 + }, + { + "epoch": 1.0443467838593687, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2614 + }, + { + "epoch": 1.0447463044346783, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 2615 + }, + { + "epoch": 1.045145825009988, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2616 + }, + { + "epoch": 1.0455453455852977, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2617 + }, + { + "epoch": 1.0459448661606072, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2618 + }, + { + "epoch": 1.046344386735917, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2619 + }, + { + "epoch": 1.0467439073112266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2620 + }, + { + "epoch": 1.0471434278865361, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2621 + }, + { + "epoch": 1.0475429484618457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2622 + }, + { + "epoch": 1.0479424690371555, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2623 + }, + { + "epoch": 1.048341989612465, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2624 + }, + { + "epoch": 1.0487415101877746, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2625 + }, + { + "epoch": 1.0491410307630844, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 2626 + }, + { + "epoch": 1.049540551338394, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 2627 + }, + { + "epoch": 1.0499400719137035, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2628 + }, + { + "epoch": 1.050339592489013, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 2629 + }, + { + "epoch": 1.0507391130643229, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2630 + }, + { + "epoch": 1.0511386336396324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2631 + }, + { + "epoch": 1.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2632 + }, + { + "epoch": 1.0519376747902518, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2633 + }, + { + "epoch": 1.0523371953655614, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2634 + }, + { + "epoch": 1.052736715940871, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2635 + }, + { + "epoch": 1.0531362365161805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2636 + }, + { + "epoch": 1.0535357570914903, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2637 + }, + { + "epoch": 1.0539352776667998, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2638 + }, + { + "epoch": 1.0543347982421094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2639 + }, + { + "epoch": 1.0547343188174192, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 2640 + }, + { + "epoch": 1.0551338393927288, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2641 + }, + { + "epoch": 1.0555333599680383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2642 + }, + { + "epoch": 1.0559328805433479, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2643 + }, + { + "epoch": 1.0563324011186577, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2644 + }, + { + "epoch": 1.0567319216939672, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2645 + }, + { + "epoch": 1.0571314422692768, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2646 + }, + { + "epoch": 1.0575309628445866, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2647 + }, + { + "epoch": 1.0579304834198962, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2648 + }, + { + "epoch": 1.0583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2649 + }, + { + "epoch": 1.0587295245705153, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2650 + }, + { + "epoch": 1.059129045145825, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 2651 + }, + { + "epoch": 1.0595285657211346, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2652 + }, + { + "epoch": 1.0599280862964442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2653 + }, + { + "epoch": 1.060327606871754, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2654 + }, + { + "epoch": 1.0607271274470635, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2655 + }, + { + "epoch": 1.061126648022373, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2656 + }, + { + "epoch": 1.0615261685976827, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2657 + }, + { + "epoch": 1.0619256891729925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2658 + }, + { + "epoch": 1.062325209748302, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2659 + }, + { + "epoch": 1.0627247303236116, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 2660 + }, + { + "epoch": 1.0631242508989214, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2661 + }, + { + "epoch": 1.063523771474231, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2662 + }, + { + "epoch": 1.0639232920495405, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2663 + }, + { + "epoch": 1.0643228126248503, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2664 + }, + { + "epoch": 1.0647223332001599, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2665 + }, + { + "epoch": 1.0651218537754694, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2666 + }, + { + "epoch": 1.065521374350779, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2667 + }, + { + "epoch": 1.0659208949260888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2668 + }, + { + "epoch": 1.0663204155013983, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 2669 + }, + { + "epoch": 1.066719936076708, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2670 + }, + { + "epoch": 1.0671194566520175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2671 + }, + { + "epoch": 1.0675189772273272, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 2672 + }, + { + "epoch": 1.0679184978026368, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 2673 + }, + { + "epoch": 1.0683180183779464, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2674 + }, + { + "epoch": 1.0687175389532562, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 2675 + }, + { + "epoch": 1.0691170595285657, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2676 + }, + { + "epoch": 1.0695165801038753, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2677 + }, + { + "epoch": 1.069916100679185, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 2678 + }, + { + "epoch": 1.0703156212544946, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2679 + }, + { + "epoch": 1.0707151418298042, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2680 + }, + { + "epoch": 1.0711146624051138, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2681 + }, + { + "epoch": 1.0715141829804236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2682 + }, + { + "epoch": 1.0719137035557331, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3486, + "step": 2683 + }, + { + "epoch": 1.0723132241310427, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 2684 + }, + { + "epoch": 1.0727127447063525, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2685 + }, + { + "epoch": 1.073112265281662, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2686 + }, + { + "epoch": 1.0735117858569716, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3507, + "step": 2687 + }, + { + "epoch": 1.0739113064322812, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2688 + }, + { + "epoch": 1.074310827007591, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2689 + }, + { + "epoch": 1.0747103475829005, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 2690 + }, + { + "epoch": 1.07510986815821, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2691 + }, + { + "epoch": 1.0755093887335199, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2692 + }, + { + "epoch": 1.0759089093088294, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2693 + }, + { + "epoch": 1.076308429884139, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 2694 + }, + { + "epoch": 1.0767079504594486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2695 + }, + { + "epoch": 1.0771074710347583, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2696 + }, + { + "epoch": 1.077506991610068, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2697 + }, + { + "epoch": 1.0779065121853775, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2698 + }, + { + "epoch": 1.0783060327606873, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 2699 + }, + { + "epoch": 1.0787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2700 + }, + { + "epoch": 1.0791050739113064, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2701 + }, + { + "epoch": 1.079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 2702 + }, + { + "epoch": 1.0799041150619257, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2703 + }, + { + "epoch": 1.0803036356372353, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 2704 + }, + { + "epoch": 1.0807031562125449, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.351, + "step": 2705 + }, + { + "epoch": 1.0811026767878547, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2706 + }, + { + "epoch": 1.0815021973631642, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2707 + }, + { + "epoch": 1.0819017179384738, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2708 + }, + { + "epoch": 1.0823012385137836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 2709 + }, + { + "epoch": 1.0827007590890931, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2710 + }, + { + "epoch": 1.0831002796644027, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 2711 + }, + { + "epoch": 1.0834998002397123, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2712 + }, + { + "epoch": 1.083899320815022, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2713 + }, + { + "epoch": 1.0842988413903316, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2714 + }, + { + "epoch": 1.0846983619656412, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2715 + }, + { + "epoch": 1.0850978825409507, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2716 + }, + { + "epoch": 1.0854974031162605, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 2717 + }, + { + "epoch": 1.08589692369157, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2718 + }, + { + "epoch": 1.0862964442668797, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2719 + }, + { + "epoch": 1.0866959648421894, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2720 + }, + { + "epoch": 1.087095485417499, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2721 + }, + { + "epoch": 1.0874950059928086, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2722 + }, + { + "epoch": 1.0878945265681184, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2723 + }, + { + "epoch": 1.088294047143428, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2724 + }, + { + "epoch": 1.0886935677187375, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 2725 + }, + { + "epoch": 1.089093088294047, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2726 + }, + { + "epoch": 1.0894926088693568, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 2727 + }, + { + "epoch": 1.0898921294446664, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2728 + }, + { + "epoch": 1.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 2729 + }, + { + "epoch": 1.0906911705952858, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2730 + }, + { + "epoch": 1.0910906911705953, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2731 + }, + { + "epoch": 1.0914902117459049, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2732 + }, + { + "epoch": 1.0918897323212144, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2733 + }, + { + "epoch": 1.0922892528965242, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2734 + }, + { + "epoch": 1.0926887734718338, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 2735 + }, + { + "epoch": 1.0930882940471434, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 2736 + }, + { + "epoch": 1.0934878146224531, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2737 + }, + { + "epoch": 1.0938873351977627, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2738 + }, + { + "epoch": 1.0942868557730723, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2739 + }, + { + "epoch": 1.0946863763483818, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2740 + }, + { + "epoch": 1.0950858969236916, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 2741 + }, + { + "epoch": 1.0954854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2742 + }, + { + "epoch": 1.0958849380743108, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.358, + "step": 2743 + }, + { + "epoch": 1.0962844586496205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 2744 + }, + { + "epoch": 1.09668397922493, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2745 + }, + { + "epoch": 1.0970834998002397, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2746 + }, + { + "epoch": 1.0974830203755492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2747 + }, + { + "epoch": 1.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2748 + }, + { + "epoch": 1.0982820615261686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2749 + }, + { + "epoch": 1.0986815821014781, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2750 + }, + { + "epoch": 1.099081102676788, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 2751 + }, + { + "epoch": 1.0994806232520975, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2752 + }, + { + "epoch": 1.099880143827407, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2753 + }, + { + "epoch": 1.1002796644027169, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3501, + "step": 2754 + }, + { + "epoch": 1.1006791849780264, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 2755 + }, + { + "epoch": 1.101078705553336, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 2756 + }, + { + "epoch": 1.1014782261286455, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2757 + }, + { + "epoch": 1.1018777467039553, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 2758 + }, + { + "epoch": 1.102277267279265, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 2759 + }, + { + "epoch": 1.1026767878545745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2760 + }, + { + "epoch": 1.103076308429884, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2761 + }, + { + "epoch": 1.1034758290051938, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2762 + }, + { + "epoch": 1.1038753495805034, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 2763 + }, + { + "epoch": 1.104274870155813, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2764 + }, + { + "epoch": 1.1046743907311227, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2765 + }, + { + "epoch": 1.1050739113064323, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2766 + }, + { + "epoch": 1.1054734318817419, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.348, + "step": 2767 + }, + { + "epoch": 1.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2768 + }, + { + "epoch": 1.1062724730323612, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 2769 + }, + { + "epoch": 1.1066719936076708, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2770 + }, + { + "epoch": 1.1070715141829803, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2771 + }, + { + "epoch": 1.1074710347582901, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2772 + }, + { + "epoch": 1.1078705553335997, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2773 + }, + { + "epoch": 1.1082700759089092, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 2774 + }, + { + "epoch": 1.108669596484219, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2775 + }, + { + "epoch": 1.1090691170595286, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 2776 + }, + { + "epoch": 1.1094686376348382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2777 + }, + { + "epoch": 1.1098681582101477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2778 + }, + { + "epoch": 1.1102676787854575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 2779 + }, + { + "epoch": 1.110667199360767, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2780 + }, + { + "epoch": 1.1110667199360766, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 2781 + }, + { + "epoch": 1.1114662405113864, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2782 + }, + { + "epoch": 1.111865761086696, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 2783 + }, + { + "epoch": 1.1122652816620056, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 2784 + }, + { + "epoch": 1.1126648022373151, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 2785 + }, + { + "epoch": 1.113064322812625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2786 + }, + { + "epoch": 1.1134638433879345, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 2787 + }, + { + "epoch": 1.113863363963244, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2788 + }, + { + "epoch": 1.1142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2789 + }, + { + "epoch": 1.1146624051138634, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2790 + }, + { + "epoch": 1.115061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2791 + }, + { + "epoch": 1.1154614462644825, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2792 + }, + { + "epoch": 1.1158609668397923, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2793 + }, + { + "epoch": 1.1162604874151019, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2794 + }, + { + "epoch": 1.1166600079904114, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 2795 + }, + { + "epoch": 1.1170595285657212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2796 + }, + { + "epoch": 1.1174590491410308, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2797 + }, + { + "epoch": 1.1178585697163403, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 2798 + }, + { + "epoch": 1.1182580902916501, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 2799 + }, + { + "epoch": 1.1186576108669597, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 2800 + }, + { + "epoch": 1.1190571314422693, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2801 + }, + { + "epoch": 1.1194566520175788, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 2802 + }, + { + "epoch": 1.1198561725928886, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 2803 + }, + { + "epoch": 1.1202556931681982, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3548, + "step": 2804 + }, + { + "epoch": 1.1206552137435077, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 2805 + }, + { + "epoch": 1.1210547343188173, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2806 + }, + { + "epoch": 1.121454254894127, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2807 + }, + { + "epoch": 1.1218537754694367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2808 + }, + { + "epoch": 1.1222532960447462, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2809 + }, + { + "epoch": 1.122652816620056, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2810 + }, + { + "epoch": 1.1230523371953656, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2811 + }, + { + "epoch": 1.1234518577706751, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3546, + "step": 2812 + }, + { + "epoch": 1.123851378345985, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2813 + }, + { + "epoch": 1.1242508989212945, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 2814 + }, + { + "epoch": 1.124650419496604, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 2815 + }, + { + "epoch": 1.1250499400719136, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 2816 + }, + { + "epoch": 1.1254494606472234, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 2817 + }, + { + "epoch": 1.125848981222533, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2818 + }, + { + "epoch": 1.1262485017978425, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2819 + }, + { + "epoch": 1.1266480223731523, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 2820 + }, + { + "epoch": 1.1270475429484619, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 2821 + }, + { + "epoch": 1.1274470635237714, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 2822 + }, + { + "epoch": 1.127846584099081, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 2823 + }, + { + "epoch": 1.1282461046743908, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 2824 + }, + { + "epoch": 1.1286456252497004, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2825 + }, + { + "epoch": 1.12904514582501, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.352, + "step": 2826 + }, + { + "epoch": 1.1294446664003197, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2827 + }, + { + "epoch": 1.1298441869756293, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2828 + }, + { + "epoch": 1.1302437075509388, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2829 + }, + { + "epoch": 1.1306432281262486, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2830 + }, + { + "epoch": 1.1310427487015582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3495, + "step": 2831 + }, + { + "epoch": 1.1314422692768678, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 2832 + }, + { + "epoch": 1.1318417898521773, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2833 + }, + { + "epoch": 1.132241310427487, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 2834 + }, + { + "epoch": 1.1326408310027967, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2835 + }, + { + "epoch": 1.1330403515781062, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 2836 + }, + { + "epoch": 1.1334398721534158, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 2837 + }, + { + "epoch": 1.1338393927287256, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3519, + "step": 2838 + }, + { + "epoch": 1.1342389133040351, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 2839 + }, + { + "epoch": 1.1346384338793447, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2840 + }, + { + "epoch": 1.1350379544546545, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 2841 + }, + { + "epoch": 1.135437475029964, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2842 + }, + { + "epoch": 1.1358369956052736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2843 + }, + { + "epoch": 1.1362365161805834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2844 + }, + { + "epoch": 1.136636036755893, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2845 + }, + { + "epoch": 1.1370355573312025, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 2846 + }, + { + "epoch": 1.137435077906512, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2847 + }, + { + "epoch": 1.137834598481822, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2848 + }, + { + "epoch": 1.1382341190571315, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 2849 + }, + { + "epoch": 1.138633639632441, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 2850 + }, + { + "epoch": 1.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3578, + "step": 2851 + }, + { + "epoch": 1.1394326807830604, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2852 + }, + { + "epoch": 1.13983220135837, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2853 + }, + { + "epoch": 1.1402317219336795, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 2854 + }, + { + "epoch": 1.1406312425089893, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 2855 + }, + { + "epoch": 1.1410307630842988, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2856 + }, + { + "epoch": 1.1414302836596084, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2857 + }, + { + "epoch": 1.1418298042349182, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2858 + }, + { + "epoch": 1.1422293248102278, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2859 + }, + { + "epoch": 1.1426288453855373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 2860 + }, + { + "epoch": 1.143028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 2861 + }, + { + "epoch": 1.1434278865361567, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2862 + }, + { + "epoch": 1.1438274071114662, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 2863 + }, + { + "epoch": 1.1442269276867758, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 2864 + }, + { + "epoch": 1.1446264482620856, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 2865 + }, + { + "epoch": 1.1450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2866 + }, + { + "epoch": 1.1454254894127047, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2867 + }, + { + "epoch": 1.1458250099880143, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 2868 + }, + { + "epoch": 1.146224530563324, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 2869 + }, + { + "epoch": 1.1466240511386336, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2870 + }, + { + "epoch": 1.1470235717139432, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2871 + }, + { + "epoch": 1.147423092289253, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 2872 + }, + { + "epoch": 1.1478226128645626, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2873 + }, + { + "epoch": 1.1482221334398721, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2874 + }, + { + "epoch": 1.148621654015182, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2875 + }, + { + "epoch": 1.1490211745904915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 2876 + }, + { + "epoch": 1.149420695165801, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 2877 + }, + { + "epoch": 1.1498202157411106, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3634, + "step": 2878 + }, + { + "epoch": 1.1502197363164204, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 2879 + }, + { + "epoch": 1.15061925689173, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 2880 + }, + { + "epoch": 1.1510187774670395, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 2881 + }, + { + "epoch": 1.151418298042349, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 2882 + }, + { + "epoch": 1.1518178186176589, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2883 + }, + { + "epoch": 1.1522173391929684, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 2884 + }, + { + "epoch": 1.152616859768278, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3581, + "step": 2885 + }, + { + "epoch": 1.1530163803435878, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2886 + }, + { + "epoch": 1.1534159009188973, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 2887 + }, + { + "epoch": 1.153815421494207, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 2888 + }, + { + "epoch": 1.1542149420695167, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 2889 + }, + { + "epoch": 1.1546144626448263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2890 + }, + { + "epoch": 1.1550139832201358, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 2891 + }, + { + "epoch": 1.1554135037954454, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 2892 + }, + { + "epoch": 1.1558130243707552, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2893 + }, + { + "epoch": 1.1562125449460647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3572, + "step": 2894 + }, + { + "epoch": 1.1566120655213743, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 2895 + }, + { + "epoch": 1.1570115860966839, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2896 + }, + { + "epoch": 1.1574111066719937, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 2897 + }, + { + "epoch": 1.1578106272473032, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 2898 + }, + { + "epoch": 1.1582101478226128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2899 + }, + { + "epoch": 1.1586096683979226, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2900 + }, + { + "epoch": 1.1590091889732321, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 2901 + }, + { + "epoch": 1.1594087095485417, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2902 + }, + { + "epoch": 1.1598082301238515, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 2903 + }, + { + "epoch": 1.160207750699161, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2904 + }, + { + "epoch": 1.1606072712744706, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 2905 + }, + { + "epoch": 1.1610067918497802, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 2906 + }, + { + "epoch": 1.16140631242509, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 2907 + }, + { + "epoch": 1.1618058330003995, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 2908 + }, + { + "epoch": 1.162205353575709, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 2909 + }, + { + "epoch": 1.1626048741510189, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 2910 + }, + { + "epoch": 1.1630043947263284, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 2911 + }, + { + "epoch": 1.163403915301638, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 2912 + }, + { + "epoch": 1.1638034358769476, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 2913 + }, + { + "epoch": 1.1642029564522574, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 2914 + }, + { + "epoch": 1.164602477027567, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 2915 + }, + { + "epoch": 1.1650019976028765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 2916 + }, + { + "epoch": 1.1654015181781863, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 2917 + }, + { + "epoch": 1.1658010387534958, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2918 + }, + { + "epoch": 1.1662005593288054, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3557, + "step": 2919 + }, + { + "epoch": 1.1666000799041152, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2920 + }, + { + "epoch": 1.1669996004794247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 2921 + }, + { + "epoch": 1.1673991210547343, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 2922 + }, + { + "epoch": 1.1677986416300439, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 2923 + }, + { + "epoch": 1.1681981622053537, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 2924 + }, + { + "epoch": 1.1685976827806632, + "grad_norm": 0.1826171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 2925 + }, + { + "epoch": 1.1689972033559728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2926 + }, + { + "epoch": 1.1693967239312824, + "grad_norm": 0.2431640625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2927 + }, + { + "epoch": 1.1697962445065921, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2928 + }, + { + "epoch": 1.1701957650819017, + "grad_norm": 0.2392578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 2929 + }, + { + "epoch": 1.1705952856572113, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 2930 + }, + { + "epoch": 1.170994806232521, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 2931 + }, + { + "epoch": 1.1713943268078306, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2932 + }, + { + "epoch": 1.1717938473831402, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 2933 + }, + { + "epoch": 1.17219336795845, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 2934 + }, + { + "epoch": 1.1725928885337595, + "grad_norm": 0.193359375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 2935 + }, + { + "epoch": 1.172992409109069, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2936 + }, + { + "epoch": 1.1733919296843787, + "grad_norm": 0.328125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 2937 + }, + { + "epoch": 1.1737914502596885, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3543, + "step": 2938 + }, + { + "epoch": 1.174190970834998, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 2939 + }, + { + "epoch": 1.1745904914103076, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 2940 + }, + { + "epoch": 1.1749900119856171, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2941 + }, + { + "epoch": 1.175389532560927, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 2942 + }, + { + "epoch": 1.1757890531362365, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2943 + }, + { + "epoch": 1.176188573711546, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 2944 + }, + { + "epoch": 1.1765880942868558, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 2945 + }, + { + "epoch": 1.1769876148621654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 2946 + }, + { + "epoch": 1.177387135437475, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 2947 + }, + { + "epoch": 1.1777866560127848, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 2948 + }, + { + "epoch": 1.1781861765880943, + "grad_norm": 0.28125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 2949 + }, + { + "epoch": 1.178585697163404, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3545, + "step": 2950 + }, + { + "epoch": 1.1789852177387135, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 2951 + }, + { + "epoch": 1.1793847383140232, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2952 + }, + { + "epoch": 1.1797842588893328, + "grad_norm": 0.2119140625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2953 + }, + { + "epoch": 1.1801837794646424, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 2954 + }, + { + "epoch": 1.180583300039952, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 2955 + }, + { + "epoch": 1.1809828206152617, + "grad_norm": 0.333984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 2956 + }, + { + "epoch": 1.1813823411905713, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 2957 + }, + { + "epoch": 1.1817818617658808, + "grad_norm": 0.443359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2958 + }, + { + "epoch": 1.1821813823411906, + "grad_norm": 0.150390625, + "learning_rate": 0.0002, + "loss": 1.3528, + "step": 2959 + }, + { + "epoch": 1.1825809029165002, + "grad_norm": 0.263671875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 2960 + }, + { + "epoch": 1.1829804234918098, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 2961 + }, + { + "epoch": 1.1833799440671195, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 2962 + }, + { + "epoch": 1.1837794646424291, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 2963 + }, + { + "epoch": 1.1841789852177387, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3616, + "step": 2964 + }, + { + "epoch": 1.1845785057930485, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 2965 + }, + { + "epoch": 1.184978026368358, + "grad_norm": 0.33984375, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 2966 + }, + { + "epoch": 1.1853775469436676, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 2967 + }, + { + "epoch": 1.1857770675189772, + "grad_norm": 0.29296875, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 2968 + }, + { + "epoch": 1.186176588094287, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 2969 + }, + { + "epoch": 1.1865761086695965, + "grad_norm": 0.37890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 2970 + }, + { + "epoch": 1.186975629244906, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 2971 + }, + { + "epoch": 1.1873751498202156, + "grad_norm": 0.279296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 2972 + }, + { + "epoch": 1.1877746703955254, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 2973 + }, + { + "epoch": 1.188174190970835, + "grad_norm": 0.318359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 2974 + }, + { + "epoch": 1.1885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 2975 + }, + { + "epoch": 1.1889732321214543, + "grad_norm": 0.296875, + "learning_rate": 0.0002, + "loss": 1.3527, + "step": 2976 + }, + { + "epoch": 1.189372752696764, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 2977 + }, + { + "epoch": 1.1897722732720735, + "grad_norm": 0.35546875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 2978 + }, + { + "epoch": 1.1901717938473833, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 2979 + }, + { + "epoch": 1.1905713144226928, + "grad_norm": 0.34375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 2980 + }, + { + "epoch": 1.1909708349980024, + "grad_norm": 0.1845703125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 2981 + }, + { + "epoch": 1.191370355573312, + "grad_norm": 0.4609375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 2982 + }, + { + "epoch": 1.1917698761486217, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 2983 + }, + { + "epoch": 1.1921693967239313, + "grad_norm": 0.322265625, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 2984 + }, + { + "epoch": 1.1925689172992409, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 2985 + }, + { + "epoch": 1.1929684378745504, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 2986 + }, + { + "epoch": 1.1933679584498602, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 2987 + }, + { + "epoch": 1.1937674790251698, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 2988 + }, + { + "epoch": 1.1941669996004793, + "grad_norm": 0.2099609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2989 + }, + { + "epoch": 1.1945665201757891, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 2990 + }, + { + "epoch": 1.1949660407510987, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 2991 + }, + { + "epoch": 1.1953655613264083, + "grad_norm": 0.365234375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 2992 + }, + { + "epoch": 1.195765081901718, + "grad_norm": 0.2578125, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 2993 + }, + { + "epoch": 1.1961646024770276, + "grad_norm": 0.51953125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 2994 + }, + { + "epoch": 1.1965641230523372, + "grad_norm": 0.1767578125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 2995 + }, + { + "epoch": 1.1969636436276467, + "grad_norm": 0.5078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 2996 + }, + { + "epoch": 1.1973631642029565, + "grad_norm": 0.216796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 2997 + }, + { + "epoch": 1.197762684778266, + "grad_norm": 0.54296875, + "learning_rate": 0.0002, + "loss": 1.3549, + "step": 2998 + }, + { + "epoch": 1.1981622053535756, + "grad_norm": 0.27734375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 2999 + }, + { + "epoch": 1.1985617259288852, + "grad_norm": 0.5234375, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3000 + }, + { + "epoch": 1.198961246504195, + "grad_norm": 0.43359375, + "learning_rate": 0.0002, + "loss": 1.3524, + "step": 3001 + }, + { + "epoch": 1.1993607670795046, + "grad_norm": 0.453125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3002 + }, + { + "epoch": 1.1997602876548141, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3003 + }, + { + "epoch": 1.200159808230124, + "grad_norm": 0.2294921875, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 3004 + }, + { + "epoch": 1.2005593288054335, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3005 + }, + { + "epoch": 1.200958849380743, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3006 + }, + { + "epoch": 1.2013583699560528, + "grad_norm": 0.19140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3007 + }, + { + "epoch": 1.2017578905313624, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3008 + }, + { + "epoch": 1.202157411106672, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3009 + }, + { + "epoch": 1.2025569316819817, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 3010 + }, + { + "epoch": 1.2029564522572913, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3563, + "step": 3011 + }, + { + "epoch": 1.2033559728326009, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3012 + }, + { + "epoch": 1.2037554934079104, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3013 + }, + { + "epoch": 1.2041550139832202, + "grad_norm": 0.1875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3014 + }, + { + "epoch": 1.2045545345585298, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3015 + }, + { + "epoch": 1.2049540551338394, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3016 + }, + { + "epoch": 1.205353575709149, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3017 + }, + { + "epoch": 1.2057530962844587, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3018 + }, + { + "epoch": 1.2061526168597683, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3019 + }, + { + "epoch": 1.2065521374350778, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3020 + }, + { + "epoch": 1.2069516580103876, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3021 + }, + { + "epoch": 1.2073511785856972, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3022 + }, + { + "epoch": 1.2077506991610067, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3023 + }, + { + "epoch": 1.2081502197363165, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3024 + }, + { + "epoch": 1.208549740311626, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3025 + }, + { + "epoch": 1.2089492608869357, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3026 + }, + { + "epoch": 1.2093487814622452, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3027 + }, + { + "epoch": 1.209748302037555, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3028 + }, + { + "epoch": 1.2101478226128646, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3539, + "step": 3029 + }, + { + "epoch": 1.2105473431881741, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3030 + }, + { + "epoch": 1.2109468637634837, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3031 + }, + { + "epoch": 1.2113463843387935, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3032 + }, + { + "epoch": 1.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3033 + }, + { + "epoch": 1.2121454254894126, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3034 + }, + { + "epoch": 1.2125449460647224, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3035 + }, + { + "epoch": 1.212944466640032, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3036 + }, + { + "epoch": 1.2133439872153415, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3037 + }, + { + "epoch": 1.2137435077906513, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3038 + }, + { + "epoch": 1.2141430283659609, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 3039 + }, + { + "epoch": 1.2145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3040 + }, + { + "epoch": 1.21494206951658, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3041 + }, + { + "epoch": 1.2153415900918898, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3042 + }, + { + "epoch": 1.2157411106671994, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3043 + }, + { + "epoch": 1.216140631242509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3044 + }, + { + "epoch": 1.2165401518178185, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3045 + }, + { + "epoch": 1.2169396723931283, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3046 + }, + { + "epoch": 1.2173391929684378, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3047 + }, + { + "epoch": 1.2177387135437474, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3048 + }, + { + "epoch": 1.2181382341190572, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3049 + }, + { + "epoch": 1.2185377546943668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3050 + }, + { + "epoch": 1.2189372752696763, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 3051 + }, + { + "epoch": 1.219336795844986, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3052 + }, + { + "epoch": 1.2197363164202957, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3053 + }, + { + "epoch": 1.2201358369956052, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3054 + }, + { + "epoch": 1.220535357570915, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 3055 + }, + { + "epoch": 1.2209348781462246, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3056 + }, + { + "epoch": 1.2213343987215342, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3057 + }, + { + "epoch": 1.2217339192968437, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3058 + }, + { + "epoch": 1.2221334398721535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3059 + }, + { + "epoch": 1.222532960447463, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3060 + }, + { + "epoch": 1.2229324810227726, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3061 + }, + { + "epoch": 1.2233320015980822, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3062 + }, + { + "epoch": 1.223731522173392, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3063 + }, + { + "epoch": 1.2241310427487015, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3064 + }, + { + "epoch": 1.2245305633240111, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3065 + }, + { + "epoch": 1.224930083899321, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3066 + }, + { + "epoch": 1.2253296044746305, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3067 + }, + { + "epoch": 1.22572912504994, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3068 + }, + { + "epoch": 1.2261286456252498, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3069 + }, + { + "epoch": 1.2265281662005594, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3070 + }, + { + "epoch": 1.226927686775869, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3615, + "step": 3071 + }, + { + "epoch": 1.2273272073511785, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3072 + }, + { + "epoch": 1.2277267279264883, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3521, + "step": 3073 + }, + { + "epoch": 1.2281262485017979, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3074 + }, + { + "epoch": 1.2285257690771074, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3075 + }, + { + "epoch": 1.228925289652417, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3594, + "step": 3076 + }, + { + "epoch": 1.2293248102277268, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3077 + }, + { + "epoch": 1.2297243308030363, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3078 + }, + { + "epoch": 1.230123851378346, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3079 + }, + { + "epoch": 1.2305233719536557, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3080 + }, + { + "epoch": 1.2309228925289653, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3081 + }, + { + "epoch": 1.2313224131042748, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3082 + }, + { + "epoch": 1.2317219336795846, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3083 + }, + { + "epoch": 1.2321214542548942, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3084 + }, + { + "epoch": 1.2325209748302037, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3085 + }, + { + "epoch": 1.2329204954055133, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3086 + }, + { + "epoch": 1.233320015980823, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3087 + }, + { + "epoch": 1.2337195365561326, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3088 + }, + { + "epoch": 1.2341190571314422, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3089 + }, + { + "epoch": 1.2345185777067518, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3090 + }, + { + "epoch": 1.2349180982820616, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3091 + }, + { + "epoch": 1.2353176188573711, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3092 + }, + { + "epoch": 1.2357171394326807, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3093 + }, + { + "epoch": 1.2361166600079905, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3094 + }, + { + "epoch": 1.2365161805833, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3095 + }, + { + "epoch": 1.2369157011586096, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3096 + }, + { + "epoch": 1.2373152217339194, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3097 + }, + { + "epoch": 1.237714742309229, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3098 + }, + { + "epoch": 1.2381142628845385, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 3099 + }, + { + "epoch": 1.2385137834598483, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 3100 + }, + { + "epoch": 1.2389133040351579, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3101 + }, + { + "epoch": 1.2393128246104674, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3102 + }, + { + "epoch": 1.239712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3103 + }, + { + "epoch": 1.2401118657610868, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3104 + }, + { + "epoch": 1.2405113863363963, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3105 + }, + { + "epoch": 1.240910906911706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3106 + }, + { + "epoch": 1.2413104274870155, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3107 + }, + { + "epoch": 1.2417099480623253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3503, + "step": 3108 + }, + { + "epoch": 1.2421094686376348, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3109 + }, + { + "epoch": 1.2425089892129444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3110 + }, + { + "epoch": 1.2429085097882542, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3111 + }, + { + "epoch": 1.2433080303635637, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3112 + }, + { + "epoch": 1.2437075509388733, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3113 + }, + { + "epoch": 1.244107071514183, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3114 + }, + { + "epoch": 1.2445065920894927, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3115 + }, + { + "epoch": 1.2449061126648022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3116 + }, + { + "epoch": 1.2453056332401118, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3117 + }, + { + "epoch": 1.2457051538154216, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3118 + }, + { + "epoch": 1.2461046743907311, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3119 + }, + { + "epoch": 1.2465041949660407, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3120 + }, + { + "epoch": 1.2469037155413503, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3121 + }, + { + "epoch": 1.24730323611666, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3122 + }, + { + "epoch": 1.2477027566919696, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3123 + }, + { + "epoch": 1.2481022772672792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3124 + }, + { + "epoch": 1.248501797842589, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3125 + }, + { + "epoch": 1.2489013184178985, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3126 + }, + { + "epoch": 1.249300838993208, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3127 + }, + { + "epoch": 1.2497003595685179, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3128 + }, + { + "epoch": 1.2500998801438274, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 3129 + }, + { + "epoch": 1.250499400719137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3130 + }, + { + "epoch": 1.2508989212944468, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3131 + }, + { + "epoch": 1.2512984418697564, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3132 + }, + { + "epoch": 1.251697962445066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3133 + }, + { + "epoch": 1.2520974830203755, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3134 + }, + { + "epoch": 1.252497003595685, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3135 + }, + { + "epoch": 1.2528965241709948, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3136 + }, + { + "epoch": 1.2532960447463044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3137 + }, + { + "epoch": 1.253695565321614, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3138 + }, + { + "epoch": 1.2540950858969238, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3139 + }, + { + "epoch": 1.2544946064722333, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3140 + }, + { + "epoch": 1.2548941270475429, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3141 + }, + { + "epoch": 1.2552936476228527, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3142 + }, + { + "epoch": 1.2556931681981622, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3143 + }, + { + "epoch": 1.2560926887734718, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3144 + }, + { + "epoch": 1.2564922093487816, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3145 + }, + { + "epoch": 1.2568917299240912, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3146 + }, + { + "epoch": 1.2572912504994007, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3147 + }, + { + "epoch": 1.2576907710747103, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3148 + }, + { + "epoch": 1.2580902916500198, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3149 + }, + { + "epoch": 1.2584898122253296, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3150 + }, + { + "epoch": 1.2588893328006392, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 3151 + }, + { + "epoch": 1.2592888533759488, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3152 + }, + { + "epoch": 1.2596883739512585, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3153 + }, + { + "epoch": 1.260087894526568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3154 + }, + { + "epoch": 1.2604874151018777, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3155 + }, + { + "epoch": 1.2608869356771875, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3156 + }, + { + "epoch": 1.261286456252497, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3157 + }, + { + "epoch": 1.2616859768278066, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3158 + }, + { + "epoch": 1.2620854974031164, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3159 + }, + { + "epoch": 1.262485017978426, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3160 + }, + { + "epoch": 1.2628845385537355, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3161 + }, + { + "epoch": 1.263284059129045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3162 + }, + { + "epoch": 1.2636835797043549, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3163 + }, + { + "epoch": 1.2640831002796644, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3164 + }, + { + "epoch": 1.264482620854974, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3165 + }, + { + "epoch": 1.2648821414302835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3166 + }, + { + "epoch": 1.2652816620055933, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3167 + }, + { + "epoch": 1.265681182580903, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3168 + }, + { + "epoch": 1.2660807031562125, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3169 + }, + { + "epoch": 1.2664802237315222, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3170 + }, + { + "epoch": 1.2668797443068318, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 3171 + }, + { + "epoch": 1.2672792648821414, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3172 + }, + { + "epoch": 1.2676787854574512, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3173 + }, + { + "epoch": 1.2680783060327607, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3174 + }, + { + "epoch": 1.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3175 + }, + { + "epoch": 1.26887734718338, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3176 + }, + { + "epoch": 1.2692768677586896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3177 + }, + { + "epoch": 1.2696763883339992, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3178 + }, + { + "epoch": 1.2700759089093088, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3179 + }, + { + "epoch": 1.2704754294846183, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3180 + }, + { + "epoch": 1.2708749500599281, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3181 + }, + { + "epoch": 1.2712744706352377, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3182 + }, + { + "epoch": 1.2716739912105472, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3183 + }, + { + "epoch": 1.272073511785857, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3184 + }, + { + "epoch": 1.2724730323611666, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3185 + }, + { + "epoch": 1.2728725529364762, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3186 + }, + { + "epoch": 1.273272073511786, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3187 + }, + { + "epoch": 1.2736715940870955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3188 + }, + { + "epoch": 1.274071114662405, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3189 + }, + { + "epoch": 1.2744706352377149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3190 + }, + { + "epoch": 1.2748701558130244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3191 + }, + { + "epoch": 1.275269676388334, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3192 + }, + { + "epoch": 1.2756691969636436, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3193 + }, + { + "epoch": 1.2760687175389531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3194 + }, + { + "epoch": 1.276468238114263, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3195 + }, + { + "epoch": 1.2768677586895725, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3196 + }, + { + "epoch": 1.277267279264882, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3197 + }, + { + "epoch": 1.2776667998401918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3198 + }, + { + "epoch": 1.2780663204155014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3199 + }, + { + "epoch": 1.278465840990811, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3200 + }, + { + "epoch": 1.2788653615661207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3201 + }, + { + "epoch": 1.2792648821414303, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3202 + }, + { + "epoch": 1.2796644027167399, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3203 + }, + { + "epoch": 1.2800639232920497, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3561, + "step": 3204 + }, + { + "epoch": 1.2804634438673592, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3205 + }, + { + "epoch": 1.2808629644426688, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3555, + "step": 3206 + }, + { + "epoch": 1.2812624850179783, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3207 + }, + { + "epoch": 1.2816620055932881, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3473, + "step": 3208 + }, + { + "epoch": 1.2820615261685977, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3209 + }, + { + "epoch": 1.2824610467439073, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3210 + }, + { + "epoch": 1.2828605673192168, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3211 + }, + { + "epoch": 1.2832600878945266, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3212 + }, + { + "epoch": 1.2836596084698362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3213 + }, + { + "epoch": 1.2840591290451457, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3214 + }, + { + "epoch": 1.2844586496204555, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3215 + }, + { + "epoch": 1.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3216 + }, + { + "epoch": 1.2852576907710747, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3217 + }, + { + "epoch": 1.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3218 + }, + { + "epoch": 1.286056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3219 + }, + { + "epoch": 1.2864562524970036, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 3220 + }, + { + "epoch": 1.2868557730723134, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3221 + }, + { + "epoch": 1.287255293647623, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3222 + }, + { + "epoch": 1.2876548142229325, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3223 + }, + { + "epoch": 1.288054334798242, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3224 + }, + { + "epoch": 1.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3225 + }, + { + "epoch": 1.2888533759488614, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3226 + }, + { + "epoch": 1.289252896524171, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3227 + }, + { + "epoch": 1.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3228 + }, + { + "epoch": 1.2900519376747903, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3229 + }, + { + "epoch": 1.2904514582500999, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3230 + }, + { + "epoch": 1.2908509788254094, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3231 + }, + { + "epoch": 1.2912504994007192, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3232 + }, + { + "epoch": 1.2916500199760288, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3233 + }, + { + "epoch": 1.2920495405513384, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3234 + }, + { + "epoch": 1.2924490611266481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3235 + }, + { + "epoch": 1.2928485817019577, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3236 + }, + { + "epoch": 1.2932481022772673, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3237 + }, + { + "epoch": 1.2936476228525768, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3238 + }, + { + "epoch": 1.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3518, + "step": 3239 + }, + { + "epoch": 1.2944466640031962, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3515, + "step": 3240 + }, + { + "epoch": 1.2948461845785058, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3241 + }, + { + "epoch": 1.2952457051538153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3242 + }, + { + "epoch": 1.295645225729125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3243 + }, + { + "epoch": 1.2960447463044347, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3244 + }, + { + "epoch": 1.2964442668797442, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 3245 + }, + { + "epoch": 1.296843787455054, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3246 + }, + { + "epoch": 1.2972433080303636, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3247 + }, + { + "epoch": 1.2976428286056731, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3248 + }, + { + "epoch": 1.298042349180983, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3249 + }, + { + "epoch": 1.2984418697562925, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3250 + }, + { + "epoch": 1.298841390331602, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3251 + }, + { + "epoch": 1.2992409109069116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3252 + }, + { + "epoch": 1.2996404314822214, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3253 + }, + { + "epoch": 1.300039952057531, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3254 + }, + { + "epoch": 1.3004394726328405, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3255 + }, + { + "epoch": 1.30083899320815, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3256 + }, + { + "epoch": 1.30123851378346, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3257 + }, + { + "epoch": 1.3016380343587695, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3258 + }, + { + "epoch": 1.302037554934079, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3259 + }, + { + "epoch": 1.3024370755093888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3260 + }, + { + "epoch": 1.3028365960846984, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3261 + }, + { + "epoch": 1.303236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3262 + }, + { + "epoch": 1.3036356372353177, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 3263 + }, + { + "epoch": 1.3040351578106273, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3264 + }, + { + "epoch": 1.3044346783859369, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3265 + }, + { + "epoch": 1.3048341989612466, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3266 + }, + { + "epoch": 1.3052337195365562, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3267 + }, + { + "epoch": 1.3056332401118658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3268 + }, + { + "epoch": 1.3060327606871753, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3269 + }, + { + "epoch": 1.306432281262485, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3270 + }, + { + "epoch": 1.3068318018377947, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3271 + }, + { + "epoch": 1.3072313224131042, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 3272 + }, + { + "epoch": 1.3076308429884138, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3273 + }, + { + "epoch": 1.3080303635637236, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3274 + }, + { + "epoch": 1.3084298841390332, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3275 + }, + { + "epoch": 1.3088294047143427, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3276 + }, + { + "epoch": 1.3092289252896525, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3277 + }, + { + "epoch": 1.309628445864962, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3278 + }, + { + "epoch": 1.3100279664402716, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3279 + }, + { + "epoch": 1.3104274870155814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3280 + }, + { + "epoch": 1.310827007590891, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3281 + }, + { + "epoch": 1.3112265281662006, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 3282 + }, + { + "epoch": 1.3116260487415101, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3283 + }, + { + "epoch": 1.3120255693168197, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3284 + }, + { + "epoch": 1.3124250898921295, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3285 + }, + { + "epoch": 1.312824610467439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3286 + }, + { + "epoch": 1.3132241310427486, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3287 + }, + { + "epoch": 1.3136236516180584, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3288 + }, + { + "epoch": 1.314023172193368, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3289 + }, + { + "epoch": 1.3144226927686775, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3290 + }, + { + "epoch": 1.3148222133439873, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3291 + }, + { + "epoch": 1.3152217339192969, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3292 + }, + { + "epoch": 1.3156212544946064, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3293 + }, + { + "epoch": 1.3160207750699162, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3294 + }, + { + "epoch": 1.3164202956452258, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3295 + }, + { + "epoch": 1.3168198162205353, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3296 + }, + { + "epoch": 1.317219336795845, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3297 + }, + { + "epoch": 1.3176188573711547, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3298 + }, + { + "epoch": 1.3180183779464643, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3299 + }, + { + "epoch": 1.3184178985217738, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3300 + }, + { + "epoch": 1.3188174190970834, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3301 + }, + { + "epoch": 1.3192169396723932, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3302 + }, + { + "epoch": 1.3196164602477027, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3303 + }, + { + "epoch": 1.3200159808230123, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3304 + }, + { + "epoch": 1.320415501398322, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3305 + }, + { + "epoch": 1.3208150219736317, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3306 + }, + { + "epoch": 1.3212145425489412, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3307 + }, + { + "epoch": 1.321614063124251, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3308 + }, + { + "epoch": 1.3220135836995606, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3309 + }, + { + "epoch": 1.3224131042748701, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3310 + }, + { + "epoch": 1.32281262485018, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3311 + }, + { + "epoch": 1.3232121454254895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3312 + }, + { + "epoch": 1.323611666000799, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3313 + }, + { + "epoch": 1.3240111865761086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3314 + }, + { + "epoch": 1.3244107071514182, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3315 + }, + { + "epoch": 1.324810227726728, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3316 + }, + { + "epoch": 1.3252097483020375, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3317 + }, + { + "epoch": 1.325609268877347, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 3318 + }, + { + "epoch": 1.3260087894526569, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 3319 + }, + { + "epoch": 1.3264083100279664, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3320 + }, + { + "epoch": 1.326807830603276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3321 + }, + { + "epoch": 1.3272073511785858, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3322 + }, + { + "epoch": 1.3276068717538954, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3323 + }, + { + "epoch": 1.328006392329205, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3324 + }, + { + "epoch": 1.3284059129045147, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3325 + }, + { + "epoch": 1.3288054334798243, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3326 + }, + { + "epoch": 1.3292049540551338, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3327 + }, + { + "epoch": 1.3296044746304434, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 3328 + }, + { + "epoch": 1.330003995205753, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3329 + }, + { + "epoch": 1.3304035157810628, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3330 + }, + { + "epoch": 1.3308030363563723, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3331 + }, + { + "epoch": 1.3312025569316819, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3332 + }, + { + "epoch": 1.3316020775069917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3593, + "step": 3333 + }, + { + "epoch": 1.3320015980823012, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3334 + }, + { + "epoch": 1.3324011186576108, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3335 + }, + { + "epoch": 1.3328006392329206, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3336 + }, + { + "epoch": 1.3332001598082301, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3337 + }, + { + "epoch": 1.3335996803835397, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3338 + }, + { + "epoch": 1.3339992009588495, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3339 + }, + { + "epoch": 1.334398721534159, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3340 + }, + { + "epoch": 1.3347982421094686, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3341 + }, + { + "epoch": 1.3351977626847782, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3342 + }, + { + "epoch": 1.335597283260088, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 3343 + }, + { + "epoch": 1.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 3344 + }, + { + "epoch": 1.336396324410707, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3345 + }, + { + "epoch": 1.3367958449860167, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 3346 + }, + { + "epoch": 1.3371953655613265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3347 + }, + { + "epoch": 1.337594886136636, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3348 + }, + { + "epoch": 1.3379944067119456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3349 + }, + { + "epoch": 1.3383939272872554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3350 + }, + { + "epoch": 1.338793447862565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3351 + }, + { + "epoch": 1.3391929684378745, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3352 + }, + { + "epoch": 1.3395924890131843, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3353 + }, + { + "epoch": 1.3399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3354 + }, + { + "epoch": 1.3403915301638034, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 3355 + }, + { + "epoch": 1.3407910507391132, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3356 + }, + { + "epoch": 1.3411905713144228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3357 + }, + { + "epoch": 1.3415900918897323, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3358 + }, + { + "epoch": 1.341989612465042, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 3359 + }, + { + "epoch": 1.3423891330403515, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3360 + }, + { + "epoch": 1.3427886536156612, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 3361 + }, + { + "epoch": 1.3431881741909708, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3362 + }, + { + "epoch": 1.3435876947662804, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3363 + }, + { + "epoch": 1.3439872153415902, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3364 + }, + { + "epoch": 1.3443867359168997, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3365 + }, + { + "epoch": 1.3447862564922093, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3366 + }, + { + "epoch": 1.345185777067519, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3367 + }, + { + "epoch": 1.3455852976428286, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3368 + }, + { + "epoch": 1.3459848182181382, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 3369 + }, + { + "epoch": 1.346384338793448, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3370 + }, + { + "epoch": 1.3467838593687576, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 3371 + }, + { + "epoch": 1.3471833799440671, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3372 + }, + { + "epoch": 1.3475829005193767, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3373 + }, + { + "epoch": 1.3479824210946862, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3374 + }, + { + "epoch": 1.348381941669996, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 3375 + }, + { + "epoch": 1.3487814622453056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3376 + }, + { + "epoch": 1.3491809828206152, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3377 + }, + { + "epoch": 1.349580503395925, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 3378 + }, + { + "epoch": 1.3499800239712345, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 3379 + }, + { + "epoch": 1.350379544546544, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3380 + }, + { + "epoch": 1.3507790651218539, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 3381 + }, + { + "epoch": 1.3511785856971634, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3382 + }, + { + "epoch": 1.351578106272473, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3383 + }, + { + "epoch": 1.3519776268477828, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3384 + }, + { + "epoch": 1.3523771474230923, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 3385 + }, + { + "epoch": 1.352776667998402, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3386 + }, + { + "epoch": 1.3531761885737115, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3387 + }, + { + "epoch": 1.3535757091490213, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3388 + }, + { + "epoch": 1.3539752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3389 + }, + { + "epoch": 1.3543747502996404, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3390 + }, + { + "epoch": 1.35477427087495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3391 + }, + { + "epoch": 1.3551737914502597, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3484, + "step": 3392 + }, + { + "epoch": 1.3555733120255693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3393 + }, + { + "epoch": 1.3559728326008789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3394 + }, + { + "epoch": 1.3563723531761886, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 3395 + }, + { + "epoch": 1.3567718737514982, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3396 + }, + { + "epoch": 1.3571713943268078, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3397 + }, + { + "epoch": 1.3575709149021176, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3398 + }, + { + "epoch": 1.3579704354774271, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 3399 + }, + { + "epoch": 1.3583699560527367, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3400 + }, + { + "epoch": 1.3587694766280465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3401 + }, + { + "epoch": 1.359168997203356, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 3402 + }, + { + "epoch": 1.3595685177786656, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3403 + }, + { + "epoch": 1.3599680383539752, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3404 + }, + { + "epoch": 1.3603675589292847, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3405 + }, + { + "epoch": 1.3607670795045945, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 3406 + }, + { + "epoch": 1.361166600079904, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3407 + }, + { + "epoch": 1.3615661206552137, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3408 + }, + { + "epoch": 1.3619656412305234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3409 + }, + { + "epoch": 1.362365161805833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3410 + }, + { + "epoch": 1.3627646823811426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3411 + }, + { + "epoch": 1.3631642029564524, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 3412 + }, + { + "epoch": 1.363563723531762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3413 + }, + { + "epoch": 1.3639632441070715, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3414 + }, + { + "epoch": 1.3643627646823813, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3415 + }, + { + "epoch": 1.3647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3416 + }, + { + "epoch": 1.3651618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3417 + }, + { + "epoch": 1.36556132640831, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3418 + }, + { + "epoch": 1.3659608469836195, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3419 + }, + { + "epoch": 1.3663603675589293, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3420 + }, + { + "epoch": 1.3667598881342389, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3421 + }, + { + "epoch": 1.3671594087095484, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3422 + }, + { + "epoch": 1.3675589292848582, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3423 + }, + { + "epoch": 1.3679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 3424 + }, + { + "epoch": 1.3683579704354774, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3425 + }, + { + "epoch": 1.3687574910107871, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3426 + }, + { + "epoch": 1.3691570115860967, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3427 + }, + { + "epoch": 1.3695565321614063, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3428 + }, + { + "epoch": 1.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3429 + }, + { + "epoch": 1.3703555733120256, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3430 + }, + { + "epoch": 1.3707550938873352, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3431 + }, + { + "epoch": 1.3711546144626447, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3432 + }, + { + "epoch": 1.3715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3433 + }, + { + "epoch": 1.371953655613264, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3522, + "step": 3434 + }, + { + "epoch": 1.3723531761885737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3435 + }, + { + "epoch": 1.3727526967638832, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3436 + }, + { + "epoch": 1.373152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 3437 + }, + { + "epoch": 1.3735517379145026, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3438 + }, + { + "epoch": 1.3739512584898121, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3439 + }, + { + "epoch": 1.374350779065122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3440 + }, + { + "epoch": 1.3747502996404315, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3441 + }, + { + "epoch": 1.375149820215741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 3442 + }, + { + "epoch": 1.3755493407910508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 3443 + }, + { + "epoch": 1.3759488613663604, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3444 + }, + { + "epoch": 1.37634838194167, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3445 + }, + { + "epoch": 1.3767479025169798, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3446 + }, + { + "epoch": 1.3771474230922893, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3447 + }, + { + "epoch": 1.3775469436675989, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3448 + }, + { + "epoch": 1.3779464642429085, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 3449 + }, + { + "epoch": 1.378345984818218, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3450 + }, + { + "epoch": 1.3787455053935278, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 3451 + }, + { + "epoch": 1.3791450259688374, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3452 + }, + { + "epoch": 1.379544546544147, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3453 + }, + { + "epoch": 1.3799440671194567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3454 + }, + { + "epoch": 1.3803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3455 + }, + { + "epoch": 1.3807431082700758, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3456 + }, + { + "epoch": 1.3811426288453856, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3457 + }, + { + "epoch": 1.3815421494206952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3458 + }, + { + "epoch": 1.3819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3459 + }, + { + "epoch": 1.3823411905713145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3460 + }, + { + "epoch": 1.3827407111466241, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3461 + }, + { + "epoch": 1.3831402317219337, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3462 + }, + { + "epoch": 1.3835397522972432, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3463 + }, + { + "epoch": 1.3839392728725528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3475, + "step": 3464 + }, + { + "epoch": 1.3843387934478626, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3538, + "step": 3465 + }, + { + "epoch": 1.3847383140231722, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3466 + }, + { + "epoch": 1.3851378345984817, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3467 + }, + { + "epoch": 1.3855373551737915, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3468 + }, + { + "epoch": 1.385936875749101, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3469 + }, + { + "epoch": 1.3863363963244106, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3470 + }, + { + "epoch": 1.3867359168997204, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3471 + }, + { + "epoch": 1.38713543747503, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3472 + }, + { + "epoch": 1.3875349580503396, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 3473 + }, + { + "epoch": 1.3879344786256493, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3474 + }, + { + "epoch": 1.388333999200959, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3475 + }, + { + "epoch": 1.3887335197762685, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3476 + }, + { + "epoch": 1.389133040351578, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3477 + }, + { + "epoch": 1.3895325609268878, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3478 + }, + { + "epoch": 1.3899320815021974, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3479 + }, + { + "epoch": 1.390331602077507, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3480 + }, + { + "epoch": 1.3907311226528165, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3481 + }, + { + "epoch": 1.3911306432281263, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3482 + }, + { + "epoch": 1.3915301638034359, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3483 + }, + { + "epoch": 1.3919296843787454, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3484 + }, + { + "epoch": 1.3923292049540552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3485 + }, + { + "epoch": 1.3927287255293648, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3486 + }, + { + "epoch": 1.3931282461046743, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3487 + }, + { + "epoch": 1.3935277666799841, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 3488 + }, + { + "epoch": 1.3939272872552937, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3489 + }, + { + "epoch": 1.3943268078306033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3490 + }, + { + "epoch": 1.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3491 + }, + { + "epoch": 1.3951258489812226, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3492 + }, + { + "epoch": 1.3955253695565322, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3493 + }, + { + "epoch": 1.3959248901318417, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3494 + }, + { + "epoch": 1.3963244107071513, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3495 + }, + { + "epoch": 1.396723931282461, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3496 + }, + { + "epoch": 1.3971234518577706, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3497 + }, + { + "epoch": 1.3975229724330802, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 3498 + }, + { + "epoch": 1.39792249300839, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3499 + }, + { + "epoch": 1.3983220135836996, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 3500 + }, + { + "epoch": 1.3987215341590091, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 3501 + }, + { + "epoch": 1.399121054734319, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3502 + }, + { + "epoch": 1.3995205753096285, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3503 + }, + { + "epoch": 1.399920095884938, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3504 + }, + { + "epoch": 1.4003196164602478, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3505 + }, + { + "epoch": 1.4007191370355574, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3506 + }, + { + "epoch": 1.401118657610867, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3507 + }, + { + "epoch": 1.4015181781861765, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3508 + }, + { + "epoch": 1.401917698761486, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3509 + }, + { + "epoch": 1.4023172193367959, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3510 + }, + { + "epoch": 1.4027167399121054, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3511 + }, + { + "epoch": 1.403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3512 + }, + { + "epoch": 1.4035157810627248, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3513 + }, + { + "epoch": 1.4039153016380344, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3477, + "step": 3514 + }, + { + "epoch": 1.404314822213344, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3515 + }, + { + "epoch": 1.4047143427886537, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3516 + }, + { + "epoch": 1.4051138633639633, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3517 + }, + { + "epoch": 1.4055133839392728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3518 + }, + { + "epoch": 1.4059129045145826, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3519 + }, + { + "epoch": 1.4063124250898922, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3520 + }, + { + "epoch": 1.4067119456652017, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3521 + }, + { + "epoch": 1.4071114662405113, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3522 + }, + { + "epoch": 1.407510986815821, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3523 + }, + { + "epoch": 1.4079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3524 + }, + { + "epoch": 1.4083100279664402, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3525 + }, + { + "epoch": 1.4087095485417498, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3526 + }, + { + "epoch": 1.4091090691170596, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3527 + }, + { + "epoch": 1.4095085896923691, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 3528 + }, + { + "epoch": 1.4099081102676787, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3529 + }, + { + "epoch": 1.4103076308429885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 3530 + }, + { + "epoch": 1.410707151418298, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3531 + }, + { + "epoch": 1.4111066719936076, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 3532 + }, + { + "epoch": 1.4115061925689174, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3533 + }, + { + "epoch": 1.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3534 + }, + { + "epoch": 1.4123052337195365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3535 + }, + { + "epoch": 1.4127047542948463, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3536 + }, + { + "epoch": 1.4131042748701559, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3537 + }, + { + "epoch": 1.4135037954454654, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3476, + "step": 3538 + }, + { + "epoch": 1.413903316020775, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3539 + }, + { + "epoch": 1.4143028365960846, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3540 + }, + { + "epoch": 1.4147023571713944, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 3541 + }, + { + "epoch": 1.415101877746704, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3542 + }, + { + "epoch": 1.4155013983220135, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 3543 + }, + { + "epoch": 1.4159009188973233, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.35, + "step": 3544 + }, + { + "epoch": 1.4163004394726328, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3545 + }, + { + "epoch": 1.4166999600479424, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 3546 + }, + { + "epoch": 1.4170994806232522, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3547 + }, + { + "epoch": 1.4174990011985618, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3548 + }, + { + "epoch": 1.4178985217738713, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3549 + }, + { + "epoch": 1.418298042349181, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 3550 + }, + { + "epoch": 1.4186975629244907, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3551 + }, + { + "epoch": 1.4190970834998002, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3552 + }, + { + "epoch": 1.4194966040751098, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3553 + }, + { + "epoch": 1.4198961246504194, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3554 + }, + { + "epoch": 1.4202956452257292, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3555 + }, + { + "epoch": 1.4206951658010387, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3556 + }, + { + "epoch": 1.4210946863763483, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3557 + }, + { + "epoch": 1.421494206951658, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3558 + }, + { + "epoch": 1.4218937275269676, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3559 + }, + { + "epoch": 1.4222932481022772, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 3560 + }, + { + "epoch": 1.422692768677587, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3561 + }, + { + "epoch": 1.4230922892528965, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3562 + }, + { + "epoch": 1.4234918098282061, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 3563 + }, + { + "epoch": 1.423891330403516, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3564 + }, + { + "epoch": 1.4242908509788255, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3565 + }, + { + "epoch": 1.424690371554135, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3566 + }, + { + "epoch": 1.4250898921294446, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3567 + }, + { + "epoch": 1.4254894127047544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3568 + }, + { + "epoch": 1.425888933280064, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 3569 + }, + { + "epoch": 1.4262884538553735, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 3570 + }, + { + "epoch": 1.426687974430683, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3571 + }, + { + "epoch": 1.4270874950059929, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3572 + }, + { + "epoch": 1.4274870155813024, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3573 + }, + { + "epoch": 1.427886536156612, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3574 + }, + { + "epoch": 1.4282860567319218, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3575 + }, + { + "epoch": 1.4286855773072313, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 3576 + }, + { + "epoch": 1.429085097882541, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 3577 + }, + { + "epoch": 1.4294846184578507, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3578 + }, + { + "epoch": 1.4298841390331603, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3579 + }, + { + "epoch": 1.4302836596084698, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3580 + }, + { + "epoch": 1.4306831801837796, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3479, + "step": 3581 + }, + { + "epoch": 1.4310827007590892, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3582 + }, + { + "epoch": 1.4314822213343987, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3583 + }, + { + "epoch": 1.4318817419097083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3584 + }, + { + "epoch": 1.4322812624850179, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3585 + }, + { + "epoch": 1.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3586 + }, + { + "epoch": 1.4330803036356372, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3587 + }, + { + "epoch": 1.4334798242109468, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3588 + }, + { + "epoch": 1.4338793447862566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3589 + }, + { + "epoch": 1.4342788653615661, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3590 + }, + { + "epoch": 1.4346783859368757, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3591 + }, + { + "epoch": 1.4350779065121855, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 3592 + }, + { + "epoch": 1.435477427087495, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3593 + }, + { + "epoch": 1.4358769476628046, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3594 + }, + { + "epoch": 1.4362764682381144, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 3595 + }, + { + "epoch": 1.436675988813424, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3596 + }, + { + "epoch": 1.4370755093887335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3597 + }, + { + "epoch": 1.437475029964043, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 3598 + }, + { + "epoch": 1.4378745505393526, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 3599 + }, + { + "epoch": 1.4382740711146624, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3600 + }, + { + "epoch": 1.438673591689972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3601 + }, + { + "epoch": 1.4390731122652816, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3602 + }, + { + "epoch": 1.4394726328405913, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3603 + }, + { + "epoch": 1.439872153415901, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 3604 + }, + { + "epoch": 1.4402716739912105, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3605 + }, + { + "epoch": 1.4406711945665203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3606 + }, + { + "epoch": 1.4410707151418298, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3607 + }, + { + "epoch": 1.4414702357171394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3608 + }, + { + "epoch": 1.4418697562924492, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3609 + }, + { + "epoch": 1.4422692768677587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3610 + }, + { + "epoch": 1.4426687974430683, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3611 + }, + { + "epoch": 1.4430683180183779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3612 + }, + { + "epoch": 1.4434678385936877, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3613 + }, + { + "epoch": 1.4438673591689972, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3614 + }, + { + "epoch": 1.4442668797443068, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3497, + "step": 3615 + }, + { + "epoch": 1.4446664003196164, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 3616 + }, + { + "epoch": 1.4450659208949261, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3617 + }, + { + "epoch": 1.4454654414702357, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 3618 + }, + { + "epoch": 1.4458649620455453, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3619 + }, + { + "epoch": 1.446264482620855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 3620 + }, + { + "epoch": 1.4466640031961646, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 3621 + }, + { + "epoch": 1.4470635237714742, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3622 + }, + { + "epoch": 1.447463044346784, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 3623 + }, + { + "epoch": 1.4478625649220935, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3624 + }, + { + "epoch": 1.448262085497403, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3625 + }, + { + "epoch": 1.4486616060727129, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3626 + }, + { + "epoch": 1.4490611266480224, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 3627 + }, + { + "epoch": 1.449460647223332, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3628 + }, + { + "epoch": 1.4498601677986416, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3629 + }, + { + "epoch": 1.4502596883739511, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3630 + }, + { + "epoch": 1.450659208949261, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3631 + }, + { + "epoch": 1.4510587295245705, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3632 + }, + { + "epoch": 1.45145825009988, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3633 + }, + { + "epoch": 1.4518577706751898, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3634 + }, + { + "epoch": 1.4522572912504994, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3635 + }, + { + "epoch": 1.452656811825809, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 3636 + }, + { + "epoch": 1.4530563324011188, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3637 + }, + { + "epoch": 1.4534558529764283, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 3638 + }, + { + "epoch": 1.4538553735517379, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3639 + }, + { + "epoch": 1.4542548941270477, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3640 + }, + { + "epoch": 1.4546544147023572, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 3641 + }, + { + "epoch": 1.4550539352776668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3642 + }, + { + "epoch": 1.4554534558529764, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 3643 + }, + { + "epoch": 1.455852976428286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3644 + }, + { + "epoch": 1.4562524970035957, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3645 + }, + { + "epoch": 1.4566520175789053, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 3646 + }, + { + "epoch": 1.4570515381542148, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3647 + }, + { + "epoch": 1.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3648 + }, + { + "epoch": 1.4578505793048342, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3649 + }, + { + "epoch": 1.4582500998801438, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3650 + }, + { + "epoch": 1.4586496204554535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3651 + }, + { + "epoch": 1.459049141030763, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3652 + }, + { + "epoch": 1.4594486616060727, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 3653 + }, + { + "epoch": 1.4598481821813825, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 3654 + }, + { + "epoch": 1.460247702756692, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 3655 + }, + { + "epoch": 1.4606472233320016, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3656 + }, + { + "epoch": 1.4610467439073112, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 3657 + }, + { + "epoch": 1.461446264482621, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3658 + }, + { + "epoch": 1.4618457850579305, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3659 + }, + { + "epoch": 1.46224530563324, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 3660 + }, + { + "epoch": 1.4626448262085496, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 3661 + }, + { + "epoch": 1.4630443467838594, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3662 + }, + { + "epoch": 1.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3663 + }, + { + "epoch": 1.4638433879344785, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3664 + }, + { + "epoch": 1.4642429085097883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3665 + }, + { + "epoch": 1.464642429085098, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3666 + }, + { + "epoch": 1.4650419496604075, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3667 + }, + { + "epoch": 1.4654414702357172, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 3668 + }, + { + "epoch": 1.4658409908110268, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3669 + }, + { + "epoch": 1.4662405113863364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3670 + }, + { + "epoch": 1.4666400319616462, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3671 + }, + { + "epoch": 1.4670395525369557, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3672 + }, + { + "epoch": 1.4674390731122653, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3673 + }, + { + "epoch": 1.4678385936875749, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3674 + }, + { + "epoch": 1.4682381142628844, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 3675 + }, + { + "epoch": 1.4686376348381942, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3511, + "step": 3676 + }, + { + "epoch": 1.4690371554135038, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 3677 + }, + { + "epoch": 1.4694366759888133, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3678 + }, + { + "epoch": 1.4698361965641231, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3679 + }, + { + "epoch": 1.4702357171394327, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3680 + }, + { + "epoch": 1.4706352377147422, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 3681 + }, + { + "epoch": 1.471034758290052, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3498, + "step": 3682 + }, + { + "epoch": 1.4714342788653616, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3683 + }, + { + "epoch": 1.4718337994406712, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3684 + }, + { + "epoch": 1.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3685 + }, + { + "epoch": 1.4726328405912905, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3686 + }, + { + "epoch": 1.4730323611666, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3687 + }, + { + "epoch": 1.4734318817419096, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3688 + }, + { + "epoch": 1.4738314023172192, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3689 + }, + { + "epoch": 1.474230922892529, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3690 + }, + { + "epoch": 1.4746304434678386, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3691 + }, + { + "epoch": 1.4750299640431481, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3692 + }, + { + "epoch": 1.475429484618458, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3693 + }, + { + "epoch": 1.4758290051937675, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3694 + }, + { + "epoch": 1.476228525769077, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3695 + }, + { + "epoch": 1.4766280463443868, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3696 + }, + { + "epoch": 1.4770275669196964, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 3697 + }, + { + "epoch": 1.477427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3698 + }, + { + "epoch": 1.4778266080703157, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3699 + }, + { + "epoch": 1.4782261286456253, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3700 + }, + { + "epoch": 1.4786256492209349, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 3701 + }, + { + "epoch": 1.4790251697962444, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 3702 + }, + { + "epoch": 1.479424690371554, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 3703 + }, + { + "epoch": 1.4798242109468638, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3704 + }, + { + "epoch": 1.4802237315221733, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 3705 + }, + { + "epoch": 1.480623252097483, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3506, + "step": 3706 + }, + { + "epoch": 1.4810227726727927, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3707 + }, + { + "epoch": 1.4814222932481023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3708 + }, + { + "epoch": 1.4818218138234118, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 3709 + }, + { + "epoch": 1.4822213343987216, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 3710 + }, + { + "epoch": 1.4826208549740312, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3711 + }, + { + "epoch": 1.4830203755493407, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 3712 + }, + { + "epoch": 1.4834198961246505, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3713 + }, + { + "epoch": 1.48381941669996, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 3714 + }, + { + "epoch": 1.4842189372752697, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3715 + }, + { + "epoch": 1.4846184578505794, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3716 + }, + { + "epoch": 1.485017978425889, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 3717 + }, + { + "epoch": 1.4854174990011986, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3718 + }, + { + "epoch": 1.4858170195765081, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3719 + }, + { + "epoch": 1.4862165401518177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 3720 + }, + { + "epoch": 1.4866160607271275, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3721 + }, + { + "epoch": 1.487015581302437, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3722 + }, + { + "epoch": 1.4874151018777466, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3723 + }, + { + "epoch": 1.4878146224530564, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3724 + }, + { + "epoch": 1.488214143028366, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3725 + }, + { + "epoch": 1.4886136636036755, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3726 + }, + { + "epoch": 1.4890131841789853, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3727 + }, + { + "epoch": 1.4894127047542949, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3728 + }, + { + "epoch": 1.4898122253296044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3729 + }, + { + "epoch": 1.4902117459049142, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3730 + }, + { + "epoch": 1.4906112664802238, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3731 + }, + { + "epoch": 1.4910107870555334, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3732 + }, + { + "epoch": 1.491410307630843, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3733 + }, + { + "epoch": 1.4918098282061525, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3734 + }, + { + "epoch": 1.4922093487814623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3735 + }, + { + "epoch": 1.4926088693567718, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3736 + }, + { + "epoch": 1.4930083899320814, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 3737 + }, + { + "epoch": 1.4934079105073912, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3738 + }, + { + "epoch": 1.4938074310827008, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3739 + }, + { + "epoch": 1.4942069516580103, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3740 + }, + { + "epoch": 1.49460647223332, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3458, + "step": 3741 + }, + { + "epoch": 1.4950059928086297, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3742 + }, + { + "epoch": 1.4954055133839392, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3743 + }, + { + "epoch": 1.495805033959249, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3744 + }, + { + "epoch": 1.4962045545345586, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 3745 + }, + { + "epoch": 1.4966040751098681, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3746 + }, + { + "epoch": 1.4970035956851777, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 3747 + }, + { + "epoch": 1.4974031162604873, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3748 + }, + { + "epoch": 1.497802636835797, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3749 + }, + { + "epoch": 1.4982021574111066, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 3750 + }, + { + "epoch": 1.4986016779864162, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 3751 + }, + { + "epoch": 1.499001198561726, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3752 + }, + { + "epoch": 1.4994007191370355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3753 + }, + { + "epoch": 1.499800239712345, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 3754 + }, + { + "epoch": 1.500199760287655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3755 + }, + { + "epoch": 1.5005992808629645, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 3756 + }, + { + "epoch": 1.500998801438274, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3757 + }, + { + "epoch": 1.5013983220135838, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3758 + }, + { + "epoch": 1.5017978425888934, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 3759 + }, + { + "epoch": 1.502197363164203, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3481, + "step": 3760 + }, + { + "epoch": 1.5025968837395127, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 3761 + }, + { + "epoch": 1.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3762 + }, + { + "epoch": 1.5033959248901319, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3763 + }, + { + "epoch": 1.5037954454654414, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3764 + }, + { + "epoch": 1.504194966040751, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 3765 + }, + { + "epoch": 1.5045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 3766 + }, + { + "epoch": 1.5049940071913703, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 3767 + }, + { + "epoch": 1.50539352776668, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3768 + }, + { + "epoch": 1.5057930483419897, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3769 + }, + { + "epoch": 1.5061925689172992, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 3770 + }, + { + "epoch": 1.5065920894926088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 3771 + }, + { + "epoch": 1.5069916100679186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3772 + }, + { + "epoch": 1.5073911306432282, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 3773 + }, + { + "epoch": 1.5077906512185377, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 3774 + }, + { + "epoch": 1.5081901717938475, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3775 + }, + { + "epoch": 1.5085896923691569, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 3776 + }, + { + "epoch": 1.5089892129444666, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 3777 + }, + { + "epoch": 1.5093887335197764, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 3778 + }, + { + "epoch": 1.5097882540950858, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 3779 + }, + { + "epoch": 1.5101877746703956, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3780 + }, + { + "epoch": 1.5105872952457051, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3781 + }, + { + "epoch": 1.5109868158210147, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3782 + }, + { + "epoch": 1.5113863363963245, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3783 + }, + { + "epoch": 1.511785856971634, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3784 + }, + { + "epoch": 1.5121853775469436, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 3785 + }, + { + "epoch": 1.5125848981222534, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 3786 + }, + { + "epoch": 1.512984418697563, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 3787 + }, + { + "epoch": 1.5133839392728725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3788 + }, + { + "epoch": 1.5137834598481823, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 3789 + }, + { + "epoch": 1.5141829804234919, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3790 + }, + { + "epoch": 1.5145825009988014, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 3791 + }, + { + "epoch": 1.5149820215741112, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3792 + }, + { + "epoch": 1.5153815421494206, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 3793 + }, + { + "epoch": 1.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 3794 + }, + { + "epoch": 1.51618058330004, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3795 + }, + { + "epoch": 1.5165801038753495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3796 + }, + { + "epoch": 1.5169796244506593, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 3797 + }, + { + "epoch": 1.5173791450259688, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3798 + }, + { + "epoch": 1.5177786656012784, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 3799 + }, + { + "epoch": 1.5181781861765882, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 3800 + }, + { + "epoch": 1.5185777067518977, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 3801 + }, + { + "epoch": 1.5189772273272073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3802 + }, + { + "epoch": 1.519376747902517, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3803 + }, + { + "epoch": 1.5197762684778267, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3804 + }, + { + "epoch": 1.5201757890531362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 3805 + }, + { + "epoch": 1.520575309628446, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 3806 + }, + { + "epoch": 1.5209748302037553, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3807 + }, + { + "epoch": 1.5213743507790651, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 3808 + }, + { + "epoch": 1.5217738713543747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3809 + }, + { + "epoch": 1.5221733919296843, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 3810 + }, + { + "epoch": 1.522572912504994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 3811 + }, + { + "epoch": 1.5229724330803036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3812 + }, + { + "epoch": 1.5233719536556132, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 3813 + }, + { + "epoch": 1.523771474230923, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 3814 + }, + { + "epoch": 1.5241709948062325, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 3815 + }, + { + "epoch": 1.524570515381542, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 3816 + }, + { + "epoch": 1.5249700359568519, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 3817 + }, + { + "epoch": 1.5253695565321614, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3468, + "step": 3818 + }, + { + "epoch": 1.525769077107471, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3819 + }, + { + "epoch": 1.5261685976827808, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3820 + }, + { + "epoch": 1.5265681182580901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3537, + "step": 3821 + }, + { + "epoch": 1.5269676388334, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3526, + "step": 3822 + }, + { + "epoch": 1.5273671594087097, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3823 + }, + { + "epoch": 1.527766679984019, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 3824 + }, + { + "epoch": 1.5281662005593288, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 3825 + }, + { + "epoch": 1.5285657211346384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3826 + }, + { + "epoch": 1.528965241709948, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 3827 + }, + { + "epoch": 1.5293647622852578, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3828 + }, + { + "epoch": 1.5297642828605673, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 3829 + }, + { + "epoch": 1.5301638034358769, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 3830 + }, + { + "epoch": 1.5305633240111867, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3831 + }, + { + "epoch": 1.5309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 3832 + }, + { + "epoch": 1.5313623651618058, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 3833 + }, + { + "epoch": 1.5317618857371156, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 3834 + }, + { + "epoch": 1.5321614063124251, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 3835 + }, + { + "epoch": 1.5325609268877347, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3836 + }, + { + "epoch": 1.5329604474630445, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3837 + }, + { + "epoch": 1.5333599680383538, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3838 + }, + { + "epoch": 1.5337594886136636, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3839 + }, + { + "epoch": 1.5341590091889732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 3840 + }, + { + "epoch": 1.5345585297642828, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3841 + }, + { + "epoch": 1.5349580503395925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3842 + }, + { + "epoch": 1.535357570914902, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 3843 + }, + { + "epoch": 1.5357570914902117, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3844 + }, + { + "epoch": 1.5361566120655215, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 3845 + }, + { + "epoch": 1.536556132640831, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3846 + }, + { + "epoch": 1.5369556532161406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3847 + }, + { + "epoch": 1.5373551737914504, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3848 + }, + { + "epoch": 1.53775469436676, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 3849 + }, + { + "epoch": 1.5381542149420695, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3850 + }, + { + "epoch": 1.5385537355173793, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3851 + }, + { + "epoch": 1.5389532560926886, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3852 + }, + { + "epoch": 1.5393527766679984, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 3853 + }, + { + "epoch": 1.539752297243308, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3854 + }, + { + "epoch": 1.5401518178186175, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 3855 + }, + { + "epoch": 1.5405513383939273, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 3856 + }, + { + "epoch": 1.540950858969237, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3857 + }, + { + "epoch": 1.5413503795445465, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 3858 + }, + { + "epoch": 1.5417499001198562, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3859 + }, + { + "epoch": 1.5421494206951658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 3860 + }, + { + "epoch": 1.5425489412704754, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3861 + }, + { + "epoch": 1.5429484618457852, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 3862 + }, + { + "epoch": 1.5433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3863 + }, + { + "epoch": 1.5437475029964043, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3864 + }, + { + "epoch": 1.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3865 + }, + { + "epoch": 1.5445465441470234, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3866 + }, + { + "epoch": 1.5449460647223332, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3867 + }, + { + "epoch": 1.545345585297643, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 3868 + }, + { + "epoch": 1.5457451058729523, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 3869 + }, + { + "epoch": 1.5461446264482621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 3870 + }, + { + "epoch": 1.5465441470235717, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 3871 + }, + { + "epoch": 1.5469436675988812, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 3872 + }, + { + "epoch": 1.547343188174191, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3873 + }, + { + "epoch": 1.5477427087495006, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 3874 + }, + { + "epoch": 1.5481422293248102, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 3875 + }, + { + "epoch": 1.54854174990012, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 3876 + }, + { + "epoch": 1.5489412704754295, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3877 + }, + { + "epoch": 1.549340791050739, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 3878 + }, + { + "epoch": 1.5497403116260489, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 3879 + }, + { + "epoch": 1.5501398322013584, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 3880 + }, + { + "epoch": 1.550539352776668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3881 + }, + { + "epoch": 1.5509388733519778, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 3882 + }, + { + "epoch": 1.5513383939272871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 3883 + }, + { + "epoch": 1.551737914502597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3552, + "step": 3884 + }, + { + "epoch": 1.5521374350779065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3885 + }, + { + "epoch": 1.552536955653216, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 3886 + }, + { + "epoch": 1.5529364762285258, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3887 + }, + { + "epoch": 1.5533359968038354, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3888 + }, + { + "epoch": 1.553735517379145, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3889 + }, + { + "epoch": 1.5541350379544547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 3890 + }, + { + "epoch": 1.5545345585297643, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3891 + }, + { + "epoch": 1.5549340791050739, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3892 + }, + { + "epoch": 1.5553335996803836, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 3893 + }, + { + "epoch": 1.5557331202556932, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3894 + }, + { + "epoch": 1.5561326408310028, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 3895 + }, + { + "epoch": 1.5565321614063126, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 3896 + }, + { + "epoch": 1.556931681981622, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 3897 + }, + { + "epoch": 1.5573312025569317, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 3898 + }, + { + "epoch": 1.5577307231322413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 3899 + }, + { + "epoch": 1.5581302437075508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3900 + }, + { + "epoch": 1.5585297642828606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3901 + }, + { + "epoch": 1.5589292848581702, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3902 + }, + { + "epoch": 1.5593288054334797, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 3903 + }, + { + "epoch": 1.5597283260087895, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 3904 + }, + { + "epoch": 1.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 3905 + }, + { + "epoch": 1.5605273671594087, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 3906 + }, + { + "epoch": 1.5609268877347184, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 3907 + }, + { + "epoch": 1.561326408310028, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3908 + }, + { + "epoch": 1.5617259288853376, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3909 + }, + { + "epoch": 1.5621254494606474, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3910 + }, + { + "epoch": 1.5625249700359567, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 3911 + }, + { + "epoch": 1.5629244906112665, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 3912 + }, + { + "epoch": 1.5633240111865763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 3913 + }, + { + "epoch": 1.5637235317618856, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3914 + }, + { + "epoch": 1.5641230523371954, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 3915 + }, + { + "epoch": 1.564522572912505, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 3916 + }, + { + "epoch": 1.5649220934878145, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3917 + }, + { + "epoch": 1.5653216140631243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 3918 + }, + { + "epoch": 1.5657211346384339, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 3919 + }, + { + "epoch": 1.5661206552137434, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 3920 + }, + { + "epoch": 1.5665201757890532, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 3921 + }, + { + "epoch": 1.5669196963643628, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 3922 + }, + { + "epoch": 1.5673192169396724, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 3923 + }, + { + "epoch": 1.5677187375149821, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 3924 + }, + { + "epoch": 1.5681182580902915, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3925 + }, + { + "epoch": 1.5685177786656013, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 3926 + }, + { + "epoch": 1.568917299240911, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 3927 + }, + { + "epoch": 1.5693168198162204, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 3928 + }, + { + "epoch": 1.5697163403915302, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 3929 + }, + { + "epoch": 1.5701158609668397, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 3930 + }, + { + "epoch": 1.5705153815421493, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 3931 + }, + { + "epoch": 1.570914902117459, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3932 + }, + { + "epoch": 1.5713144226927687, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 3933 + }, + { + "epoch": 1.5717139432680782, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 3934 + }, + { + "epoch": 1.572113463843388, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 3935 + }, + { + "epoch": 1.5725129844186976, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 3936 + }, + { + "epoch": 1.5729125049940071, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 3937 + }, + { + "epoch": 1.573312025569317, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 3938 + }, + { + "epoch": 1.5737115461446265, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 3939 + }, + { + "epoch": 1.574111066719936, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3940 + }, + { + "epoch": 1.5745105872952458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 3941 + }, + { + "epoch": 1.5749101078705552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 3942 + }, + { + "epoch": 1.575309628445865, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 3943 + }, + { + "epoch": 1.5757091490211745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 3944 + }, + { + "epoch": 1.576108669596484, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 3945 + }, + { + "epoch": 1.5765081901717939, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 3946 + }, + { + "epoch": 1.5769077107471035, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3947 + }, + { + "epoch": 1.577307231322413, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3504, + "step": 3948 + }, + { + "epoch": 1.5777067518977228, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3949 + }, + { + "epoch": 1.5781062724730324, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3950 + }, + { + "epoch": 1.578505793048342, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 3951 + }, + { + "epoch": 1.5789053136236517, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 3952 + }, + { + "epoch": 1.5793048341989613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 3953 + }, + { + "epoch": 1.5797043547742708, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 3954 + }, + { + "epoch": 1.5801038753495806, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 3955 + }, + { + "epoch": 1.58050339592489, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 3956 + }, + { + "epoch": 1.5809029165001998, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 3957 + }, + { + "epoch": 1.5813024370755095, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 3958 + }, + { + "epoch": 1.581701957650819, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 3959 + }, + { + "epoch": 1.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 3960 + }, + { + "epoch": 1.5825009988014382, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 3961 + }, + { + "epoch": 1.5829005193767478, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 3962 + }, + { + "epoch": 1.5833000399520576, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 3963 + }, + { + "epoch": 1.5836995605273672, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 3964 + }, + { + "epoch": 1.5840990811026767, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 3965 + }, + { + "epoch": 1.5844986016779865, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 3966 + }, + { + "epoch": 1.584898122253296, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 3967 + }, + { + "epoch": 1.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 3968 + }, + { + "epoch": 1.5856971634039154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 3969 + }, + { + "epoch": 1.5860966839792248, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 3970 + }, + { + "epoch": 1.5864962045545346, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 3971 + }, + { + "epoch": 1.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 3972 + }, + { + "epoch": 1.5872952457051537, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 3973 + }, + { + "epoch": 1.5876947662804635, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 3974 + }, + { + "epoch": 1.588094286855773, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 3975 + }, + { + "epoch": 1.5884938074310826, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 3976 + }, + { + "epoch": 1.5888933280063924, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 3977 + }, + { + "epoch": 1.589292848581702, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 3978 + }, + { + "epoch": 1.5896923691570115, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 3979 + }, + { + "epoch": 1.5900918897323213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 3980 + }, + { + "epoch": 1.5904914103076309, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 3981 + }, + { + "epoch": 1.5908909308829404, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 3982 + }, + { + "epoch": 1.5912904514582502, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 3983 + }, + { + "epoch": 1.5916899720335598, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 3984 + }, + { + "epoch": 1.5920894926088693, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 3985 + }, + { + "epoch": 1.5924890131841791, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 3986 + }, + { + "epoch": 1.5928885337594885, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 3987 + }, + { + "epoch": 1.5932880543347983, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 3988 + }, + { + "epoch": 1.5936875749101078, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 3989 + }, + { + "epoch": 1.5940870954854174, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 3990 + }, + { + "epoch": 1.5944866160607272, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 3991 + }, + { + "epoch": 1.5948861366360367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 3992 + }, + { + "epoch": 1.5952856572113463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 3993 + }, + { + "epoch": 1.595685177786656, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 3994 + }, + { + "epoch": 1.5960846983619656, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3459, + "step": 3995 + }, + { + "epoch": 1.5964842189372752, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 3996 + }, + { + "epoch": 1.596883739512585, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.354, + "step": 3997 + }, + { + "epoch": 1.5972832600878946, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 3998 + }, + { + "epoch": 1.5976827806632041, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 3999 + }, + { + "epoch": 1.598082301238514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4000 + }, + { + "epoch": 1.5984818218138233, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4001 + }, + { + "epoch": 1.598881342389133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4002 + }, + { + "epoch": 1.5992808629644428, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4003 + }, + { + "epoch": 1.5996803835397522, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4004 + }, + { + "epoch": 1.600079904115062, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4005 + }, + { + "epoch": 1.6004794246903715, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4006 + }, + { + "epoch": 1.600878945265681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4007 + }, + { + "epoch": 1.6012784658409909, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4008 + }, + { + "epoch": 1.6016779864163004, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4009 + }, + { + "epoch": 1.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4010 + }, + { + "epoch": 1.6024770275669198, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4011 + }, + { + "epoch": 1.6028765481422294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4012 + }, + { + "epoch": 1.603276068717539, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4013 + }, + { + "epoch": 1.6036755892928487, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4014 + }, + { + "epoch": 1.604075109868158, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4015 + }, + { + "epoch": 1.6044746304434678, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3525, + "step": 4016 + }, + { + "epoch": 1.6048741510187776, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4017 + }, + { + "epoch": 1.605273671594087, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4018 + }, + { + "epoch": 1.6056731921693967, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4019 + }, + { + "epoch": 1.6060727127447063, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4020 + }, + { + "epoch": 1.6064722333200159, + "grad_norm": 0.2158203125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4021 + }, + { + "epoch": 1.6068717538953257, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4022 + }, + { + "epoch": 1.6072712744706352, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4023 + }, + { + "epoch": 1.6076707950459448, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4024 + }, + { + "epoch": 1.6080703156212546, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4025 + }, + { + "epoch": 1.6084698361965641, + "grad_norm": 0.185546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4026 + }, + { + "epoch": 1.6088693567718737, + "grad_norm": 0.212890625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4027 + }, + { + "epoch": 1.6092688773471835, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4028 + }, + { + "epoch": 1.609668397922493, + "grad_norm": 0.24609375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4029 + }, + { + "epoch": 1.6100679184978026, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4030 + }, + { + "epoch": 1.6104674390731124, + "grad_norm": 0.44140625, + "learning_rate": 0.0002, + "loss": 1.353, + "step": 4031 + }, + { + "epoch": 1.6108669596484217, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4032 + }, + { + "epoch": 1.6112664802237315, + "grad_norm": 0.3125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4033 + }, + { + "epoch": 1.611666000799041, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 4034 + }, + { + "epoch": 1.6120655213743507, + "grad_norm": 0.41015625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4035 + }, + { + "epoch": 1.6124650419496604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4036 + }, + { + "epoch": 1.61286456252497, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4037 + }, + { + "epoch": 1.6132640831002796, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3487, + "step": 4038 + }, + { + "epoch": 1.6136636036755894, + "grad_norm": 0.2353515625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4039 + }, + { + "epoch": 1.614063124250899, + "grad_norm": 0.1884765625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4040 + }, + { + "epoch": 1.6144626448262085, + "grad_norm": 0.2109375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4041 + }, + { + "epoch": 1.6148621654015183, + "grad_norm": 0.224609375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4042 + }, + { + "epoch": 1.6152616859768278, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4043 + }, + { + "epoch": 1.6156612065521374, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4044 + }, + { + "epoch": 1.6160607271274472, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4045 + }, + { + "epoch": 1.6164602477027565, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4046 + }, + { + "epoch": 1.6168597682780663, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4047 + }, + { + "epoch": 1.617259288853376, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4048 + }, + { + "epoch": 1.6176588094286855, + "grad_norm": 0.203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4049 + }, + { + "epoch": 1.6180583300039952, + "grad_norm": 0.2314453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4050 + }, + { + "epoch": 1.6184578505793048, + "grad_norm": 0.1982421875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4051 + }, + { + "epoch": 1.6188573711546144, + "grad_norm": 0.236328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4052 + }, + { + "epoch": 1.6192568917299242, + "grad_norm": 0.1865234375, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4053 + }, + { + "epoch": 1.6196564123052337, + "grad_norm": 0.21875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4054 + }, + { + "epoch": 1.6200559328805433, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4055 + }, + { + "epoch": 1.620455453455853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4056 + }, + { + "epoch": 1.6208549740311626, + "grad_norm": 0.1748046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4057 + }, + { + "epoch": 1.6212544946064722, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4058 + }, + { + "epoch": 1.621654015181782, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4059 + }, + { + "epoch": 1.6220535357570913, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4060 + }, + { + "epoch": 1.6224530563324011, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4061 + }, + { + "epoch": 1.622852576907711, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4062 + }, + { + "epoch": 1.6232520974830202, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4063 + }, + { + "epoch": 1.62365161805833, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4064 + }, + { + "epoch": 1.6240511386336396, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4065 + }, + { + "epoch": 1.6244506592089492, + "grad_norm": 0.1962890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 4066 + }, + { + "epoch": 1.624850179784259, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4067 + }, + { + "epoch": 1.6252497003595685, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4068 + }, + { + "epoch": 1.625649220934878, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4069 + }, + { + "epoch": 1.6260487415101879, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4070 + }, + { + "epoch": 1.6264482620854974, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4071 + }, + { + "epoch": 1.626847782660807, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4072 + }, + { + "epoch": 1.6272473032361168, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4073 + }, + { + "epoch": 1.6276468238114263, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4074 + }, + { + "epoch": 1.628046344386736, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3489, + "step": 4075 + }, + { + "epoch": 1.6284458649620457, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4076 + }, + { + "epoch": 1.628845385537355, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4077 + }, + { + "epoch": 1.6292449061126648, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4078 + }, + { + "epoch": 1.6296444266879744, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.349, + "step": 4079 + }, + { + "epoch": 1.630043947263284, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4080 + }, + { + "epoch": 1.6304434678385937, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4081 + }, + { + "epoch": 1.6308429884139033, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4082 + }, + { + "epoch": 1.6312425089892129, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4083 + }, + { + "epoch": 1.6316420295645226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4084 + }, + { + "epoch": 1.6320415501398322, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4085 + }, + { + "epoch": 1.6324410707151418, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4086 + }, + { + "epoch": 1.6328405912904516, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4087 + }, + { + "epoch": 1.6332401118657611, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4088 + }, + { + "epoch": 1.6336396324410707, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4089 + }, + { + "epoch": 1.6340391530163805, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4090 + }, + { + "epoch": 1.6344386735916898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4091 + }, + { + "epoch": 1.6348381941669996, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4092 + }, + { + "epoch": 1.6352377147423094, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4093 + }, + { + "epoch": 1.6356372353176187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4094 + }, + { + "epoch": 1.6360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4095 + }, + { + "epoch": 1.636436276468238, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4096 + }, + { + "epoch": 1.6368357970435476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3508, + "step": 4097 + }, + { + "epoch": 1.6372353176188574, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4098 + }, + { + "epoch": 1.637634838194167, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4099 + }, + { + "epoch": 1.6380343587694766, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 4100 + }, + { + "epoch": 1.6384338793447863, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4101 + }, + { + "epoch": 1.638833399920096, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4102 + }, + { + "epoch": 1.6392329204954055, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 4103 + }, + { + "epoch": 1.6396324410707153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4104 + }, + { + "epoch": 1.6400319616460246, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4105 + }, + { + "epoch": 1.6404314822213344, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4106 + }, + { + "epoch": 1.6408310027966442, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4107 + }, + { + "epoch": 1.6412305233719535, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4108 + }, + { + "epoch": 1.6416300439472633, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4109 + }, + { + "epoch": 1.6420295645225729, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4110 + }, + { + "epoch": 1.6424290850978824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 4111 + }, + { + "epoch": 1.6428286056731922, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4112 + }, + { + "epoch": 1.6432281262485018, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4113 + }, + { + "epoch": 1.6436276468238114, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4114 + }, + { + "epoch": 1.6440271673991211, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4115 + }, + { + "epoch": 1.6444266879744307, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4116 + }, + { + "epoch": 1.6448262085497403, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4117 + }, + { + "epoch": 1.64522572912505, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4118 + }, + { + "epoch": 1.6456252497003596, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4119 + }, + { + "epoch": 1.6460247702756692, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4120 + }, + { + "epoch": 1.646424290850979, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4121 + }, + { + "epoch": 1.6468238114262883, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4122 + }, + { + "epoch": 1.647223332001598, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4123 + }, + { + "epoch": 1.6476228525769077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4124 + }, + { + "epoch": 1.6480223731522172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4125 + }, + { + "epoch": 1.648421893727527, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4126 + }, + { + "epoch": 1.6488214143028366, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4127 + }, + { + "epoch": 1.6492209348781461, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4128 + }, + { + "epoch": 1.649620455453456, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4129 + }, + { + "epoch": 1.6500199760287655, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4130 + }, + { + "epoch": 1.650419496604075, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 4131 + }, + { + "epoch": 1.6508190171793848, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4132 + }, + { + "epoch": 1.6512185377546944, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4133 + }, + { + "epoch": 1.651618058330004, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4134 + }, + { + "epoch": 1.6520175789053138, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4135 + }, + { + "epoch": 1.652417099480623, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3565, + "step": 4136 + }, + { + "epoch": 1.6528166200559329, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4137 + }, + { + "epoch": 1.6532161406312427, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4138 + }, + { + "epoch": 1.653615661206552, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4139 + }, + { + "epoch": 1.6540151817818618, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4140 + }, + { + "epoch": 1.6544147023571714, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4141 + }, + { + "epoch": 1.654814222932481, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4142 + }, + { + "epoch": 1.6552137435077907, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4143 + }, + { + "epoch": 1.6556132640831003, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4144 + }, + { + "epoch": 1.6560127846584098, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4145 + }, + { + "epoch": 1.6564123052337196, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4146 + }, + { + "epoch": 1.6568118258090292, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4147 + }, + { + "epoch": 1.6572113463843388, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4148 + }, + { + "epoch": 1.6576108669596485, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4149 + }, + { + "epoch": 1.6580103875349579, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4150 + }, + { + "epoch": 1.6584099081102677, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4151 + }, + { + "epoch": 1.6588094286855775, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4152 + }, + { + "epoch": 1.6592089492608868, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4153 + }, + { + "epoch": 1.6596084698361966, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4154 + }, + { + "epoch": 1.6600079904115062, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4155 + }, + { + "epoch": 1.6604075109868157, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4156 + }, + { + "epoch": 1.6608070315621255, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4157 + }, + { + "epoch": 1.661206552137435, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4158 + }, + { + "epoch": 1.6616060727127446, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4159 + }, + { + "epoch": 1.6620055932880544, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4160 + }, + { + "epoch": 1.662405113863364, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4161 + }, + { + "epoch": 1.6628046344386735, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4162 + }, + { + "epoch": 1.6632041550139833, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4163 + }, + { + "epoch": 1.663603675589293, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4164 + }, + { + "epoch": 1.6640031961646025, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4165 + }, + { + "epoch": 1.6644027167399122, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4166 + }, + { + "epoch": 1.6648022373152216, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4167 + }, + { + "epoch": 1.6652017578905314, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4168 + }, + { + "epoch": 1.665601278465841, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 4169 + }, + { + "epoch": 1.6660007990411505, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4170 + }, + { + "epoch": 1.6664003196164603, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4171 + }, + { + "epoch": 1.6667998401917699, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 4172 + }, + { + "epoch": 1.6671993607670794, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4173 + }, + { + "epoch": 1.6675988813423892, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4174 + }, + { + "epoch": 1.6679984019176988, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4175 + }, + { + "epoch": 1.6683979224930083, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 4176 + }, + { + "epoch": 1.6687974430683181, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4177 + }, + { + "epoch": 1.6691969636436277, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4178 + }, + { + "epoch": 1.6695964842189372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4179 + }, + { + "epoch": 1.669996004794247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 4180 + }, + { + "epoch": 1.6703955253695564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4181 + }, + { + "epoch": 1.6707950459448662, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4182 + }, + { + "epoch": 1.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4183 + }, + { + "epoch": 1.6715940870954853, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4184 + }, + { + "epoch": 1.671993607670795, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4185 + }, + { + "epoch": 1.6723931282461046, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.346, + "step": 4186 + }, + { + "epoch": 1.6727926488214142, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 4187 + }, + { + "epoch": 1.673192169396724, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4188 + }, + { + "epoch": 1.6735916899720336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4189 + }, + { + "epoch": 1.6739912105473431, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4190 + }, + { + "epoch": 1.674390731122653, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4191 + }, + { + "epoch": 1.6747902516979625, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4192 + }, + { + "epoch": 1.675189772273272, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4193 + }, + { + "epoch": 1.6755892928485818, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4194 + }, + { + "epoch": 1.6759888134238912, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4195 + }, + { + "epoch": 1.676388333999201, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4196 + }, + { + "epoch": 1.6767878545745107, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4197 + }, + { + "epoch": 1.67718737514982, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4198 + }, + { + "epoch": 1.6775868957251299, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4199 + }, + { + "epoch": 1.6779864163004394, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4200 + }, + { + "epoch": 1.678385936875749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 4201 + }, + { + "epoch": 1.6787854574510588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4202 + }, + { + "epoch": 1.6791849780263683, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4203 + }, + { + "epoch": 1.679584498601678, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3509, + "step": 4204 + }, + { + "epoch": 1.6799840191769877, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4205 + }, + { + "epoch": 1.6803835397522973, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4206 + }, + { + "epoch": 1.6807830603276068, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4207 + }, + { + "epoch": 1.6811825809029166, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 4208 + }, + { + "epoch": 1.6815821014782262, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4209 + }, + { + "epoch": 1.6819816220535357, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4210 + }, + { + "epoch": 1.6823811426288455, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4211 + }, + { + "epoch": 1.6827806632041549, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4212 + }, + { + "epoch": 1.6831801837794647, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3488, + "step": 4213 + }, + { + "epoch": 1.6835797043547742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4214 + }, + { + "epoch": 1.6839792249300838, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3513, + "step": 4215 + }, + { + "epoch": 1.6843787455053936, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4216 + }, + { + "epoch": 1.6847782660807031, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3516, + "step": 4217 + }, + { + "epoch": 1.6851777866560127, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4218 + }, + { + "epoch": 1.6855773072313225, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4219 + }, + { + "epoch": 1.685976827806632, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.344, + "step": 4220 + }, + { + "epoch": 1.6863763483819416, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 4221 + }, + { + "epoch": 1.6867758689572514, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4222 + }, + { + "epoch": 1.687175389532561, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4223 + }, + { + "epoch": 1.6875749101078705, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4224 + }, + { + "epoch": 1.6879744306831803, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4225 + }, + { + "epoch": 1.6883739512584897, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4226 + }, + { + "epoch": 1.6887734718337994, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4227 + }, + { + "epoch": 1.6891729924091092, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4228 + }, + { + "epoch": 1.6895725129844186, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4229 + }, + { + "epoch": 1.6899720335597284, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4230 + }, + { + "epoch": 1.690371554135038, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4231 + }, + { + "epoch": 1.6907710747103475, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4232 + }, + { + "epoch": 1.6911705952856573, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4233 + }, + { + "epoch": 1.6915701158609668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4234 + }, + { + "epoch": 1.6919696364362764, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4235 + }, + { + "epoch": 1.6923691570115862, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4236 + }, + { + "epoch": 1.6927686775868958, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4237 + }, + { + "epoch": 1.6931681981622053, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4238 + }, + { + "epoch": 1.693567718737515, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4239 + }, + { + "epoch": 1.6939672393128244, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4240 + }, + { + "epoch": 1.6943667598881342, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4241 + }, + { + "epoch": 1.694766280463444, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4242 + }, + { + "epoch": 1.6951658010387534, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4243 + }, + { + "epoch": 1.6955653216140631, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4244 + }, + { + "epoch": 1.6959648421893727, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4245 + }, + { + "epoch": 1.6963643627646823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4246 + }, + { + "epoch": 1.696763883339992, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4247 + }, + { + "epoch": 1.6971634039153016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4248 + }, + { + "epoch": 1.6975629244906112, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4249 + }, + { + "epoch": 1.697962445065921, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4250 + }, + { + "epoch": 1.6983619656412305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4251 + }, + { + "epoch": 1.69876148621654, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4252 + }, + { + "epoch": 1.69916100679185, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4253 + }, + { + "epoch": 1.6995605273671595, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 4254 + }, + { + "epoch": 1.699960047942469, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4255 + }, + { + "epoch": 1.7003595685177788, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4256 + }, + { + "epoch": 1.7007590890930881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4257 + }, + { + "epoch": 1.701158609668398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4258 + }, + { + "epoch": 1.7015581302437075, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4259 + }, + { + "epoch": 1.701957650819017, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4260 + }, + { + "epoch": 1.7023571713943269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4261 + }, + { + "epoch": 1.7027566919696364, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4262 + }, + { + "epoch": 1.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4263 + }, + { + "epoch": 1.7035557331202558, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4264 + }, + { + "epoch": 1.7039552536955653, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4265 + }, + { + "epoch": 1.704354774270875, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4266 + }, + { + "epoch": 1.7047542948461847, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4267 + }, + { + "epoch": 1.7051538154214942, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4268 + }, + { + "epoch": 1.7055533359968038, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4269 + }, + { + "epoch": 1.7059528565721136, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4270 + }, + { + "epoch": 1.706352377147423, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4271 + }, + { + "epoch": 1.7067518977227327, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4272 + }, + { + "epoch": 1.7071514182980425, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4273 + }, + { + "epoch": 1.7075509388733519, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4274 + }, + { + "epoch": 1.7079504594486616, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4275 + }, + { + "epoch": 1.7083499800239712, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4276 + }, + { + "epoch": 1.7087495005992808, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4277 + }, + { + "epoch": 1.7091490211745906, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4278 + }, + { + "epoch": 1.7095485417499001, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4279 + }, + { + "epoch": 1.7099480623252097, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4280 + }, + { + "epoch": 1.7103475829005195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4281 + }, + { + "epoch": 1.710747103475829, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4282 + }, + { + "epoch": 1.7111466240511386, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4283 + }, + { + "epoch": 1.7115461446264484, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4284 + }, + { + "epoch": 1.7119456652017577, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4285 + }, + { + "epoch": 1.7123451857770675, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 4286 + }, + { + "epoch": 1.7127447063523773, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4287 + }, + { + "epoch": 1.7131442269276866, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4288 + }, + { + "epoch": 1.7135437475029964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4289 + }, + { + "epoch": 1.713943268078306, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4290 + }, + { + "epoch": 1.7143427886536156, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4291 + }, + { + "epoch": 1.7147423092289253, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4292 + }, + { + "epoch": 1.715141829804235, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4293 + }, + { + "epoch": 1.7155413503795445, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4294 + }, + { + "epoch": 1.7159408709548543, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4295 + }, + { + "epoch": 1.7163403915301638, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4296 + }, + { + "epoch": 1.7167399121054734, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4297 + }, + { + "epoch": 1.7171394326807832, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4298 + }, + { + "epoch": 1.7175389532560927, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4299 + }, + { + "epoch": 1.7179384738314023, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3512, + "step": 4300 + }, + { + "epoch": 1.718337994406712, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4301 + }, + { + "epoch": 1.7187375149820214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 4302 + }, + { + "epoch": 1.7191370355573312, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4303 + }, + { + "epoch": 1.7195365561326408, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4304 + }, + { + "epoch": 1.7199360767079503, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4305 + }, + { + "epoch": 1.7203355972832601, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4306 + }, + { + "epoch": 1.7207351178585697, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4307 + }, + { + "epoch": 1.7211346384338793, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 4308 + }, + { + "epoch": 1.721534159009189, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4309 + }, + { + "epoch": 1.7219336795844986, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4310 + }, + { + "epoch": 1.7223332001598082, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4311 + }, + { + "epoch": 1.722732720735118, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4312 + }, + { + "epoch": 1.7231322413104275, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4313 + }, + { + "epoch": 1.723531761885737, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 4314 + }, + { + "epoch": 1.7239312824610469, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4315 + }, + { + "epoch": 1.7243308030363562, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4316 + }, + { + "epoch": 1.724730323611666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4317 + }, + { + "epoch": 1.7251298441869758, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4318 + }, + { + "epoch": 1.7255293647622851, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4319 + }, + { + "epoch": 1.725928885337595, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4320 + }, + { + "epoch": 1.7263284059129045, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4321 + }, + { + "epoch": 1.726727926488214, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4322 + }, + { + "epoch": 1.7271274470635238, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4323 + }, + { + "epoch": 1.7275269676388334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4324 + }, + { + "epoch": 1.727926488214143, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4325 + }, + { + "epoch": 1.7283260087894528, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4326 + }, + { + "epoch": 1.7287255293647623, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4327 + }, + { + "epoch": 1.7291250499400719, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4328 + }, + { + "epoch": 1.7295245705153817, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4329 + }, + { + "epoch": 1.729924091090691, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4330 + }, + { + "epoch": 1.7303236116660008, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 4331 + }, + { + "epoch": 1.7307231322413106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4332 + }, + { + "epoch": 1.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4333 + }, + { + "epoch": 1.7315221733919297, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4334 + }, + { + "epoch": 1.7319216939672393, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4335 + }, + { + "epoch": 1.7323212145425488, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4336 + }, + { + "epoch": 1.7327207351178586, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 4337 + }, + { + "epoch": 1.7331202556931682, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4338 + }, + { + "epoch": 1.7335197762684778, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4339 + }, + { + "epoch": 1.7339192968437875, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4340 + }, + { + "epoch": 1.734318817419097, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4341 + }, + { + "epoch": 1.7347183379944067, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4342 + }, + { + "epoch": 1.7351178585697165, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4343 + }, + { + "epoch": 1.735517379145026, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4344 + }, + { + "epoch": 1.7359168997203356, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4345 + }, + { + "epoch": 1.7363164202956454, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4346 + }, + { + "epoch": 1.7367159408709547, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4347 + }, + { + "epoch": 1.7371154614462645, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4348 + }, + { + "epoch": 1.737514982021574, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4349 + }, + { + "epoch": 1.7379145025968836, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4350 + }, + { + "epoch": 1.7383140231721934, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4351 + }, + { + "epoch": 1.738713543747503, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4352 + }, + { + "epoch": 1.7391130643228125, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4353 + }, + { + "epoch": 1.7395125848981223, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4354 + }, + { + "epoch": 1.739912105473432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4355 + }, + { + "epoch": 1.7403116260487415, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4356 + }, + { + "epoch": 1.7407111466240512, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 4357 + }, + { + "epoch": 1.7411106671993608, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4358 + }, + { + "epoch": 1.7415101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4359 + }, + { + "epoch": 1.7419097083499802, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4360 + }, + { + "epoch": 1.7423092289252895, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4361 + }, + { + "epoch": 1.7427087495005993, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4362 + }, + { + "epoch": 1.743108270075909, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4363 + }, + { + "epoch": 1.7435077906512184, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4364 + }, + { + "epoch": 1.7439073112265282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4365 + }, + { + "epoch": 1.7443068318018378, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4366 + }, + { + "epoch": 1.7447063523771473, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4367 + }, + { + "epoch": 1.7451058729524571, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4368 + }, + { + "epoch": 1.7455053935277667, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4369 + }, + { + "epoch": 1.7459049141030762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4370 + }, + { + "epoch": 1.746304434678386, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4371 + }, + { + "epoch": 1.7467039552536956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4372 + }, + { + "epoch": 1.7471034758290052, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4373 + }, + { + "epoch": 1.747502996404315, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4374 + }, + { + "epoch": 1.7479025169796243, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4375 + }, + { + "epoch": 1.748302037554934, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4376 + }, + { + "epoch": 1.7487015581302439, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4377 + }, + { + "epoch": 1.7491010787055532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4378 + }, + { + "epoch": 1.749500599280863, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4379 + }, + { + "epoch": 1.7499001198561726, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4380 + }, + { + "epoch": 1.7502996404314821, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4381 + }, + { + "epoch": 1.750699161006792, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4382 + }, + { + "epoch": 1.7510986815821015, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4383 + }, + { + "epoch": 1.751498202157411, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4384 + }, + { + "epoch": 1.7518977227327208, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4385 + }, + { + "epoch": 1.7522972433080304, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4386 + }, + { + "epoch": 1.75269676388334, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4387 + }, + { + "epoch": 1.7530962844586497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4388 + }, + { + "epoch": 1.7534958050339593, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4389 + }, + { + "epoch": 1.7538953256092689, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4390 + }, + { + "epoch": 1.7542948461845786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4391 + }, + { + "epoch": 1.754694366759888, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4392 + }, + { + "epoch": 1.7550938873351978, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3466, + "step": 4393 + }, + { + "epoch": 1.7554934079105073, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4394 + }, + { + "epoch": 1.755892928485817, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4395 + }, + { + "epoch": 1.7562924490611267, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4396 + }, + { + "epoch": 1.7566919696364363, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4397 + }, + { + "epoch": 1.7570914902117458, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4398 + }, + { + "epoch": 1.7574910107870556, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4399 + }, + { + "epoch": 1.7578905313623652, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 4400 + }, + { + "epoch": 1.7582900519376747, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4401 + }, + { + "epoch": 1.7586895725129845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4402 + }, + { + "epoch": 1.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4403 + }, + { + "epoch": 1.7594886136636037, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4404 + }, + { + "epoch": 1.7598881342389134, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4405 + }, + { + "epoch": 1.7602876548142228, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 4406 + }, + { + "epoch": 1.7606871753895326, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 4407 + }, + { + "epoch": 1.7610866959648424, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4408 + }, + { + "epoch": 1.7614862165401517, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 4409 + }, + { + "epoch": 1.7618857371154615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4410 + }, + { + "epoch": 1.762285257690771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4411 + }, + { + "epoch": 1.7626847782660806, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4412 + }, + { + "epoch": 1.7630842988413904, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4413 + }, + { + "epoch": 1.7634838194167, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4414 + }, + { + "epoch": 1.7638833399920095, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4415 + }, + { + "epoch": 1.7642828605673193, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4416 + }, + { + "epoch": 1.7646823811426289, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 4417 + }, + { + "epoch": 1.7650819017179384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4418 + }, + { + "epoch": 1.7654814222932482, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4419 + }, + { + "epoch": 1.7658809428685576, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 4420 + }, + { + "epoch": 1.7662804634438674, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4421 + }, + { + "epoch": 1.7666799840191771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4422 + }, + { + "epoch": 1.7670795045944865, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4423 + }, + { + "epoch": 1.7674790251697963, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4424 + }, + { + "epoch": 1.7678785457451058, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4425 + }, + { + "epoch": 1.7682780663204154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 4426 + }, + { + "epoch": 1.7686775868957252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4427 + }, + { + "epoch": 1.7690771074710347, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4428 + }, + { + "epoch": 1.7694766280463443, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4429 + }, + { + "epoch": 1.769876148621654, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4430 + }, + { + "epoch": 1.7702756691969637, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 4431 + }, + { + "epoch": 1.7706751897722732, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4432 + }, + { + "epoch": 1.771074710347583, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4433 + }, + { + "epoch": 1.7714742309228926, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4434 + }, + { + "epoch": 1.7718737514982021, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4435 + }, + { + "epoch": 1.772273272073512, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 4436 + }, + { + "epoch": 1.7726727926488213, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4437 + }, + { + "epoch": 1.773072313224131, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4438 + }, + { + "epoch": 1.7734718337994406, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4439 + }, + { + "epoch": 1.7738713543747502, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 4440 + }, + { + "epoch": 1.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4441 + }, + { + "epoch": 1.7746703955253695, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4442 + }, + { + "epoch": 1.775069916100679, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 4443 + }, + { + "epoch": 1.7754694366759889, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 4444 + }, + { + "epoch": 1.7758689572512985, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4445 + }, + { + "epoch": 1.776268477826608, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3494, + "step": 4446 + }, + { + "epoch": 1.7766679984019178, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3493, + "step": 4447 + }, + { + "epoch": 1.7770675189772274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4448 + }, + { + "epoch": 1.777467039552537, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 4449 + }, + { + "epoch": 1.7778665601278467, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 4450 + }, + { + "epoch": 1.778266080703156, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4451 + }, + { + "epoch": 1.7786656012784658, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4452 + }, + { + "epoch": 1.7790651218537756, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4453 + }, + { + "epoch": 1.779464642429085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4454 + }, + { + "epoch": 1.7798641630043948, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4455 + }, + { + "epoch": 1.7802636835797043, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4456 + }, + { + "epoch": 1.780663204155014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4457 + }, + { + "epoch": 1.7810627247303237, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4458 + }, + { + "epoch": 1.7814622453056332, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 4459 + }, + { + "epoch": 1.7818617658809428, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4460 + }, + { + "epoch": 1.7822612864562526, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4461 + }, + { + "epoch": 1.7826608070315622, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4462 + }, + { + "epoch": 1.7830603276068717, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4463 + }, + { + "epoch": 1.7834598481821815, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4464 + }, + { + "epoch": 1.7838593687574908, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4465 + }, + { + "epoch": 1.7842588893328006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4466 + }, + { + "epoch": 1.7846584099081104, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4467 + }, + { + "epoch": 1.7850579304834198, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 4468 + }, + { + "epoch": 1.7854574510587295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 4469 + }, + { + "epoch": 1.7858569716340391, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4470 + }, + { + "epoch": 1.7862564922093487, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4471 + }, + { + "epoch": 1.7866560127846585, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4472 + }, + { + "epoch": 1.787055533359968, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 4473 + }, + { + "epoch": 1.7874550539352776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4474 + }, + { + "epoch": 1.7878545745105874, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4475 + }, + { + "epoch": 1.788254095085897, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4476 + }, + { + "epoch": 1.7886536156612065, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4477 + }, + { + "epoch": 1.7890531362365163, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4478 + }, + { + "epoch": 1.7894526568118259, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4479 + }, + { + "epoch": 1.7898521773871354, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 4480 + }, + { + "epoch": 1.7902516979624452, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4481 + }, + { + "epoch": 1.7906512185377546, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3491, + "step": 4482 + }, + { + "epoch": 1.7910507391130643, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4483 + }, + { + "epoch": 1.791450259688374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4484 + }, + { + "epoch": 1.7918497802636835, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4485 + }, + { + "epoch": 1.7922493008389933, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4486 + }, + { + "epoch": 1.7926488214143028, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4487 + }, + { + "epoch": 1.7930483419896124, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4488 + }, + { + "epoch": 1.7934478625649222, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4489 + }, + { + "epoch": 1.7938473831402317, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4490 + }, + { + "epoch": 1.7942469037155413, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4491 + }, + { + "epoch": 1.794646424290851, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4492 + }, + { + "epoch": 1.7950459448661606, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4493 + }, + { + "epoch": 1.7954454654414702, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4494 + }, + { + "epoch": 1.79584498601678, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4495 + }, + { + "epoch": 1.7962445065920893, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4496 + }, + { + "epoch": 1.7966440271673991, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4497 + }, + { + "epoch": 1.797043547742709, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4498 + }, + { + "epoch": 1.7974430683180183, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4499 + }, + { + "epoch": 1.797842588893328, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4500 + }, + { + "epoch": 1.7982421094686376, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4501 + }, + { + "epoch": 1.7986416300439472, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4502 + }, + { + "epoch": 1.799041150619257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4503 + }, + { + "epoch": 1.7994406711945665, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4504 + }, + { + "epoch": 1.799840191769876, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 4505 + }, + { + "epoch": 1.8002397123451859, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4506 + }, + { + "epoch": 1.8006392329204954, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 4507 + }, + { + "epoch": 1.801038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 4508 + }, + { + "epoch": 1.8014382740711148, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4509 + }, + { + "epoch": 1.8018377946464241, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4510 + }, + { + "epoch": 1.802237315221734, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4511 + }, + { + "epoch": 1.8026368357970437, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4512 + }, + { + "epoch": 1.803036356372353, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 4513 + }, + { + "epoch": 1.8034358769476628, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 4514 + }, + { + "epoch": 1.8038353975229724, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4515 + }, + { + "epoch": 1.804234918098282, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3485, + "step": 4516 + }, + { + "epoch": 1.8046344386735917, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4517 + }, + { + "epoch": 1.8050339592489013, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4518 + }, + { + "epoch": 1.8054334798242109, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4519 + }, + { + "epoch": 1.8058330003995207, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4520 + }, + { + "epoch": 1.8062325209748302, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4521 + }, + { + "epoch": 1.8066320415501398, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 4522 + }, + { + "epoch": 1.8070315621254496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4523 + }, + { + "epoch": 1.8074310827007591, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4524 + }, + { + "epoch": 1.8078306032760687, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4525 + }, + { + "epoch": 1.8082301238513785, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4526 + }, + { + "epoch": 1.8086296444266878, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4527 + }, + { + "epoch": 1.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4528 + }, + { + "epoch": 1.8094286855773072, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 4529 + }, + { + "epoch": 1.8098282061526167, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4530 + }, + { + "epoch": 1.8102277267279265, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4531 + }, + { + "epoch": 1.810627247303236, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4532 + }, + { + "epoch": 1.8110267678785457, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4533 + }, + { + "epoch": 1.8114262884538554, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4534 + }, + { + "epoch": 1.811825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 4535 + }, + { + "epoch": 1.8122253296044746, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4536 + }, + { + "epoch": 1.8126248501797844, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4537 + }, + { + "epoch": 1.813024370755094, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 4538 + }, + { + "epoch": 1.8134238913304035, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4539 + }, + { + "epoch": 1.8138234119057133, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 4540 + }, + { + "epoch": 1.8142229324810226, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3465, + "step": 4541 + }, + { + "epoch": 1.8146224530563324, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4542 + }, + { + "epoch": 1.8150219736316422, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4543 + }, + { + "epoch": 1.8154214942069515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4544 + }, + { + "epoch": 1.8158210147822613, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 4545 + }, + { + "epoch": 1.8162205353575709, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 4546 + }, + { + "epoch": 1.8166200559328805, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4547 + }, + { + "epoch": 1.8170195765081902, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 4548 + }, + { + "epoch": 1.8174190970834998, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4549 + }, + { + "epoch": 1.8178186176588094, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4550 + }, + { + "epoch": 1.8182181382341192, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 4551 + }, + { + "epoch": 1.8186176588094287, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4552 + }, + { + "epoch": 1.8190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4553 + }, + { + "epoch": 1.819416699960048, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4554 + }, + { + "epoch": 1.8198162205353574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 4555 + }, + { + "epoch": 1.8202157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 4556 + }, + { + "epoch": 1.820615261685977, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 4557 + }, + { + "epoch": 1.8210147822612863, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4558 + }, + { + "epoch": 1.821414302836596, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4559 + }, + { + "epoch": 1.8218138234119057, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4560 + }, + { + "epoch": 1.8222133439872152, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4561 + }, + { + "epoch": 1.822612864562525, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4562 + }, + { + "epoch": 1.8230123851378346, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4563 + }, + { + "epoch": 1.8234119057131442, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4564 + }, + { + "epoch": 1.823811426288454, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4565 + }, + { + "epoch": 1.8242109468637635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4566 + }, + { + "epoch": 1.824610467439073, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 4567 + }, + { + "epoch": 1.8250099880143829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4568 + }, + { + "epoch": 1.8254095085896924, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4569 + }, + { + "epoch": 1.825809029165002, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4570 + }, + { + "epoch": 1.8262085497403118, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4571 + }, + { + "epoch": 1.8266080703156211, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4572 + }, + { + "epoch": 1.827007590890931, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4573 + }, + { + "epoch": 1.8274071114662405, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4574 + }, + { + "epoch": 1.82780663204155, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4575 + }, + { + "epoch": 1.8282061526168598, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4576 + }, + { + "epoch": 1.8286056731921694, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4577 + }, + { + "epoch": 1.829005193767479, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4578 + }, + { + "epoch": 1.8294047143427887, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4579 + }, + { + "epoch": 1.8298042349180983, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 4580 + }, + { + "epoch": 1.8302037554934079, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4581 + }, + { + "epoch": 1.8306032760687176, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4582 + }, + { + "epoch": 1.8310027966440272, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4583 + }, + { + "epoch": 1.8314023172193368, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4584 + }, + { + "epoch": 1.8318018377946466, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4585 + }, + { + "epoch": 1.832201358369956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4586 + }, + { + "epoch": 1.8326008789452657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4587 + }, + { + "epoch": 1.8330003995205755, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4588 + }, + { + "epoch": 1.8333999200958848, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4589 + }, + { + "epoch": 1.8337994406711946, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4590 + }, + { + "epoch": 1.8341989612465042, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4591 + }, + { + "epoch": 1.8345984818218137, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4592 + }, + { + "epoch": 1.8349980023971235, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 4593 + }, + { + "epoch": 1.835397522972433, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4594 + }, + { + "epoch": 1.8357970435477426, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4595 + }, + { + "epoch": 1.8361965641230524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 4596 + }, + { + "epoch": 1.836596084698362, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4597 + }, + { + "epoch": 1.8369956052736716, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 4598 + }, + { + "epoch": 1.8373951258489813, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4599 + }, + { + "epoch": 1.8377946464242907, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4600 + }, + { + "epoch": 1.8381941669996005, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4601 + }, + { + "epoch": 1.8385936875749103, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4602 + }, + { + "epoch": 1.8389932081502196, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4603 + }, + { + "epoch": 1.8393927287255294, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4604 + }, + { + "epoch": 1.839792249300839, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 4605 + }, + { + "epoch": 1.8401917698761485, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 4606 + }, + { + "epoch": 1.8405912904514583, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4607 + }, + { + "epoch": 1.8409908110267679, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 4608 + }, + { + "epoch": 1.8413903316020774, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4609 + }, + { + "epoch": 1.8417898521773872, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4610 + }, + { + "epoch": 1.8421893727526968, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4611 + }, + { + "epoch": 1.8425888933280063, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4612 + }, + { + "epoch": 1.8429884139033161, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4613 + }, + { + "epoch": 1.8433879344786257, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4614 + }, + { + "epoch": 1.8437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4615 + }, + { + "epoch": 1.844186975629245, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 4616 + }, + { + "epoch": 1.8445864962045544, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4617 + }, + { + "epoch": 1.8449860167798642, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4618 + }, + { + "epoch": 1.8453855373551737, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 4619 + }, + { + "epoch": 1.8457850579304833, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4620 + }, + { + "epoch": 1.846184578505793, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4621 + }, + { + "epoch": 1.8465840990811027, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 4622 + }, + { + "epoch": 1.8469836196564122, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4623 + }, + { + "epoch": 1.847383140231722, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4624 + }, + { + "epoch": 1.8477826608070316, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4625 + }, + { + "epoch": 1.8481821813823411, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4626 + }, + { + "epoch": 1.848581701957651, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4627 + }, + { + "epoch": 1.8489812225329605, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4628 + }, + { + "epoch": 1.84938074310827, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 4629 + }, + { + "epoch": 1.8497802636835798, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 4630 + }, + { + "epoch": 1.8501797842588892, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4631 + }, + { + "epoch": 1.850579304834199, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4632 + }, + { + "epoch": 1.8509788254095088, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 4633 + }, + { + "epoch": 1.851378345984818, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4634 + }, + { + "epoch": 1.8517778665601279, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4635 + }, + { + "epoch": 1.8521773871354374, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 4636 + }, + { + "epoch": 1.852576907710747, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 4637 + }, + { + "epoch": 1.8529764282860568, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4638 + }, + { + "epoch": 1.8533759488613664, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4639 + }, + { + "epoch": 1.853775469436676, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4640 + }, + { + "epoch": 1.8541749900119857, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4641 + }, + { + "epoch": 1.8545745105872953, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 4642 + }, + { + "epoch": 1.8549740311626048, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4643 + }, + { + "epoch": 1.8553735517379146, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4644 + }, + { + "epoch": 1.855773072313224, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4645 + }, + { + "epoch": 1.8561725928885338, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4646 + }, + { + "epoch": 1.8565721134638435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4647 + }, + { + "epoch": 1.8569716340391529, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4648 + }, + { + "epoch": 1.8573711546144627, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4649 + }, + { + "epoch": 1.8577706751897722, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4650 + }, + { + "epoch": 1.8581701957650818, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4651 + }, + { + "epoch": 1.8585697163403916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4652 + }, + { + "epoch": 1.8589692369157012, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4653 + }, + { + "epoch": 1.8593687574910107, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4654 + }, + { + "epoch": 1.8597682780663205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 4655 + }, + { + "epoch": 1.86016779864163, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4656 + }, + { + "epoch": 1.8605673192169396, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4657 + }, + { + "epoch": 1.8609668397922494, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4658 + }, + { + "epoch": 1.861366360367559, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4659 + }, + { + "epoch": 1.8617658809428685, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 4660 + }, + { + "epoch": 1.8621654015181783, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 4661 + }, + { + "epoch": 1.8625649220934877, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4662 + }, + { + "epoch": 1.8629644426687975, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4663 + }, + { + "epoch": 1.863363963244107, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3447, + "step": 4664 + }, + { + "epoch": 1.8637634838194166, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 4665 + }, + { + "epoch": 1.8641630043947264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 4666 + }, + { + "epoch": 1.864562524970036, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 4667 + }, + { + "epoch": 1.8649620455453455, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4668 + }, + { + "epoch": 1.8653615661206553, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4669 + }, + { + "epoch": 1.8657610866959649, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4670 + }, + { + "epoch": 1.8661606072712744, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4671 + }, + { + "epoch": 1.8665601278465842, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4672 + }, + { + "epoch": 1.8669596484218938, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4673 + }, + { + "epoch": 1.8673591689972033, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4674 + }, + { + "epoch": 1.8677586895725131, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4675 + }, + { + "epoch": 1.8681582101478225, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4676 + }, + { + "epoch": 1.8685577307231322, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4677 + }, + { + "epoch": 1.868957251298442, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4678 + }, + { + "epoch": 1.8693567718737514, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4679 + }, + { + "epoch": 1.8697562924490612, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4680 + }, + { + "epoch": 1.8701558130243707, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 4681 + }, + { + "epoch": 1.8705553335996803, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4682 + }, + { + "epoch": 1.87095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 4683 + }, + { + "epoch": 1.8713543747502996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4684 + }, + { + "epoch": 1.8717538953256092, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 4685 + }, + { + "epoch": 1.872153415900919, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 4686 + }, + { + "epoch": 1.8725529364762286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4687 + }, + { + "epoch": 1.8729524570515381, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4688 + }, + { + "epoch": 1.873351977626848, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 4689 + }, + { + "epoch": 1.8737514982021573, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4690 + }, + { + "epoch": 1.874151018777467, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 4691 + }, + { + "epoch": 1.8745505393527768, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4692 + }, + { + "epoch": 1.8749500599280862, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4693 + }, + { + "epoch": 1.875349580503396, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 4694 + }, + { + "epoch": 1.8757491010787055, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 4695 + }, + { + "epoch": 1.876148621654015, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4696 + }, + { + "epoch": 1.8765481422293249, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4697 + }, + { + "epoch": 1.8769476628046344, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 4698 + }, + { + "epoch": 1.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4699 + }, + { + "epoch": 1.8777467039552538, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 4700 + }, + { + "epoch": 1.8781462245305633, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4701 + }, + { + "epoch": 1.878545745105873, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4702 + }, + { + "epoch": 1.8789452656811827, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 4703 + }, + { + "epoch": 1.8793447862564923, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4704 + }, + { + "epoch": 1.8797443068318018, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4705 + }, + { + "epoch": 1.8801438274071116, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4706 + }, + { + "epoch": 1.880543347982421, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 4707 + }, + { + "epoch": 1.8809428685577307, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4708 + }, + { + "epoch": 1.8813423891330403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4709 + }, + { + "epoch": 1.8817419097083499, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4710 + }, + { + "epoch": 1.8821414302836597, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4711 + }, + { + "epoch": 1.8825409508589692, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4712 + }, + { + "epoch": 1.8829404714342788, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 4713 + }, + { + "epoch": 1.8833399920095886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4714 + }, + { + "epoch": 1.8837395125848981, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 4715 + }, + { + "epoch": 1.8841390331602077, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4716 + }, + { + "epoch": 1.8845385537355175, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4717 + }, + { + "epoch": 1.884938074310827, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 4718 + }, + { + "epoch": 1.8853375948861366, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4719 + }, + { + "epoch": 1.8857371154614464, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 4720 + }, + { + "epoch": 1.8861366360367557, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4721 + }, + { + "epoch": 1.8865361566120655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4722 + }, + { + "epoch": 1.8869356771873753, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 4723 + }, + { + "epoch": 1.8873351977626847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4724 + }, + { + "epoch": 1.8877347183379944, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4725 + }, + { + "epoch": 1.888134238913304, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4726 + }, + { + "epoch": 1.8885337594886136, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4727 + }, + { + "epoch": 1.8889332800639234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4728 + }, + { + "epoch": 1.889332800639233, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4729 + }, + { + "epoch": 1.8897323212145425, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4730 + }, + { + "epoch": 1.8901318417898523, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4731 + }, + { + "epoch": 1.8905313623651618, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4732 + }, + { + "epoch": 1.8909308829404714, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 4733 + }, + { + "epoch": 1.8913304035157812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 4734 + }, + { + "epoch": 1.8917299240910905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4735 + }, + { + "epoch": 1.8921294446664003, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 4736 + }, + { + "epoch": 1.89252896524171, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4737 + }, + { + "epoch": 1.8929284858170194, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4738 + }, + { + "epoch": 1.8933280063923292, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4739 + }, + { + "epoch": 1.8937275269676388, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 4740 + }, + { + "epoch": 1.8941270475429484, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4741 + }, + { + "epoch": 1.8945265681182581, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4742 + }, + { + "epoch": 1.8949260886935677, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4743 + }, + { + "epoch": 1.8953256092688773, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4744 + }, + { + "epoch": 1.895725129844187, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4745 + }, + { + "epoch": 1.8961246504194966, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4746 + }, + { + "epoch": 1.8965241709948062, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4747 + }, + { + "epoch": 1.896923691570116, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4748 + }, + { + "epoch": 1.8973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 4749 + }, + { + "epoch": 1.897722732720735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 4750 + }, + { + "epoch": 1.898122253296045, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4751 + }, + { + "epoch": 1.8985217738713542, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 4752 + }, + { + "epoch": 1.898921294446664, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 4753 + }, + { + "epoch": 1.8993208150219736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 4754 + }, + { + "epoch": 1.8997203355972831, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4755 + }, + { + "epoch": 1.900119856172593, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 4756 + }, + { + "epoch": 1.9005193767479025, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 4757 + }, + { + "epoch": 1.900918897323212, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4758 + }, + { + "epoch": 1.9013184178985219, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 4759 + }, + { + "epoch": 1.9017179384738314, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4760 + }, + { + "epoch": 1.902117459049141, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4761 + }, + { + "epoch": 1.9025169796244508, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 4762 + }, + { + "epoch": 1.9029165001997603, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 4763 + }, + { + "epoch": 1.90331602077507, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 4764 + }, + { + "epoch": 1.9037155413503797, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4765 + }, + { + "epoch": 1.904115061925689, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 4766 + }, + { + "epoch": 1.9045145825009988, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 4767 + }, + { + "epoch": 1.9049141030763086, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 4768 + }, + { + "epoch": 1.905313623651618, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 4769 + }, + { + "epoch": 1.9057131442269277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4770 + }, + { + "epoch": 1.9061126648022373, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4771 + }, + { + "epoch": 1.9065121853775469, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4772 + }, + { + "epoch": 1.9069117059528566, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4773 + }, + { + "epoch": 1.9073112265281662, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 4774 + }, + { + "epoch": 1.9077107471034758, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 4775 + }, + { + "epoch": 1.9081102676787856, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4776 + }, + { + "epoch": 1.9085097882540951, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4777 + }, + { + "epoch": 1.9089093088294047, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 4778 + }, + { + "epoch": 1.9093088294047145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4779 + }, + { + "epoch": 1.9097083499800238, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 4780 + }, + { + "epoch": 1.9101078705553336, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4781 + }, + { + "epoch": 1.9105073911306434, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 4782 + }, + { + "epoch": 1.9109069117059527, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4783 + }, + { + "epoch": 1.9113064322812625, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.345, + "step": 4784 + }, + { + "epoch": 1.911705952856572, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 4785 + }, + { + "epoch": 1.9121054734318816, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 4786 + }, + { + "epoch": 1.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 4787 + }, + { + "epoch": 1.912904514582501, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4788 + }, + { + "epoch": 1.9133040351578106, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4789 + }, + { + "epoch": 1.9137035557331203, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 4790 + }, + { + "epoch": 1.91410307630843, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 4791 + }, + { + "epoch": 1.9145025968837395, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4792 + }, + { + "epoch": 1.9149021174590493, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4793 + }, + { + "epoch": 1.9153016380343588, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4794 + }, + { + "epoch": 1.9157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4795 + }, + { + "epoch": 1.9161006791849782, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3421, + "step": 4796 + }, + { + "epoch": 1.9165001997602875, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 4797 + }, + { + "epoch": 1.9168997203355973, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 4798 + }, + { + "epoch": 1.9172992409109069, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 4799 + }, + { + "epoch": 1.9176987614862164, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 4800 + }, + { + "epoch": 1.9180982820615262, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 4801 + }, + { + "epoch": 1.9184978026368358, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 4802 + }, + { + "epoch": 1.9188973232121453, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 4803 + }, + { + "epoch": 1.9192968437874551, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 4804 + }, + { + "epoch": 1.9196963643627647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4805 + }, + { + "epoch": 1.9200958849380743, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4806 + }, + { + "epoch": 1.920495405513384, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 4807 + }, + { + "epoch": 1.9208949260886936, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 4808 + }, + { + "epoch": 1.9212944466640032, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4809 + }, + { + "epoch": 1.921693967239313, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 4810 + }, + { + "epoch": 1.9220934878146223, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 4811 + }, + { + "epoch": 1.922493008389932, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 4812 + }, + { + "epoch": 1.9228925289652419, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 4813 + }, + { + "epoch": 1.9232920495405512, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 4814 + }, + { + "epoch": 1.923691570115861, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3517, + "step": 4815 + }, + { + "epoch": 1.9240910906911706, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 4816 + }, + { + "epoch": 1.9244906112664801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4817 + }, + { + "epoch": 1.92489013184179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4818 + }, + { + "epoch": 1.9252896524170995, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4819 + }, + { + "epoch": 1.925689172992409, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4820 + }, + { + "epoch": 1.9260886935677188, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4821 + }, + { + "epoch": 1.9264882141430284, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 4822 + }, + { + "epoch": 1.926887734718338, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 4823 + }, + { + "epoch": 1.9272872552936477, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 4824 + }, + { + "epoch": 1.927686775868957, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4825 + }, + { + "epoch": 1.9280862964442669, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 4826 + }, + { + "epoch": 1.9284858170195767, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 4827 + }, + { + "epoch": 1.928885337594886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 4828 + }, + { + "epoch": 1.9292848581701958, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 4829 + }, + { + "epoch": 1.9296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4830 + }, + { + "epoch": 1.930083899320815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 4831 + }, + { + "epoch": 1.9304834198961247, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 4832 + }, + { + "epoch": 1.9308829404714343, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4833 + }, + { + "epoch": 1.9312824610467438, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4834 + }, + { + "epoch": 1.9316819816220536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4835 + }, + { + "epoch": 1.9320815021973632, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4836 + }, + { + "epoch": 1.9324810227726728, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 4837 + }, + { + "epoch": 1.9328805433479825, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4838 + }, + { + "epoch": 1.933280063923292, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4839 + }, + { + "epoch": 1.9336795844986017, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 4840 + }, + { + "epoch": 1.9340791050739115, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 4841 + }, + { + "epoch": 1.9344786256492208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 4842 + }, + { + "epoch": 1.9348781462245306, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4843 + }, + { + "epoch": 1.9352776667998401, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 4844 + }, + { + "epoch": 1.9356771873751497, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4845 + }, + { + "epoch": 1.9360767079504595, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 4846 + }, + { + "epoch": 1.936476228525769, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 4847 + }, + { + "epoch": 1.9368757491010786, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 4848 + }, + { + "epoch": 1.9372752696763884, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4849 + }, + { + "epoch": 1.937674790251698, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4850 + }, + { + "epoch": 1.9380743108270075, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3454, + "step": 4851 + }, + { + "epoch": 1.9384738314023173, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 4852 + }, + { + "epoch": 1.938873351977627, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 4853 + }, + { + "epoch": 1.9392728725529365, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4854 + }, + { + "epoch": 1.9396723931282462, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4855 + }, + { + "epoch": 1.9400719137035556, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 4856 + }, + { + "epoch": 1.9404714342788654, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4857 + }, + { + "epoch": 1.940870954854175, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 4858 + }, + { + "epoch": 1.9412704754294845, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 4859 + }, + { + "epoch": 1.9416699960047943, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4860 + }, + { + "epoch": 1.9420695165801038, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4861 + }, + { + "epoch": 1.9424690371554134, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4862 + }, + { + "epoch": 1.9428685577307232, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4863 + }, + { + "epoch": 1.9432680783060328, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4864 + }, + { + "epoch": 1.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4865 + }, + { + "epoch": 1.9440671194566521, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 4866 + }, + { + "epoch": 1.9444666400319617, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4867 + }, + { + "epoch": 1.9448661606072712, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 4868 + }, + { + "epoch": 1.945265681182581, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4869 + }, + { + "epoch": 1.9456652017578904, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 4870 + }, + { + "epoch": 1.9460647223332002, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4871 + }, + { + "epoch": 1.94646424290851, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4872 + }, + { + "epoch": 1.9468637634838193, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4873 + }, + { + "epoch": 1.947263284059129, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4874 + }, + { + "epoch": 1.9476628046344386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4875 + }, + { + "epoch": 1.9480623252097482, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4876 + }, + { + "epoch": 1.948461845785058, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 4877 + }, + { + "epoch": 1.9488613663603676, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 4878 + }, + { + "epoch": 1.9492608869356771, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4879 + }, + { + "epoch": 1.949660407510987, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 4880 + }, + { + "epoch": 1.9500599280862965, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 4881 + }, + { + "epoch": 1.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4882 + }, + { + "epoch": 1.9508589692369158, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 4883 + }, + { + "epoch": 1.9512584898122254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 4884 + }, + { + "epoch": 1.951658010387535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 4885 + }, + { + "epoch": 1.9520575309628447, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4886 + }, + { + "epoch": 1.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4887 + }, + { + "epoch": 1.9528565721134639, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 4888 + }, + { + "epoch": 1.9532560926887734, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4889 + }, + { + "epoch": 1.953655613264083, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3496, + "step": 4890 + }, + { + "epoch": 1.9540551338393928, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4891 + }, + { + "epoch": 1.9544546544147023, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 4892 + }, + { + "epoch": 1.954854174990012, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4893 + }, + { + "epoch": 1.9552536955653217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 4894 + }, + { + "epoch": 1.9556532161406313, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4895 + }, + { + "epoch": 1.9560527367159408, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4896 + }, + { + "epoch": 1.9564522572912506, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 4897 + }, + { + "epoch": 1.9568517778665602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 4898 + }, + { + "epoch": 1.9572512984418697, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 4899 + }, + { + "epoch": 1.9576508190171795, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4900 + }, + { + "epoch": 1.9580503395924889, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 4901 + }, + { + "epoch": 1.9584498601677987, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 4902 + }, + { + "epoch": 1.9588493807431082, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 4903 + }, + { + "epoch": 1.9592489013184178, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4904 + }, + { + "epoch": 1.9596484218937276, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 4905 + }, + { + "epoch": 1.9600479424690371, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 4906 + }, + { + "epoch": 1.9604474630443467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 4907 + }, + { + "epoch": 1.9608469836196565, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 4908 + }, + { + "epoch": 1.961246504194966, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 4909 + }, + { + "epoch": 1.9616460247702756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 4910 + }, + { + "epoch": 1.9620455453455854, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 4911 + }, + { + "epoch": 1.962445065920895, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4912 + }, + { + "epoch": 1.9628445864962045, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 4913 + }, + { + "epoch": 1.9632441070715143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 4914 + }, + { + "epoch": 1.9636436276468237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 4915 + }, + { + "epoch": 1.9640431482221334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 4916 + }, + { + "epoch": 1.9644426687974432, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 4917 + }, + { + "epoch": 1.9648421893727526, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 4918 + }, + { + "epoch": 1.9652417099480624, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 4919 + }, + { + "epoch": 1.965641230523372, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 4920 + }, + { + "epoch": 1.9660407510986815, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 4921 + }, + { + "epoch": 1.9664402716739913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 4922 + }, + { + "epoch": 1.9668397922493008, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4923 + }, + { + "epoch": 1.9672393128246104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4924 + }, + { + "epoch": 1.9676388333999202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 4925 + }, + { + "epoch": 1.9680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 4926 + }, + { + "epoch": 1.9684378745505393, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4927 + }, + { + "epoch": 1.968837395125849, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4928 + }, + { + "epoch": 1.9692369157011587, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4929 + }, + { + "epoch": 1.9696364362764682, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4930 + }, + { + "epoch": 1.970035956851778, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 4931 + }, + { + "epoch": 1.9704354774270874, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4932 + }, + { + "epoch": 1.9708349980023971, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4933 + }, + { + "epoch": 1.9712345185777067, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 4934 + }, + { + "epoch": 1.9716340391530163, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 4935 + }, + { + "epoch": 1.972033559728326, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 4936 + }, + { + "epoch": 1.9724330803036356, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4937 + }, + { + "epoch": 1.9728326008789452, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 4938 + }, + { + "epoch": 1.973232121454255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 4939 + }, + { + "epoch": 1.9736316420295645, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 4940 + }, + { + "epoch": 1.974031162604874, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3445, + "step": 4941 + }, + { + "epoch": 1.9744306831801839, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4942 + }, + { + "epoch": 1.9748302037554935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 4943 + }, + { + "epoch": 1.975229724330803, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4944 + }, + { + "epoch": 1.9756292449061128, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4945 + }, + { + "epoch": 1.9760287654814221, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 4946 + }, + { + "epoch": 1.976428286056732, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 4947 + }, + { + "epoch": 1.9768278066320415, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 4948 + }, + { + "epoch": 1.977227327207351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 4949 + }, + { + "epoch": 1.9776268477826608, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 4950 + }, + { + "epoch": 1.9780263683579704, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4951 + }, + { + "epoch": 1.97842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 4952 + }, + { + "epoch": 1.9788254095085898, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 4953 + }, + { + "epoch": 1.9792249300838993, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 4954 + }, + { + "epoch": 1.979624450659209, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4955 + }, + { + "epoch": 1.9800239712345187, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 4956 + }, + { + "epoch": 1.9804234918098282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 4957 + }, + { + "epoch": 1.9808230123851378, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 4958 + }, + { + "epoch": 1.9812225329604476, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 4959 + }, + { + "epoch": 1.981622053535757, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 4960 + }, + { + "epoch": 1.9820215741110667, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 4961 + }, + { + "epoch": 1.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 4962 + }, + { + "epoch": 1.9828206152616858, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 4963 + }, + { + "epoch": 1.9832201358369956, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4964 + }, + { + "epoch": 1.9836196564123052, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 4965 + }, + { + "epoch": 1.9840191769876148, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 4966 + }, + { + "epoch": 1.9844186975629245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 4967 + }, + { + "epoch": 1.9848182181382341, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 4968 + }, + { + "epoch": 1.9852177387135437, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 4969 + }, + { + "epoch": 1.9856172592888535, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 4970 + }, + { + "epoch": 1.986016779864163, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 4971 + }, + { + "epoch": 1.9864163004394726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4972 + }, + { + "epoch": 1.9868158210147824, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 4973 + }, + { + "epoch": 1.987215341590092, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 4974 + }, + { + "epoch": 1.9876148621654015, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 4975 + }, + { + "epoch": 1.9880143827407113, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 4976 + }, + { + "epoch": 1.9884139033160206, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 4977 + }, + { + "epoch": 1.9888134238913304, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 4978 + }, + { + "epoch": 1.98921294446664, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 4979 + }, + { + "epoch": 1.9896124650419496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 4980 + }, + { + "epoch": 1.9900119856172593, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4981 + }, + { + "epoch": 1.990411506192569, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4982 + }, + { + "epoch": 1.9908110267678785, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 4983 + }, + { + "epoch": 1.9912105473431883, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 4984 + }, + { + "epoch": 1.9916100679184978, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 4985 + }, + { + "epoch": 1.9920095884938074, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 4986 + }, + { + "epoch": 1.9924091090691172, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 4987 + }, + { + "epoch": 1.9928086296444267, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 4988 + }, + { + "epoch": 1.9932081502197363, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 4989 + }, + { + "epoch": 1.993607670795046, + "grad_norm": 0.04541015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 4990 + }, + { + "epoch": 1.9940071913703554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 4991 + }, + { + "epoch": 1.9944067119456652, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 4992 + }, + { + "epoch": 1.9948062325209748, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 4993 + }, + { + "epoch": 1.9952057530962843, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 4994 + }, + { + "epoch": 1.9956052736715941, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 4995 + }, + { + "epoch": 1.9960047942469037, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 4996 + }, + { + "epoch": 1.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 4997 + }, + { + "epoch": 1.996803835397523, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 4998 + }, + { + "epoch": 1.9972033559728326, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 4999 + }, + { + "epoch": 1.9976028765481422, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5000 + }, + { + "epoch": 1.998002397123452, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5001 + }, + { + "epoch": 1.9984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5002 + }, + { + "epoch": 1.998801438274071, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5003 + }, + { + "epoch": 1.9992009588493809, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5004 + }, + { + "epoch": 1.9996004794246902, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5005 + }, + { + "epoch": 2.0, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5006 + }, + { + "epoch": 2.00039952057531, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5007 + }, + { + "epoch": 2.000799041150619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5008 + }, + { + "epoch": 2.001198561725929, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5009 + }, + { + "epoch": 2.0015980823012387, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5010 + }, + { + "epoch": 2.001997602876548, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5011 + }, + { + "epoch": 2.002397123451858, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5012 + }, + { + "epoch": 2.0027966440271676, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5013 + }, + { + "epoch": 2.003196164602477, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 5014 + }, + { + "epoch": 2.0035956851777867, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5015 + }, + { + "epoch": 2.003995205753096, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5016 + }, + { + "epoch": 2.004394726328406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5017 + }, + { + "epoch": 2.0047942469037157, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5018 + }, + { + "epoch": 2.005193767479025, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5019 + }, + { + "epoch": 2.005593288054335, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5020 + }, + { + "epoch": 2.0059928086296446, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5021 + }, + { + "epoch": 2.006392329204954, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5022 + }, + { + "epoch": 2.0067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5023 + }, + { + "epoch": 2.0071913703555735, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 5024 + }, + { + "epoch": 2.007590890930883, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5025 + }, + { + "epoch": 2.0079904115061926, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5026 + }, + { + "epoch": 2.0083899320815024, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5027 + }, + { + "epoch": 2.0087894526568117, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3439, + "step": 5028 + }, + { + "epoch": 2.0091889732321215, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 5029 + }, + { + "epoch": 2.009588493807431, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5030 + }, + { + "epoch": 2.0099880143827407, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5031 + }, + { + "epoch": 2.0103875349580504, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5032 + }, + { + "epoch": 2.01078705553336, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5033 + }, + { + "epoch": 2.0111865761086696, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5034 + }, + { + "epoch": 2.0115860966839794, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5035 + }, + { + "epoch": 2.0119856172592887, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5036 + }, + { + "epoch": 2.0123851378345985, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5037 + }, + { + "epoch": 2.0127846584099083, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5038 + }, + { + "epoch": 2.0131841789852176, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5039 + }, + { + "epoch": 2.0135836995605274, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5040 + }, + { + "epoch": 2.013983220135837, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 5041 + }, + { + "epoch": 2.0143827407111465, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 5042 + }, + { + "epoch": 2.0147822612864563, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3437, + "step": 5043 + }, + { + "epoch": 2.0151817818617657, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5044 + }, + { + "epoch": 2.0155813024370755, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5045 + }, + { + "epoch": 2.0159808230123852, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5046 + }, + { + "epoch": 2.0163803435876946, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5047 + }, + { + "epoch": 2.0167798641630044, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5048 + }, + { + "epoch": 2.017179384738314, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5049 + }, + { + "epoch": 2.0175789053136235, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5050 + }, + { + "epoch": 2.0179784258889333, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5051 + }, + { + "epoch": 2.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5052 + }, + { + "epoch": 2.0187774670395524, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5053 + }, + { + "epoch": 2.019176987614862, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5054 + }, + { + "epoch": 2.019576508190172, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5055 + }, + { + "epoch": 2.0199760287654813, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5056 + }, + { + "epoch": 2.020375549340791, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5057 + }, + { + "epoch": 2.020775069916101, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5058 + }, + { + "epoch": 2.0211745904914102, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5059 + }, + { + "epoch": 2.02157411106672, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5060 + }, + { + "epoch": 2.0219736316420294, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5061 + }, + { + "epoch": 2.022373152217339, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3433, + "step": 5062 + }, + { + "epoch": 2.022772672792649, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5063 + }, + { + "epoch": 2.0231721933679583, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5064 + }, + { + "epoch": 2.023571713943268, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5065 + }, + { + "epoch": 2.023971234518578, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5066 + }, + { + "epoch": 2.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5067 + }, + { + "epoch": 2.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5068 + }, + { + "epoch": 2.0251697962445068, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5069 + }, + { + "epoch": 2.025569316819816, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5070 + }, + { + "epoch": 2.025968837395126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5071 + }, + { + "epoch": 2.0263683579704357, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5072 + }, + { + "epoch": 2.026767878545745, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5073 + }, + { + "epoch": 2.027167399121055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5074 + }, + { + "epoch": 2.027566919696364, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5075 + }, + { + "epoch": 2.027966440271674, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5076 + }, + { + "epoch": 2.0283659608469837, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5077 + }, + { + "epoch": 2.028765481422293, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5078 + }, + { + "epoch": 2.029165001997603, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5079 + }, + { + "epoch": 2.0295645225729126, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5080 + }, + { + "epoch": 2.029964043148222, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5081 + }, + { + "epoch": 2.0303635637235318, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5082 + }, + { + "epoch": 2.0307630842988416, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5083 + }, + { + "epoch": 2.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5084 + }, + { + "epoch": 2.0315621254494607, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 5085 + }, + { + "epoch": 2.0319616460247705, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5086 + }, + { + "epoch": 2.03236116660008, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5087 + }, + { + "epoch": 2.0327606871753896, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5088 + }, + { + "epoch": 2.033160207750699, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5089 + }, + { + "epoch": 2.0335597283260087, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5090 + }, + { + "epoch": 2.0339592489013185, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5091 + }, + { + "epoch": 2.034358769476628, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5092 + }, + { + "epoch": 2.0347582900519376, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5093 + }, + { + "epoch": 2.0351578106272474, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5094 + }, + { + "epoch": 2.0355573312025568, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5095 + }, + { + "epoch": 2.0359568517778666, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5096 + }, + { + "epoch": 2.0363563723531763, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5097 + }, + { + "epoch": 2.0367558929284857, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5098 + }, + { + "epoch": 2.0371554135037955, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5099 + }, + { + "epoch": 2.0375549340791053, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 5100 + }, + { + "epoch": 2.0379544546544146, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5101 + }, + { + "epoch": 2.0383539752297244, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5102 + }, + { + "epoch": 2.038753495805034, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5103 + }, + { + "epoch": 2.0391530163803435, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5104 + }, + { + "epoch": 2.0395525369556533, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5105 + }, + { + "epoch": 2.0399520575309626, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5106 + }, + { + "epoch": 2.0403515781062724, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5107 + }, + { + "epoch": 2.040751098681582, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 5108 + }, + { + "epoch": 2.0411506192568916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.347, + "step": 5109 + }, + { + "epoch": 2.0415501398322013, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5110 + }, + { + "epoch": 2.041949660407511, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5111 + }, + { + "epoch": 2.0423491809828205, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5112 + }, + { + "epoch": 2.0427487015581303, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5113 + }, + { + "epoch": 2.04314822213344, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5114 + }, + { + "epoch": 2.0435477427087494, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 5115 + }, + { + "epoch": 2.043947263284059, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 5116 + }, + { + "epoch": 2.044346783859369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5117 + }, + { + "epoch": 2.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5118 + }, + { + "epoch": 2.045145825009988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5119 + }, + { + "epoch": 2.0455453455852974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 5120 + }, + { + "epoch": 2.0459448661606072, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5121 + }, + { + "epoch": 2.046344386735917, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5122 + }, + { + "epoch": 2.0467439073112264, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5123 + }, + { + "epoch": 2.047143427886536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5124 + }, + { + "epoch": 2.047542948461846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5125 + }, + { + "epoch": 2.0479424690371553, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5126 + }, + { + "epoch": 2.048341989612465, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5127 + }, + { + "epoch": 2.048741510187775, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5128 + }, + { + "epoch": 2.049141030763084, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5129 + }, + { + "epoch": 2.049540551338394, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5130 + }, + { + "epoch": 2.0499400719137038, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5131 + }, + { + "epoch": 2.050339592489013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5132 + }, + { + "epoch": 2.050739113064323, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5133 + }, + { + "epoch": 2.0511386336396322, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5134 + }, + { + "epoch": 2.051538154214942, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5135 + }, + { + "epoch": 2.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3438, + "step": 5136 + }, + { + "epoch": 2.052337195365561, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5137 + }, + { + "epoch": 2.052736715940871, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3452, + "step": 5138 + }, + { + "epoch": 2.0531362365161807, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5139 + }, + { + "epoch": 2.05353575709149, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5140 + }, + { + "epoch": 2.0539352776668, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5141 + }, + { + "epoch": 2.0543347982421096, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5142 + }, + { + "epoch": 2.054734318817419, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5143 + }, + { + "epoch": 2.0551338393927288, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5144 + }, + { + "epoch": 2.0555333599680385, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 5145 + }, + { + "epoch": 2.055932880543348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5146 + }, + { + "epoch": 2.0563324011186577, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5147 + }, + { + "epoch": 2.0567319216939675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5148 + }, + { + "epoch": 2.057131442269277, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 5149 + }, + { + "epoch": 2.0575309628445866, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5150 + }, + { + "epoch": 2.057930483419896, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5151 + }, + { + "epoch": 2.0583300039952057, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5152 + }, + { + "epoch": 2.0587295245705155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5153 + }, + { + "epoch": 2.059129045145825, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5154 + }, + { + "epoch": 2.0595285657211346, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5155 + }, + { + "epoch": 2.0599280862964444, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5156 + }, + { + "epoch": 2.0603276068717538, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5157 + }, + { + "epoch": 2.0607271274470635, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5158 + }, + { + "epoch": 2.0611266480223733, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5159 + }, + { + "epoch": 2.0615261685976827, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5160 + }, + { + "epoch": 2.0619256891729925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 5161 + }, + { + "epoch": 2.0623252097483022, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5162 + }, + { + "epoch": 2.0627247303236116, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5163 + }, + { + "epoch": 2.0631242508989214, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5164 + }, + { + "epoch": 2.0635237714742307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5165 + }, + { + "epoch": 2.0639232920495405, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 5166 + }, + { + "epoch": 2.0643228126248503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5167 + }, + { + "epoch": 2.0647223332001596, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5168 + }, + { + "epoch": 2.0651218537754694, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5169 + }, + { + "epoch": 2.065521374350779, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5170 + }, + { + "epoch": 2.0659208949260885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5171 + }, + { + "epoch": 2.0663204155013983, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5172 + }, + { + "epoch": 2.066719936076708, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5173 + }, + { + "epoch": 2.0671194566520175, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 5174 + }, + { + "epoch": 2.0675189772273272, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5175 + }, + { + "epoch": 2.067918497802637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5176 + }, + { + "epoch": 2.0683180183779464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 5177 + }, + { + "epoch": 2.068717538953256, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5178 + }, + { + "epoch": 2.0691170595285655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 5179 + }, + { + "epoch": 2.0695165801038753, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5180 + }, + { + "epoch": 2.069916100679185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5181 + }, + { + "epoch": 2.0703156212544944, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5182 + }, + { + "epoch": 2.070715141829804, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5183 + }, + { + "epoch": 2.071114662405114, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5184 + }, + { + "epoch": 2.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5185 + }, + { + "epoch": 2.071913703555733, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5186 + }, + { + "epoch": 2.072313224131043, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5187 + }, + { + "epoch": 2.0727127447063523, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5188 + }, + { + "epoch": 2.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5189 + }, + { + "epoch": 2.073511785856972, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 5190 + }, + { + "epoch": 2.073911306432281, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5191 + }, + { + "epoch": 2.074310827007591, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5192 + }, + { + "epoch": 2.0747103475829007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5193 + }, + { + "epoch": 2.07510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5194 + }, + { + "epoch": 2.07550938873352, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5195 + }, + { + "epoch": 2.075908909308829, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5196 + }, + { + "epoch": 2.076308429884139, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5197 + }, + { + "epoch": 2.076707950459449, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5198 + }, + { + "epoch": 2.077107471034758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5199 + }, + { + "epoch": 2.077506991610068, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5200 + }, + { + "epoch": 2.0779065121853777, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5201 + }, + { + "epoch": 2.078306032760687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5202 + }, + { + "epoch": 2.078705553335997, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5203 + }, + { + "epoch": 2.0791050739113066, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5204 + }, + { + "epoch": 2.079504594486616, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3499, + "step": 5205 + }, + { + "epoch": 2.0799041150619257, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5206 + }, + { + "epoch": 2.0803036356372355, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5207 + }, + { + "epoch": 2.080703156212545, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5208 + }, + { + "epoch": 2.0811026767878547, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 5209 + }, + { + "epoch": 2.081502197363164, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5210 + }, + { + "epoch": 2.081901717938474, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5211 + }, + { + "epoch": 2.0823012385137836, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5212 + }, + { + "epoch": 2.082700759089093, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5213 + }, + { + "epoch": 2.0831002796644027, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 5214 + }, + { + "epoch": 2.0834998002397125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5215 + }, + { + "epoch": 2.083899320815022, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5216 + }, + { + "epoch": 2.0842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5217 + }, + { + "epoch": 2.0846983619656414, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5218 + }, + { + "epoch": 2.0850978825409507, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5219 + }, + { + "epoch": 2.0854974031162605, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 5220 + }, + { + "epoch": 2.0858969236915703, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5221 + }, + { + "epoch": 2.0862964442668797, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3478, + "step": 5222 + }, + { + "epoch": 2.0866959648421894, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5223 + }, + { + "epoch": 2.087095485417499, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 5224 + }, + { + "epoch": 2.0874950059928086, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5225 + }, + { + "epoch": 2.0878945265681184, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5226 + }, + { + "epoch": 2.0882940471434277, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5227 + }, + { + "epoch": 2.0886935677187375, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5228 + }, + { + "epoch": 2.0890930882940473, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5229 + }, + { + "epoch": 2.0894926088693566, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5230 + }, + { + "epoch": 2.0898921294446664, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5231 + }, + { + "epoch": 2.090291650019976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5232 + }, + { + "epoch": 2.0906911705952855, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5233 + }, + { + "epoch": 2.0910906911705953, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5234 + }, + { + "epoch": 2.091490211745905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5235 + }, + { + "epoch": 2.0918897323212144, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5236 + }, + { + "epoch": 2.0922892528965242, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5237 + }, + { + "epoch": 2.092688773471834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5238 + }, + { + "epoch": 2.0930882940471434, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 5239 + }, + { + "epoch": 2.093487814622453, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5240 + }, + { + "epoch": 2.0938873351977625, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5241 + }, + { + "epoch": 2.0942868557730723, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5242 + }, + { + "epoch": 2.094686376348382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5243 + }, + { + "epoch": 2.0950858969236914, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5244 + }, + { + "epoch": 2.095485417499001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5245 + }, + { + "epoch": 2.095884938074311, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5246 + }, + { + "epoch": 2.0962844586496203, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5247 + }, + { + "epoch": 2.09668397922493, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5248 + }, + { + "epoch": 2.09708349980024, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5249 + }, + { + "epoch": 2.0974830203755492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5250 + }, + { + "epoch": 2.097882540950859, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 5251 + }, + { + "epoch": 2.098282061526169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3483, + "step": 5252 + }, + { + "epoch": 2.098681582101478, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5253 + }, + { + "epoch": 2.099081102676788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5254 + }, + { + "epoch": 2.0994806232520973, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5255 + }, + { + "epoch": 2.099880143827407, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5256 + }, + { + "epoch": 2.100279664402717, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5257 + }, + { + "epoch": 2.100679184978026, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5258 + }, + { + "epoch": 2.101078705553336, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5259 + }, + { + "epoch": 2.1014782261286458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5260 + }, + { + "epoch": 2.101877746703955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5261 + }, + { + "epoch": 2.102277267279265, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5262 + }, + { + "epoch": 2.1026767878545747, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5263 + }, + { + "epoch": 2.103076308429884, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5264 + }, + { + "epoch": 2.103475829005194, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5265 + }, + { + "epoch": 2.1038753495805036, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5266 + }, + { + "epoch": 2.104274870155813, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5267 + }, + { + "epoch": 2.1046743907311227, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5268 + }, + { + "epoch": 2.105073911306432, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5269 + }, + { + "epoch": 2.105473431881742, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5270 + }, + { + "epoch": 2.1058729524570516, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5271 + }, + { + "epoch": 2.106272473032361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5272 + }, + { + "epoch": 2.1066719936076708, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5273 + }, + { + "epoch": 2.1070715141829806, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 5274 + }, + { + "epoch": 2.10747103475829, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5275 + }, + { + "epoch": 2.1078705553335997, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5276 + }, + { + "epoch": 2.1082700759089095, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5277 + }, + { + "epoch": 2.108669596484219, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5278 + }, + { + "epoch": 2.1090691170595286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5279 + }, + { + "epoch": 2.1094686376348384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5280 + }, + { + "epoch": 2.1098681582101477, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5281 + }, + { + "epoch": 2.1102676787854575, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5282 + }, + { + "epoch": 2.1106671993607673, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5283 + }, + { + "epoch": 2.1110667199360766, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5284 + }, + { + "epoch": 2.1114662405113864, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5285 + }, + { + "epoch": 2.1118657610866958, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5286 + }, + { + "epoch": 2.1122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5287 + }, + { + "epoch": 2.1126648022373153, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5288 + }, + { + "epoch": 2.1130643228126247, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5289 + }, + { + "epoch": 2.1134638433879345, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5290 + }, + { + "epoch": 2.1138633639632443, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 5291 + }, + { + "epoch": 2.1142628845385536, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5292 + }, + { + "epoch": 2.1146624051138634, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5293 + }, + { + "epoch": 2.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5294 + }, + { + "epoch": 2.1154614462644825, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5295 + }, + { + "epoch": 2.1158609668397923, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5296 + }, + { + "epoch": 2.116260487415102, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 5297 + }, + { + "epoch": 2.1166600079904114, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5298 + }, + { + "epoch": 2.117059528565721, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5299 + }, + { + "epoch": 2.1174590491410306, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5300 + }, + { + "epoch": 2.1178585697163403, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5301 + }, + { + "epoch": 2.11825809029165, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5302 + }, + { + "epoch": 2.1186576108669595, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5303 + }, + { + "epoch": 2.1190571314422693, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5304 + }, + { + "epoch": 2.119456652017579, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5305 + }, + { + "epoch": 2.1198561725928884, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5306 + }, + { + "epoch": 2.120255693168198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5307 + }, + { + "epoch": 2.120655213743508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5308 + }, + { + "epoch": 2.1210547343188173, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5309 + }, + { + "epoch": 2.121454254894127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5310 + }, + { + "epoch": 2.121853775469437, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5311 + }, + { + "epoch": 2.122253296044746, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5312 + }, + { + "epoch": 2.122652816620056, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5313 + }, + { + "epoch": 2.1230523371953653, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5314 + }, + { + "epoch": 2.123451857770675, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5315 + }, + { + "epoch": 2.123851378345985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5316 + }, + { + "epoch": 2.1242508989212943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5317 + }, + { + "epoch": 2.124650419496604, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5318 + }, + { + "epoch": 2.125049940071914, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5319 + }, + { + "epoch": 2.125449460647223, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5320 + }, + { + "epoch": 2.125848981222533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5321 + }, + { + "epoch": 2.1262485017978427, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5322 + }, + { + "epoch": 2.126648022373152, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5323 + }, + { + "epoch": 2.127047542948462, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5324 + }, + { + "epoch": 2.1274470635237717, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5325 + }, + { + "epoch": 2.127846584099081, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5326 + }, + { + "epoch": 2.128246104674391, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5327 + }, + { + "epoch": 2.1286456252497006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5328 + }, + { + "epoch": 2.12904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5329 + }, + { + "epoch": 2.1294446664003197, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 5330 + }, + { + "epoch": 2.129844186975629, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5331 + }, + { + "epoch": 2.130243707550939, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5332 + }, + { + "epoch": 2.1306432281262486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5333 + }, + { + "epoch": 2.131042748701558, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5334 + }, + { + "epoch": 2.1314422692768678, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5335 + }, + { + "epoch": 2.1318417898521775, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5336 + }, + { + "epoch": 2.132241310427487, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5337 + }, + { + "epoch": 2.1326408310027967, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5338 + }, + { + "epoch": 2.1330403515781065, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5339 + }, + { + "epoch": 2.133439872153416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5340 + }, + { + "epoch": 2.1338393927287256, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5341 + }, + { + "epoch": 2.134238913304035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5342 + }, + { + "epoch": 2.1346384338793447, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5343 + }, + { + "epoch": 2.1350379544546545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5344 + }, + { + "epoch": 2.135437475029964, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5345 + }, + { + "epoch": 2.1358369956052736, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 5346 + }, + { + "epoch": 2.1362365161805834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5347 + }, + { + "epoch": 2.1366360367558928, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5348 + }, + { + "epoch": 2.1370355573312025, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5349 + }, + { + "epoch": 2.1374350779065123, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5350 + }, + { + "epoch": 2.1378345984818217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 5351 + }, + { + "epoch": 2.1382341190571315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 5352 + }, + { + "epoch": 2.1386336396324412, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5353 + }, + { + "epoch": 2.1390331602077506, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5354 + }, + { + "epoch": 2.1394326807830604, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5355 + }, + { + "epoch": 2.13983220135837, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5356 + }, + { + "epoch": 2.1402317219336795, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3472, + "step": 5357 + }, + { + "epoch": 2.1406312425089893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5358 + }, + { + "epoch": 2.1410307630842986, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5359 + }, + { + "epoch": 2.1414302836596084, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5360 + }, + { + "epoch": 2.141829804234918, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5361 + }, + { + "epoch": 2.1422293248102275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5362 + }, + { + "epoch": 2.1426288453855373, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 5363 + }, + { + "epoch": 2.143028365960847, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5364 + }, + { + "epoch": 2.1434278865361565, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5365 + }, + { + "epoch": 2.1438274071114662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5366 + }, + { + "epoch": 2.144226927686776, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5367 + }, + { + "epoch": 2.1446264482620854, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5368 + }, + { + "epoch": 2.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5369 + }, + { + "epoch": 2.145425489412705, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5370 + }, + { + "epoch": 2.1458250099880143, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5371 + }, + { + "epoch": 2.146224530563324, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5372 + }, + { + "epoch": 2.146624051138634, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5373 + }, + { + "epoch": 2.147023571713943, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5374 + }, + { + "epoch": 2.147423092289253, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5375 + }, + { + "epoch": 2.1478226128645623, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5376 + }, + { + "epoch": 2.148222133439872, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5377 + }, + { + "epoch": 2.148621654015182, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 5378 + }, + { + "epoch": 2.1490211745904912, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5379 + }, + { + "epoch": 2.149420695165801, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5380 + }, + { + "epoch": 2.149820215741111, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5381 + }, + { + "epoch": 2.15021973631642, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5382 + }, + { + "epoch": 2.15061925689173, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5383 + }, + { + "epoch": 2.1510187774670397, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5384 + }, + { + "epoch": 2.151418298042349, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5385 + }, + { + "epoch": 2.151817818617659, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5386 + }, + { + "epoch": 2.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5387 + }, + { + "epoch": 2.152616859768278, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5388 + }, + { + "epoch": 2.1530163803435878, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5389 + }, + { + "epoch": 2.153415900918897, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5390 + }, + { + "epoch": 2.153815421494207, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5391 + }, + { + "epoch": 2.1542149420695167, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5392 + }, + { + "epoch": 2.154614462644826, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5393 + }, + { + "epoch": 2.155013983220136, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5394 + }, + { + "epoch": 2.1554135037954456, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5395 + }, + { + "epoch": 2.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5396 + }, + { + "epoch": 2.1562125449460647, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5397 + }, + { + "epoch": 2.1566120655213745, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 5398 + }, + { + "epoch": 2.157011586096684, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5399 + }, + { + "epoch": 2.1574111066719937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5400 + }, + { + "epoch": 2.1578106272473034, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5401 + }, + { + "epoch": 2.158210147822613, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5402 + }, + { + "epoch": 2.1586096683979226, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5403 + }, + { + "epoch": 2.159009188973232, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5404 + }, + { + "epoch": 2.1594087095485417, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 5405 + }, + { + "epoch": 2.1598082301238515, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5406 + }, + { + "epoch": 2.160207750699161, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5407 + }, + { + "epoch": 2.1606072712744706, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5408 + }, + { + "epoch": 2.1610067918497804, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5409 + }, + { + "epoch": 2.1614063124250897, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5410 + }, + { + "epoch": 2.1618058330003995, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5411 + }, + { + "epoch": 2.1622053535757093, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5412 + }, + { + "epoch": 2.1626048741510187, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5413 + }, + { + "epoch": 2.1630043947263284, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5414 + }, + { + "epoch": 2.1634039153016382, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5415 + }, + { + "epoch": 2.1638034358769476, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5416 + }, + { + "epoch": 2.1642029564522574, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5417 + }, + { + "epoch": 2.164602477027567, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5418 + }, + { + "epoch": 2.1650019976028765, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5419 + }, + { + "epoch": 2.1654015181781863, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5420 + }, + { + "epoch": 2.1658010387534956, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5421 + }, + { + "epoch": 2.1662005593288054, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5422 + }, + { + "epoch": 2.166600079904115, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5423 + }, + { + "epoch": 2.1669996004794245, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5424 + }, + { + "epoch": 2.1673991210547343, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5425 + }, + { + "epoch": 2.167798641630044, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5426 + }, + { + "epoch": 2.1681981622053534, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5427 + }, + { + "epoch": 2.1685976827806632, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5428 + }, + { + "epoch": 2.168997203355973, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5429 + }, + { + "epoch": 2.1693967239312824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5430 + }, + { + "epoch": 2.169796244506592, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5431 + }, + { + "epoch": 2.1701957650819015, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5432 + }, + { + "epoch": 2.1705952856572113, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5433 + }, + { + "epoch": 2.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5434 + }, + { + "epoch": 2.1713943268078304, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5435 + }, + { + "epoch": 2.17179384738314, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 5436 + }, + { + "epoch": 2.17219336795845, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 5437 + }, + { + "epoch": 2.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5438 + }, + { + "epoch": 2.172992409109069, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5439 + }, + { + "epoch": 2.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5440 + }, + { + "epoch": 2.1737914502596882, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5441 + }, + { + "epoch": 2.174190970834998, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5442 + }, + { + "epoch": 2.174590491410308, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3393, + "step": 5443 + }, + { + "epoch": 2.174990011985617, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5444 + }, + { + "epoch": 2.175389532560927, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5445 + }, + { + "epoch": 2.1757890531362367, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5446 + }, + { + "epoch": 2.176188573711546, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5447 + }, + { + "epoch": 2.176588094286856, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5448 + }, + { + "epoch": 2.176987614862165, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5449 + }, + { + "epoch": 2.177387135437475, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5450 + }, + { + "epoch": 2.1777866560127848, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5451 + }, + { + "epoch": 2.178186176588094, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 5452 + }, + { + "epoch": 2.178585697163404, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 5453 + }, + { + "epoch": 2.1789852177387137, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5454 + }, + { + "epoch": 2.179384738314023, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5455 + }, + { + "epoch": 2.179784258889333, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5456 + }, + { + "epoch": 2.1801837794646426, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5457 + }, + { + "epoch": 2.180583300039952, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5458 + }, + { + "epoch": 2.1809828206152617, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 5459 + }, + { + "epoch": 2.1813823411905715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5460 + }, + { + "epoch": 2.181781861765881, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5461 + }, + { + "epoch": 2.1821813823411906, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5462 + }, + { + "epoch": 2.1825809029165004, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5463 + }, + { + "epoch": 2.1829804234918098, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5464 + }, + { + "epoch": 2.1833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5465 + }, + { + "epoch": 2.183779464642429, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 5466 + }, + { + "epoch": 2.1841789852177387, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 5467 + }, + { + "epoch": 2.1845785057930485, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5468 + }, + { + "epoch": 2.184978026368358, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5469 + }, + { + "epoch": 2.1853775469436676, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5470 + }, + { + "epoch": 2.1857770675189774, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5471 + }, + { + "epoch": 2.1861765880942867, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5472 + }, + { + "epoch": 2.1865761086695965, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 5473 + }, + { + "epoch": 2.1869756292449063, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5474 + }, + { + "epoch": 2.1873751498202156, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5475 + }, + { + "epoch": 2.1877746703955254, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 5476 + }, + { + "epoch": 2.1881741909708348, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5477 + }, + { + "epoch": 2.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5478 + }, + { + "epoch": 2.1889732321214543, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5479 + }, + { + "epoch": 2.1893727526967637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 5480 + }, + { + "epoch": 2.1897722732720735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 5481 + }, + { + "epoch": 2.1901717938473833, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5482 + }, + { + "epoch": 2.1905713144226926, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5483 + }, + { + "epoch": 2.1909708349980024, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5484 + }, + { + "epoch": 2.191370355573312, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5485 + }, + { + "epoch": 2.1917698761486215, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5486 + }, + { + "epoch": 2.1921693967239313, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 5487 + }, + { + "epoch": 2.192568917299241, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5488 + }, + { + "epoch": 2.1929684378745504, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3482, + "step": 5489 + }, + { + "epoch": 2.19336795844986, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3431, + "step": 5490 + }, + { + "epoch": 2.19376747902517, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5491 + }, + { + "epoch": 2.1941669996004793, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5492 + }, + { + "epoch": 2.194566520175789, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5493 + }, + { + "epoch": 2.1949660407510985, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5494 + }, + { + "epoch": 2.1953655613264083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5495 + }, + { + "epoch": 2.195765081901718, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5496 + }, + { + "epoch": 2.1961646024770274, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5497 + }, + { + "epoch": 2.196564123052337, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5498 + }, + { + "epoch": 2.196963643627647, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5499 + }, + { + "epoch": 2.1973631642029563, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5500 + }, + { + "epoch": 2.197762684778266, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5501 + }, + { + "epoch": 2.198162205353576, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5502 + }, + { + "epoch": 2.198561725928885, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5503 + }, + { + "epoch": 2.198961246504195, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5504 + }, + { + "epoch": 2.199360767079505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5505 + }, + { + "epoch": 2.199760287654814, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5506 + }, + { + "epoch": 2.200159808230124, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5507 + }, + { + "epoch": 2.2005593288054337, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5508 + }, + { + "epoch": 2.200958849380743, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5509 + }, + { + "epoch": 2.201358369956053, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5510 + }, + { + "epoch": 2.201757890531362, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 5511 + }, + { + "epoch": 2.202157411106672, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5512 + }, + { + "epoch": 2.2025569316819817, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5513 + }, + { + "epoch": 2.202956452257291, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5514 + }, + { + "epoch": 2.203355972832601, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5515 + }, + { + "epoch": 2.2037554934079107, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5516 + }, + { + "epoch": 2.20415501398322, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 5517 + }, + { + "epoch": 2.20455453455853, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5518 + }, + { + "epoch": 2.2049540551338396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5519 + }, + { + "epoch": 2.205353575709149, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5520 + }, + { + "epoch": 2.2057530962844587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 5521 + }, + { + "epoch": 2.206152616859768, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 5522 + }, + { + "epoch": 2.206552137435078, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5523 + }, + { + "epoch": 2.2069516580103876, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5524 + }, + { + "epoch": 2.207351178585697, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5525 + }, + { + "epoch": 2.2077506991610067, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 5526 + }, + { + "epoch": 2.2081502197363165, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.342, + "step": 5527 + }, + { + "epoch": 2.208549740311626, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5528 + }, + { + "epoch": 2.2089492608869357, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5529 + }, + { + "epoch": 2.2093487814622454, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5530 + }, + { + "epoch": 2.209748302037555, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5531 + }, + { + "epoch": 2.2101478226128646, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 5532 + }, + { + "epoch": 2.2105473431881744, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 5533 + }, + { + "epoch": 2.2109468637634837, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5534 + }, + { + "epoch": 2.2113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5535 + }, + { + "epoch": 2.2117459049141033, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5536 + }, + { + "epoch": 2.2121454254894126, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 5537 + }, + { + "epoch": 2.2125449460647224, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5538 + }, + { + "epoch": 2.2129444666400317, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5539 + }, + { + "epoch": 2.2133439872153415, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5540 + }, + { + "epoch": 2.2137435077906513, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5541 + }, + { + "epoch": 2.2141430283659607, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5542 + }, + { + "epoch": 2.2145425489412705, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5543 + }, + { + "epoch": 2.2149420695165802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5544 + }, + { + "epoch": 2.2153415900918896, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5545 + }, + { + "epoch": 2.2157411106671994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 5546 + }, + { + "epoch": 2.216140631242509, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5547 + }, + { + "epoch": 2.2165401518178185, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5548 + }, + { + "epoch": 2.2169396723931283, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5549 + }, + { + "epoch": 2.217339192968438, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5550 + }, + { + "epoch": 2.2177387135437474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5551 + }, + { + "epoch": 2.218138234119057, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5552 + }, + { + "epoch": 2.218537754694367, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5553 + }, + { + "epoch": 2.2189372752696763, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3474, + "step": 5554 + }, + { + "epoch": 2.219336795844986, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5555 + }, + { + "epoch": 2.2197363164202955, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5556 + }, + { + "epoch": 2.2201358369956052, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 5557 + }, + { + "epoch": 2.220535357570915, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5558 + }, + { + "epoch": 2.2209348781462244, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 5559 + }, + { + "epoch": 2.221334398721534, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5560 + }, + { + "epoch": 2.221733919296844, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5561 + }, + { + "epoch": 2.2221334398721533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5562 + }, + { + "epoch": 2.222532960447463, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5563 + }, + { + "epoch": 2.222932481022773, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5564 + }, + { + "epoch": 2.223332001598082, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5565 + }, + { + "epoch": 2.223731522173392, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 5566 + }, + { + "epoch": 2.2241310427487013, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5567 + }, + { + "epoch": 2.224530563324011, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5568 + }, + { + "epoch": 2.224930083899321, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5569 + }, + { + "epoch": 2.2253296044746302, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5570 + }, + { + "epoch": 2.22572912504994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5571 + }, + { + "epoch": 2.22612864562525, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5572 + }, + { + "epoch": 2.226528166200559, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5573 + }, + { + "epoch": 2.226927686775869, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5574 + }, + { + "epoch": 2.2273272073511787, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5575 + }, + { + "epoch": 2.227726727926488, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5576 + }, + { + "epoch": 2.228126248501798, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5577 + }, + { + "epoch": 2.2285257690771076, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5578 + }, + { + "epoch": 2.228925289652417, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5579 + }, + { + "epoch": 2.2293248102277268, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 5580 + }, + { + "epoch": 2.2297243308030366, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5581 + }, + { + "epoch": 2.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 5582 + }, + { + "epoch": 2.2305233719536557, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5583 + }, + { + "epoch": 2.230922892528965, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5584 + }, + { + "epoch": 2.231322413104275, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 5585 + }, + { + "epoch": 2.2317219336795846, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 5586 + }, + { + "epoch": 2.232121454254894, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 5587 + }, + { + "epoch": 2.2325209748302037, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5588 + }, + { + "epoch": 2.2329204954055135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 5589 + }, + { + "epoch": 2.233320015980823, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5590 + }, + { + "epoch": 2.2337195365561326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5591 + }, + { + "epoch": 2.2341190571314424, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5592 + }, + { + "epoch": 2.2345185777067518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 5593 + }, + { + "epoch": 2.2349180982820616, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5594 + }, + { + "epoch": 2.2353176188573713, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 5595 + }, + { + "epoch": 2.2357171394326807, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5596 + }, + { + "epoch": 2.2361166600079905, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5597 + }, + { + "epoch": 2.2365161805833003, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5598 + }, + { + "epoch": 2.2369157011586096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5599 + }, + { + "epoch": 2.2373152217339194, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5600 + }, + { + "epoch": 2.2377147423092287, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5601 + }, + { + "epoch": 2.2381142628845385, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5602 + }, + { + "epoch": 2.2385137834598483, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5603 + }, + { + "epoch": 2.2389133040351576, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5604 + }, + { + "epoch": 2.2393128246104674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 5605 + }, + { + "epoch": 2.239712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 5606 + }, + { + "epoch": 2.2401118657610866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5607 + }, + { + "epoch": 2.2405113863363963, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 5608 + }, + { + "epoch": 2.240910906911706, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 5609 + }, + { + "epoch": 2.2413104274870155, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5610 + }, + { + "epoch": 2.2417099480623253, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5611 + }, + { + "epoch": 2.2421094686376346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5612 + }, + { + "epoch": 2.2425089892129444, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5613 + }, + { + "epoch": 2.242908509788254, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5614 + }, + { + "epoch": 2.2433080303635635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5615 + }, + { + "epoch": 2.2437075509388733, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 5616 + }, + { + "epoch": 2.244107071514183, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5617 + }, + { + "epoch": 2.2445065920894924, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 5618 + }, + { + "epoch": 2.2449061126648022, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5619 + }, + { + "epoch": 2.245305633240112, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5620 + }, + { + "epoch": 2.2457051538154214, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 5621 + }, + { + "epoch": 2.246104674390731, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 5622 + }, + { + "epoch": 2.246504194966041, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 5623 + }, + { + "epoch": 2.2469037155413503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5624 + }, + { + "epoch": 2.24730323611666, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5625 + }, + { + "epoch": 2.24770275669197, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5626 + }, + { + "epoch": 2.248102277267279, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5627 + }, + { + "epoch": 2.248501797842589, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5628 + }, + { + "epoch": 2.2489013184178983, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5629 + }, + { + "epoch": 2.249300838993208, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5630 + }, + { + "epoch": 2.249700359568518, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5631 + }, + { + "epoch": 2.2500998801438272, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 5632 + }, + { + "epoch": 2.250499400719137, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5633 + }, + { + "epoch": 2.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5634 + }, + { + "epoch": 2.251298441869756, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5635 + }, + { + "epoch": 2.251697962445066, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 5636 + }, + { + "epoch": 2.2520974830203757, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5637 + }, + { + "epoch": 2.252497003595685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 5638 + }, + { + "epoch": 2.252896524170995, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5639 + }, + { + "epoch": 2.2532960447463046, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 5640 + }, + { + "epoch": 2.253695565321614, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3399, + "step": 5641 + }, + { + "epoch": 2.2540950858969238, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 5642 + }, + { + "epoch": 2.2544946064722335, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5643 + }, + { + "epoch": 2.254894127047543, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5644 + }, + { + "epoch": 2.2552936476228527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5645 + }, + { + "epoch": 2.255693168198162, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5646 + }, + { + "epoch": 2.256092688773472, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 5647 + }, + { + "epoch": 2.2564922093487816, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5648 + }, + { + "epoch": 2.256891729924091, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5649 + }, + { + "epoch": 2.2572912504994007, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 5650 + }, + { + "epoch": 2.2576907710747105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3442, + "step": 5651 + }, + { + "epoch": 2.25809029165002, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5652 + }, + { + "epoch": 2.2584898122253296, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 5653 + }, + { + "epoch": 2.2588893328006394, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 5654 + }, + { + "epoch": 2.2592888533759488, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5655 + }, + { + "epoch": 2.2596883739512585, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5656 + }, + { + "epoch": 2.260087894526568, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 5657 + }, + { + "epoch": 2.2604874151018777, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5658 + }, + { + "epoch": 2.2608869356771875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5659 + }, + { + "epoch": 2.2612864562524972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 5660 + }, + { + "epoch": 2.2616859768278066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5661 + }, + { + "epoch": 2.2620854974031164, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5662 + }, + { + "epoch": 2.2624850179784257, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 5663 + }, + { + "epoch": 2.2628845385537355, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5664 + }, + { + "epoch": 2.2632840591290453, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5665 + }, + { + "epoch": 2.2636835797043546, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5666 + }, + { + "epoch": 2.2640831002796644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 5667 + }, + { + "epoch": 2.264482620854974, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5668 + }, + { + "epoch": 2.2648821414302835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5669 + }, + { + "epoch": 2.2652816620055933, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5670 + }, + { + "epoch": 2.265681182580903, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5671 + }, + { + "epoch": 2.2660807031562125, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 5672 + }, + { + "epoch": 2.2664802237315222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 5673 + }, + { + "epoch": 2.2668797443068316, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5674 + }, + { + "epoch": 2.2672792648821414, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 5675 + }, + { + "epoch": 2.267678785457451, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5676 + }, + { + "epoch": 2.2680783060327605, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 5677 + }, + { + "epoch": 2.2684778266080703, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5678 + }, + { + "epoch": 2.26887734718338, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 5679 + }, + { + "epoch": 2.2692768677586894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5680 + }, + { + "epoch": 2.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5681 + }, + { + "epoch": 2.270075908909309, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5682 + }, + { + "epoch": 2.2704754294846183, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 5683 + }, + { + "epoch": 2.270874950059928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5684 + }, + { + "epoch": 2.271274470635238, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 5685 + }, + { + "epoch": 2.2716739912105472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5686 + }, + { + "epoch": 2.272073511785857, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5687 + }, + { + "epoch": 2.272473032361167, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5688 + }, + { + "epoch": 2.272872552936476, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5689 + }, + { + "epoch": 2.273272073511786, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5690 + }, + { + "epoch": 2.2736715940870953, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 5691 + }, + { + "epoch": 2.274071114662405, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 5692 + }, + { + "epoch": 2.274470635237715, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5693 + }, + { + "epoch": 2.274870155813024, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 5694 + }, + { + "epoch": 2.275269676388334, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5695 + }, + { + "epoch": 2.275669196963644, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 5696 + }, + { + "epoch": 2.276068717538953, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5697 + }, + { + "epoch": 2.276468238114263, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 5698 + }, + { + "epoch": 2.2768677586895727, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5699 + }, + { + "epoch": 2.277267279264882, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 5700 + }, + { + "epoch": 2.277666799840192, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5701 + }, + { + "epoch": 2.278066320415501, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 5702 + }, + { + "epoch": 2.278465840990811, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3429, + "step": 5703 + }, + { + "epoch": 2.2788653615661207, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5704 + }, + { + "epoch": 2.2792648821414305, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5705 + }, + { + "epoch": 2.27966440271674, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5706 + }, + { + "epoch": 2.2800639232920497, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5707 + }, + { + "epoch": 2.280463443867359, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5708 + }, + { + "epoch": 2.280862964442669, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5709 + }, + { + "epoch": 2.2812624850179786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 5710 + }, + { + "epoch": 2.281662005593288, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5711 + }, + { + "epoch": 2.2820615261685977, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5712 + }, + { + "epoch": 2.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 5713 + }, + { + "epoch": 2.282860567319217, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3426, + "step": 5714 + }, + { + "epoch": 2.2832600878945266, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5715 + }, + { + "epoch": 2.2836596084698364, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5716 + }, + { + "epoch": 2.2840591290451457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5717 + }, + { + "epoch": 2.2844586496204555, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5718 + }, + { + "epoch": 2.284858170195765, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5719 + }, + { + "epoch": 2.2852576907710747, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5720 + }, + { + "epoch": 2.2856572113463844, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5721 + }, + { + "epoch": 2.286056731921694, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5722 + }, + { + "epoch": 2.2864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5723 + }, + { + "epoch": 2.2868557730723134, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 5724 + }, + { + "epoch": 2.2872552936476227, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 5725 + }, + { + "epoch": 2.2876548142229325, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5726 + }, + { + "epoch": 2.2880543347982423, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5727 + }, + { + "epoch": 2.2884538553735516, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5728 + }, + { + "epoch": 2.2888533759488614, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5729 + }, + { + "epoch": 2.289252896524171, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5730 + }, + { + "epoch": 2.2896524170994805, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5731 + }, + { + "epoch": 2.2900519376747903, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5732 + }, + { + "epoch": 2.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5733 + }, + { + "epoch": 2.2908509788254094, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5734 + }, + { + "epoch": 2.2912504994007192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 5735 + }, + { + "epoch": 2.2916500199760286, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 5736 + }, + { + "epoch": 2.2920495405513384, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5737 + }, + { + "epoch": 2.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5738 + }, + { + "epoch": 2.2928485817019575, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5739 + }, + { + "epoch": 2.2932481022772673, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 5740 + }, + { + "epoch": 2.293647622852577, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5741 + }, + { + "epoch": 2.2940471434278864, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5742 + }, + { + "epoch": 2.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5743 + }, + { + "epoch": 2.294846184578506, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5744 + }, + { + "epoch": 2.2952457051538153, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 5745 + }, + { + "epoch": 2.295645225729125, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5746 + }, + { + "epoch": 2.2960447463044344, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5747 + }, + { + "epoch": 2.2964442668797442, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5748 + }, + { + "epoch": 2.296843787455054, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 5749 + }, + { + "epoch": 2.297243308030364, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5750 + }, + { + "epoch": 2.297642828605673, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 5751 + }, + { + "epoch": 2.298042349180983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5752 + }, + { + "epoch": 2.2984418697562923, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5753 + }, + { + "epoch": 2.298841390331602, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 5754 + }, + { + "epoch": 2.299240910906912, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 5755 + }, + { + "epoch": 2.299640431482221, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 5756 + }, + { + "epoch": 2.300039952057531, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5757 + }, + { + "epoch": 2.3004394726328408, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5758 + }, + { + "epoch": 2.30083899320815, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 5759 + }, + { + "epoch": 2.30123851378346, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5760 + }, + { + "epoch": 2.3016380343587697, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 5761 + }, + { + "epoch": 2.302037554934079, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3409, + "step": 5762 + }, + { + "epoch": 2.302437075509389, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5763 + }, + { + "epoch": 2.302836596084698, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5764 + }, + { + "epoch": 2.303236116660008, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5765 + }, + { + "epoch": 2.3036356372353177, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5766 + }, + { + "epoch": 2.304035157810627, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5767 + }, + { + "epoch": 2.304434678385937, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5768 + }, + { + "epoch": 2.3048341989612466, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5769 + }, + { + "epoch": 2.305233719536556, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 5770 + }, + { + "epoch": 2.3056332401118658, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5771 + }, + { + "epoch": 2.3060327606871756, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5772 + }, + { + "epoch": 2.306432281262485, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5773 + }, + { + "epoch": 2.3068318018377947, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5774 + }, + { + "epoch": 2.3072313224131045, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5775 + }, + { + "epoch": 2.307630842988414, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 5776 + }, + { + "epoch": 2.3080303635637236, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 5777 + }, + { + "epoch": 2.3084298841390334, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5778 + }, + { + "epoch": 2.3088294047143427, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 5779 + }, + { + "epoch": 2.3092289252896525, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 5780 + }, + { + "epoch": 2.309628445864962, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5781 + }, + { + "epoch": 2.3100279664402716, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 5782 + }, + { + "epoch": 2.3104274870155814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 5783 + }, + { + "epoch": 2.3108270075908908, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5784 + }, + { + "epoch": 2.3112265281662006, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5785 + }, + { + "epoch": 2.3116260487415103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3435, + "step": 5786 + }, + { + "epoch": 2.3120255693168197, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5787 + }, + { + "epoch": 2.3124250898921295, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 5788 + }, + { + "epoch": 2.3128246104674393, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5789 + }, + { + "epoch": 2.3132241310427486, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5790 + }, + { + "epoch": 2.3136236516180584, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5791 + }, + { + "epoch": 2.3140231721933677, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5792 + }, + { + "epoch": 2.3144226927686775, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5793 + }, + { + "epoch": 2.3148222133439873, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 5794 + }, + { + "epoch": 2.315221733919297, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5795 + }, + { + "epoch": 2.3156212544946064, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5796 + }, + { + "epoch": 2.316020775069916, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5797 + }, + { + "epoch": 2.3164202956452256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5798 + }, + { + "epoch": 2.3168198162205353, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5799 + }, + { + "epoch": 2.317219336795845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 5800 + }, + { + "epoch": 2.3176188573711545, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 5801 + }, + { + "epoch": 2.3180183779464643, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 5802 + }, + { + "epoch": 2.318417898521774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 5803 + }, + { + "epoch": 2.3188174190970834, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 5804 + }, + { + "epoch": 2.319216939672393, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 5805 + }, + { + "epoch": 2.319616460247703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 5806 + }, + { + "epoch": 2.3200159808230123, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 5807 + }, + { + "epoch": 2.320415501398322, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 5808 + }, + { + "epoch": 2.3208150219736314, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 5809 + }, + { + "epoch": 2.321214542548941, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5810 + }, + { + "epoch": 2.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 5811 + }, + { + "epoch": 2.3220135836995603, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5812 + }, + { + "epoch": 2.32241310427487, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 5813 + }, + { + "epoch": 2.32281262485018, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5814 + }, + { + "epoch": 2.3232121454254893, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 5815 + }, + { + "epoch": 2.323611666000799, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 5816 + }, + { + "epoch": 2.324011186576109, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 5817 + }, + { + "epoch": 2.324410707151418, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5818 + }, + { + "epoch": 2.324810227726728, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 5819 + }, + { + "epoch": 2.3252097483020377, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5820 + }, + { + "epoch": 2.325609268877347, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5821 + }, + { + "epoch": 2.326008789452657, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5822 + }, + { + "epoch": 2.3264083100279667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 5823 + }, + { + "epoch": 2.326807830603276, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 5824 + }, + { + "epoch": 2.327207351178586, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5825 + }, + { + "epoch": 2.327606871753895, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5826 + }, + { + "epoch": 2.328006392329205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5827 + }, + { + "epoch": 2.3284059129045147, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5828 + }, + { + "epoch": 2.328805433479824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5829 + }, + { + "epoch": 2.329204954055134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5830 + }, + { + "epoch": 2.3296044746304436, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 5831 + }, + { + "epoch": 2.330003995205753, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5832 + }, + { + "epoch": 2.3304035157810628, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 5833 + }, + { + "epoch": 2.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5834 + }, + { + "epoch": 2.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 5835 + }, + { + "epoch": 2.3316020775069917, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5836 + }, + { + "epoch": 2.332001598082301, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5837 + }, + { + "epoch": 2.332401118657611, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 5838 + }, + { + "epoch": 2.3328006392329206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5839 + }, + { + "epoch": 2.3332001598082304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 5840 + }, + { + "epoch": 2.3335996803835397, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5841 + }, + { + "epoch": 2.3339992009588495, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5842 + }, + { + "epoch": 2.334398721534159, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 5843 + }, + { + "epoch": 2.3347982421094686, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5844 + }, + { + "epoch": 2.3351977626847784, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5845 + }, + { + "epoch": 2.3355972832600878, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 5846 + }, + { + "epoch": 2.3359968038353975, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 5847 + }, + { + "epoch": 2.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5848 + }, + { + "epoch": 2.3367958449860167, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5849 + }, + { + "epoch": 2.3371953655613265, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5850 + }, + { + "epoch": 2.3375948861366362, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 5851 + }, + { + "epoch": 2.3379944067119456, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5852 + }, + { + "epoch": 2.3383939272872554, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 5853 + }, + { + "epoch": 2.3387934478625647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 5854 + }, + { + "epoch": 2.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5855 + }, + { + "epoch": 2.3395924890131843, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5856 + }, + { + "epoch": 2.3399920095884936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 5857 + }, + { + "epoch": 2.3403915301638034, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 5858 + }, + { + "epoch": 2.340791050739113, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5859 + }, + { + "epoch": 2.3411905713144225, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5860 + }, + { + "epoch": 2.3415900918897323, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5861 + }, + { + "epoch": 2.341989612465042, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 5862 + }, + { + "epoch": 2.3423891330403515, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 5863 + }, + { + "epoch": 2.3427886536156612, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5864 + }, + { + "epoch": 2.343188174190971, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 5865 + }, + { + "epoch": 2.3435876947662804, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 5866 + }, + { + "epoch": 2.34398721534159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5867 + }, + { + "epoch": 2.3443867359169, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5868 + }, + { + "epoch": 2.3447862564922093, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 5869 + }, + { + "epoch": 2.345185777067519, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 5870 + }, + { + "epoch": 2.3455852976428284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 5871 + }, + { + "epoch": 2.345984818218138, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 5872 + }, + { + "epoch": 2.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 5873 + }, + { + "epoch": 2.3467838593687573, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 5874 + }, + { + "epoch": 2.347183379944067, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5875 + }, + { + "epoch": 2.347582900519377, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 5876 + }, + { + "epoch": 2.3479824210946862, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 5877 + }, + { + "epoch": 2.348381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 5878 + }, + { + "epoch": 2.348781462245306, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 5879 + }, + { + "epoch": 2.349180982820615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 5880 + }, + { + "epoch": 2.349580503395925, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 5881 + }, + { + "epoch": 2.3499800239712343, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 5882 + }, + { + "epoch": 2.350379544546544, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 5883 + }, + { + "epoch": 2.350779065121854, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 5884 + }, + { + "epoch": 2.3511785856971636, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5885 + }, + { + "epoch": 2.351578106272473, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5886 + }, + { + "epoch": 2.3519776268477828, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 5887 + }, + { + "epoch": 2.352377147423092, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 5888 + }, + { + "epoch": 2.352776667998402, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 5889 + }, + { + "epoch": 2.3531761885737117, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5890 + }, + { + "epoch": 2.353575709149021, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5891 + }, + { + "epoch": 2.353975229724331, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 5892 + }, + { + "epoch": 2.3543747502996406, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 5893 + }, + { + "epoch": 2.35477427087495, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 5894 + }, + { + "epoch": 2.3551737914502597, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5895 + }, + { + "epoch": 2.3555733120255695, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 5896 + }, + { + "epoch": 2.355972832600879, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 5897 + }, + { + "epoch": 2.3563723531761886, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5898 + }, + { + "epoch": 2.356771873751498, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 5899 + }, + { + "epoch": 2.357171394326808, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5900 + }, + { + "epoch": 2.3575709149021176, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5901 + }, + { + "epoch": 2.357970435477427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 5902 + }, + { + "epoch": 2.3583699560527367, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5903 + }, + { + "epoch": 2.3587694766280465, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 5904 + }, + { + "epoch": 2.359168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5905 + }, + { + "epoch": 2.3595685177786656, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 5906 + }, + { + "epoch": 2.3599680383539754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5907 + }, + { + "epoch": 2.3603675589292847, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 5908 + }, + { + "epoch": 2.3607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 5909 + }, + { + "epoch": 2.361166600079904, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5910 + }, + { + "epoch": 2.3615661206552137, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5911 + }, + { + "epoch": 2.3619656412305234, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 5912 + }, + { + "epoch": 2.3623651618058332, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 5913 + }, + { + "epoch": 2.3627646823811426, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 5914 + }, + { + "epoch": 2.3631642029564524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5915 + }, + { + "epoch": 2.3635637235317617, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 5916 + }, + { + "epoch": 2.3639632441070715, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5917 + }, + { + "epoch": 2.3643627646823813, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5918 + }, + { + "epoch": 2.3647622852576906, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 5919 + }, + { + "epoch": 2.3651618058330004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3443, + "step": 5920 + }, + { + "epoch": 2.36556132640831, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5921 + }, + { + "epoch": 2.3659608469836195, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5922 + }, + { + "epoch": 2.3663603675589293, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 5923 + }, + { + "epoch": 2.366759888134239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5924 + }, + { + "epoch": 2.3671594087095484, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 5925 + }, + { + "epoch": 2.3675589292848582, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5926 + }, + { + "epoch": 2.3679584498601676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 5927 + }, + { + "epoch": 2.3683579704354774, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 5928 + }, + { + "epoch": 2.368757491010787, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 5929 + }, + { + "epoch": 2.369157011586097, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 5930 + }, + { + "epoch": 2.3695565321614063, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5931 + }, + { + "epoch": 2.369956052736716, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 5932 + }, + { + "epoch": 2.3703555733120254, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 5933 + }, + { + "epoch": 2.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 5934 + }, + { + "epoch": 2.371154614462645, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 5935 + }, + { + "epoch": 2.3715541350379543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 5936 + }, + { + "epoch": 2.371953655613264, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 5937 + }, + { + "epoch": 2.372353176188574, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 5938 + }, + { + "epoch": 2.3727526967638832, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 5939 + }, + { + "epoch": 2.373152217339193, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 5940 + }, + { + "epoch": 2.373551737914503, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 5941 + }, + { + "epoch": 2.373951258489812, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5942 + }, + { + "epoch": 2.374350779065122, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 5943 + }, + { + "epoch": 2.3747502996404313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 5944 + }, + { + "epoch": 2.375149820215741, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 5945 + }, + { + "epoch": 2.375549340791051, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 5946 + }, + { + "epoch": 2.37594886136636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 5947 + }, + { + "epoch": 2.37634838194167, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 5948 + }, + { + "epoch": 2.3767479025169798, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 5949 + }, + { + "epoch": 2.377147423092289, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 5950 + }, + { + "epoch": 2.377546943667599, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3385, + "step": 5951 + }, + { + "epoch": 2.3779464642429087, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 5952 + }, + { + "epoch": 2.378345984818218, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 5953 + }, + { + "epoch": 2.378745505393528, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5954 + }, + { + "epoch": 2.379145025968837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 5955 + }, + { + "epoch": 2.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 5956 + }, + { + "epoch": 2.3799440671194567, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 5957 + }, + { + "epoch": 2.3803435876947665, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 5958 + }, + { + "epoch": 2.380743108270076, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 5959 + }, + { + "epoch": 2.3811426288453856, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5960 + }, + { + "epoch": 2.381542149420695, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 5961 + }, + { + "epoch": 2.3819416699960048, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 5962 + }, + { + "epoch": 2.3823411905713145, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 5963 + }, + { + "epoch": 2.382740711146624, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 5964 + }, + { + "epoch": 2.3831402317219337, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 5965 + }, + { + "epoch": 2.3835397522972435, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 5966 + }, + { + "epoch": 2.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5967 + }, + { + "epoch": 2.3843387934478626, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 5968 + }, + { + "epoch": 2.3847383140231724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 5969 + }, + { + "epoch": 2.3851378345984817, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 5970 + }, + { + "epoch": 2.3855373551737915, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 5971 + }, + { + "epoch": 2.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 5972 + }, + { + "epoch": 2.3863363963244106, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 5973 + }, + { + "epoch": 2.3867359168997204, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 5974 + }, + { + "epoch": 2.38713543747503, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 5975 + }, + { + "epoch": 2.3875349580503396, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 5976 + }, + { + "epoch": 2.3879344786256493, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 5977 + }, + { + "epoch": 2.3883339992009587, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 5978 + }, + { + "epoch": 2.3887335197762685, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 5979 + }, + { + "epoch": 2.3891330403515783, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 5980 + }, + { + "epoch": 2.3895325609268876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 5981 + }, + { + "epoch": 2.3899320815021974, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 5982 + }, + { + "epoch": 2.390331602077507, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 5983 + }, + { + "epoch": 2.3907311226528165, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 5984 + }, + { + "epoch": 2.3911306432281263, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 5985 + }, + { + "epoch": 2.391530163803436, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 5986 + }, + { + "epoch": 2.3919296843787454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 5987 + }, + { + "epoch": 2.392329204954055, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 5988 + }, + { + "epoch": 2.3927287255293646, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 5989 + }, + { + "epoch": 2.3931282461046743, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 5990 + }, + { + "epoch": 2.393527766679984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 5991 + }, + { + "epoch": 2.3939272872552935, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 5992 + }, + { + "epoch": 2.3943268078306033, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 5993 + }, + { + "epoch": 2.394726328405913, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 5994 + }, + { + "epoch": 2.3951258489812224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 5995 + }, + { + "epoch": 2.395525369556532, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 5996 + }, + { + "epoch": 2.395924890131842, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 5997 + }, + { + "epoch": 2.3963244107071513, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 5998 + }, + { + "epoch": 2.396723931282461, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 5999 + }, + { + "epoch": 2.3971234518577704, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6000 + }, + { + "epoch": 2.39752297243308, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6001 + }, + { + "epoch": 2.39792249300839, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6002 + }, + { + "epoch": 2.3983220135837, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6003 + }, + { + "epoch": 2.398721534159009, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6004 + }, + { + "epoch": 2.399121054734319, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6005 + }, + { + "epoch": 2.3995205753096283, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6006 + }, + { + "epoch": 2.399920095884938, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6007 + }, + { + "epoch": 2.400319616460248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6008 + }, + { + "epoch": 2.400719137035557, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6009 + }, + { + "epoch": 2.401118657610867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6010 + }, + { + "epoch": 2.4015181781861767, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6011 + }, + { + "epoch": 2.401917698761486, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6012 + }, + { + "epoch": 2.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6013 + }, + { + "epoch": 2.4027167399121057, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6014 + }, + { + "epoch": 2.403116260487415, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6015 + }, + { + "epoch": 2.403515781062725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6016 + }, + { + "epoch": 2.403915301638034, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6017 + }, + { + "epoch": 2.404314822213344, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6018 + }, + { + "epoch": 2.4047143427886537, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6019 + }, + { + "epoch": 2.4051138633639635, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6020 + }, + { + "epoch": 2.405513383939273, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6021 + }, + { + "epoch": 2.4059129045145826, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6022 + }, + { + "epoch": 2.406312425089892, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6023 + }, + { + "epoch": 2.4067119456652017, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6024 + }, + { + "epoch": 2.4071114662405115, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6025 + }, + { + "epoch": 2.407510986815821, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6026 + }, + { + "epoch": 2.4079105073911307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6027 + }, + { + "epoch": 2.4083100279664404, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6028 + }, + { + "epoch": 2.40870954854175, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 6029 + }, + { + "epoch": 2.4091090691170596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6030 + }, + { + "epoch": 2.4095085896923694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6031 + }, + { + "epoch": 2.4099081102676787, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6032 + }, + { + "epoch": 2.4103076308429885, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 6033 + }, + { + "epoch": 2.410707151418298, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 6034 + }, + { + "epoch": 2.4111066719936076, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6035 + }, + { + "epoch": 2.4115061925689174, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6036 + }, + { + "epoch": 2.4119057131442267, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6037 + }, + { + "epoch": 2.4123052337195365, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6038 + }, + { + "epoch": 2.4127047542948463, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6039 + }, + { + "epoch": 2.4131042748701557, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6040 + }, + { + "epoch": 2.4135037954454654, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6041 + }, + { + "epoch": 2.4139033160207752, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6042 + }, + { + "epoch": 2.4143028365960846, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 6043 + }, + { + "epoch": 2.4147023571713944, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6044 + }, + { + "epoch": 2.4151018777467037, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6045 + }, + { + "epoch": 2.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6046 + }, + { + "epoch": 2.4159009188973233, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6047 + }, + { + "epoch": 2.416300439472633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 6048 + }, + { + "epoch": 2.4166999600479424, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6049 + }, + { + "epoch": 2.417099480623252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6050 + }, + { + "epoch": 2.4174990011985615, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6051 + }, + { + "epoch": 2.4178985217738713, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6052 + }, + { + "epoch": 2.418298042349181, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6053 + }, + { + "epoch": 2.4186975629244905, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6054 + }, + { + "epoch": 2.4190970834998002, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6055 + }, + { + "epoch": 2.41949660407511, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6056 + }, + { + "epoch": 2.4198961246504194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6057 + }, + { + "epoch": 2.420295645225729, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6058 + }, + { + "epoch": 2.420695165801039, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6059 + }, + { + "epoch": 2.4210946863763483, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6060 + }, + { + "epoch": 2.421494206951658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6061 + }, + { + "epoch": 2.4218937275269674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 6062 + }, + { + "epoch": 2.422293248102277, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6063 + }, + { + "epoch": 2.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6064 + }, + { + "epoch": 2.4230922892528968, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6065 + }, + { + "epoch": 2.423491809828206, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6066 + }, + { + "epoch": 2.423891330403516, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6067 + }, + { + "epoch": 2.4242908509788252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6068 + }, + { + "epoch": 2.424690371554135, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6069 + }, + { + "epoch": 2.425089892129445, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6070 + }, + { + "epoch": 2.425489412704754, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6071 + }, + { + "epoch": 2.425888933280064, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6072 + }, + { + "epoch": 2.4262884538553737, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6073 + }, + { + "epoch": 2.426687974430683, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6074 + }, + { + "epoch": 2.427087495005993, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6075 + }, + { + "epoch": 2.4274870155813026, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6076 + }, + { + "epoch": 2.427886536156612, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6077 + }, + { + "epoch": 2.4282860567319218, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6078 + }, + { + "epoch": 2.428685577307231, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6079 + }, + { + "epoch": 2.429085097882541, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6080 + }, + { + "epoch": 2.4294846184578507, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6081 + }, + { + "epoch": 2.42988413903316, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6082 + }, + { + "epoch": 2.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6083 + }, + { + "epoch": 2.4306831801837796, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6084 + }, + { + "epoch": 2.431082700759089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6085 + }, + { + "epoch": 2.4314822213343987, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6086 + }, + { + "epoch": 2.4318817419097085, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6087 + }, + { + "epoch": 2.432281262485018, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6088 + }, + { + "epoch": 2.4326807830603276, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6089 + }, + { + "epoch": 2.433080303635637, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6090 + }, + { + "epoch": 2.4334798242109468, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6091 + }, + { + "epoch": 2.4338793447862566, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 6092 + }, + { + "epoch": 2.4342788653615663, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6093 + }, + { + "epoch": 2.4346783859368757, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6094 + }, + { + "epoch": 2.4350779065121855, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3453, + "step": 6095 + }, + { + "epoch": 2.435477427087495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 6096 + }, + { + "epoch": 2.4358769476628046, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 6097 + }, + { + "epoch": 2.4362764682381144, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6098 + }, + { + "epoch": 2.4366759888134237, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6099 + }, + { + "epoch": 2.4370755093887335, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6100 + }, + { + "epoch": 2.4374750299640433, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6101 + }, + { + "epoch": 2.4378745505393526, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6102 + }, + { + "epoch": 2.4382740711146624, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6103 + }, + { + "epoch": 2.438673591689972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6104 + }, + { + "epoch": 2.4390731122652816, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6105 + }, + { + "epoch": 2.4394726328405913, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6106 + }, + { + "epoch": 2.4398721534159007, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6107 + }, + { + "epoch": 2.4402716739912105, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6108 + }, + { + "epoch": 2.4406711945665203, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6109 + }, + { + "epoch": 2.44107071514183, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6110 + }, + { + "epoch": 2.4414702357171394, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6111 + }, + { + "epoch": 2.441869756292449, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6112 + }, + { + "epoch": 2.4422692768677585, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6113 + }, + { + "epoch": 2.4426687974430683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6114 + }, + { + "epoch": 2.443068318018378, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6115 + }, + { + "epoch": 2.4434678385936874, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6116 + }, + { + "epoch": 2.443867359168997, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6117 + }, + { + "epoch": 2.444266879744307, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 6118 + }, + { + "epoch": 2.4446664003196164, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6119 + }, + { + "epoch": 2.445065920894926, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6120 + }, + { + "epoch": 2.445465441470236, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6121 + }, + { + "epoch": 2.4458649620455453, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6122 + }, + { + "epoch": 2.446264482620855, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6123 + }, + { + "epoch": 2.4466640031961644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6124 + }, + { + "epoch": 2.447063523771474, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6125 + }, + { + "epoch": 2.447463044346784, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6126 + }, + { + "epoch": 2.4478625649220933, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6127 + }, + { + "epoch": 2.448262085497403, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6128 + }, + { + "epoch": 2.448661606072713, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6129 + }, + { + "epoch": 2.4490611266480222, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6130 + }, + { + "epoch": 2.449460647223332, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6131 + }, + { + "epoch": 2.449860167798642, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6132 + }, + { + "epoch": 2.450259688373951, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6133 + }, + { + "epoch": 2.450659208949261, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6134 + }, + { + "epoch": 2.4510587295245703, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 6135 + }, + { + "epoch": 2.45145825009988, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6136 + }, + { + "epoch": 2.45185777067519, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6137 + }, + { + "epoch": 2.4522572912504996, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6138 + }, + { + "epoch": 2.452656811825809, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6139 + }, + { + "epoch": 2.4530563324011188, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6140 + }, + { + "epoch": 2.453455852976428, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6141 + }, + { + "epoch": 2.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6142 + }, + { + "epoch": 2.4542548941270477, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6143 + }, + { + "epoch": 2.454654414702357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6144 + }, + { + "epoch": 2.455053935277667, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6145 + }, + { + "epoch": 2.4554534558529766, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6146 + }, + { + "epoch": 2.455852976428286, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6147 + }, + { + "epoch": 2.4562524970035957, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6148 + }, + { + "epoch": 2.4566520175789055, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6149 + }, + { + "epoch": 2.457051538154215, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6150 + }, + { + "epoch": 2.4574510587295246, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6151 + }, + { + "epoch": 2.457850579304834, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6152 + }, + { + "epoch": 2.4582500998801438, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6153 + }, + { + "epoch": 2.4586496204554535, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 6154 + }, + { + "epoch": 2.4590491410307633, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6155 + }, + { + "epoch": 2.4594486616060727, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6156 + }, + { + "epoch": 2.4598481821813825, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6157 + }, + { + "epoch": 2.460247702756692, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6158 + }, + { + "epoch": 2.4606472233320016, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6159 + }, + { + "epoch": 2.4610467439073114, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6160 + }, + { + "epoch": 2.4614462644826207, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6161 + }, + { + "epoch": 2.4618457850579305, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6162 + }, + { + "epoch": 2.4622453056332403, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6163 + }, + { + "epoch": 2.4626448262085496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6164 + }, + { + "epoch": 2.4630443467838594, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6165 + }, + { + "epoch": 2.463443867359169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6166 + }, + { + "epoch": 2.4638433879344785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6167 + }, + { + "epoch": 2.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6168 + }, + { + "epoch": 2.4646424290850977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 6169 + }, + { + "epoch": 2.4650419496604075, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6170 + }, + { + "epoch": 2.4654414702357172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6171 + }, + { + "epoch": 2.4658409908110266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6172 + }, + { + "epoch": 2.4662405113863364, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6173 + }, + { + "epoch": 2.466640031961646, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6174 + }, + { + "epoch": 2.4670395525369555, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6175 + }, + { + "epoch": 2.4674390731122653, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6176 + }, + { + "epoch": 2.467838593687575, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6177 + }, + { + "epoch": 2.4682381142628844, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6178 + }, + { + "epoch": 2.468637634838194, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6179 + }, + { + "epoch": 2.4690371554135035, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6180 + }, + { + "epoch": 2.4694366759888133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6181 + }, + { + "epoch": 2.469836196564123, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6182 + }, + { + "epoch": 2.470235717139433, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6183 + }, + { + "epoch": 2.4706352377147422, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6184 + }, + { + "epoch": 2.471034758290052, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6185 + }, + { + "epoch": 2.4714342788653614, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.341, + "step": 6186 + }, + { + "epoch": 2.471833799440671, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6187 + }, + { + "epoch": 2.472233320015981, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6188 + }, + { + "epoch": 2.4726328405912903, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6189 + }, + { + "epoch": 2.4730323611666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6190 + }, + { + "epoch": 2.47343188174191, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6191 + }, + { + "epoch": 2.473831402317219, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6192 + }, + { + "epoch": 2.474230922892529, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6193 + }, + { + "epoch": 2.474630443467839, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6194 + }, + { + "epoch": 2.475029964043148, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6195 + }, + { + "epoch": 2.475429484618458, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6196 + }, + { + "epoch": 2.4758290051937673, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6197 + }, + { + "epoch": 2.476228525769077, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6198 + }, + { + "epoch": 2.476628046344387, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6199 + }, + { + "epoch": 2.4770275669196966, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6200 + }, + { + "epoch": 2.477427087495006, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6201 + }, + { + "epoch": 2.4778266080703157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6202 + }, + { + "epoch": 2.478226128645625, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6203 + }, + { + "epoch": 2.478625649220935, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6204 + }, + { + "epoch": 2.4790251697962447, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6205 + }, + { + "epoch": 2.479424690371554, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6206 + }, + { + "epoch": 2.479824210946864, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6207 + }, + { + "epoch": 2.4802237315221736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6208 + }, + { + "epoch": 2.480623252097483, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6209 + }, + { + "epoch": 2.4810227726727927, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6210 + }, + { + "epoch": 2.4814222932481025, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6211 + }, + { + "epoch": 2.481821813823412, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6212 + }, + { + "epoch": 2.4822213343987216, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6213 + }, + { + "epoch": 2.482620854974031, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 6214 + }, + { + "epoch": 2.4830203755493407, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6215 + }, + { + "epoch": 2.4834198961246505, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6216 + }, + { + "epoch": 2.48381941669996, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 6217 + }, + { + "epoch": 2.4842189372752697, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6218 + }, + { + "epoch": 2.4846184578505794, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6219 + }, + { + "epoch": 2.485017978425889, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6220 + }, + { + "epoch": 2.4854174990011986, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6221 + }, + { + "epoch": 2.4858170195765084, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6222 + }, + { + "epoch": 2.4862165401518177, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6223 + }, + { + "epoch": 2.4866160607271275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6224 + }, + { + "epoch": 2.487015581302437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6225 + }, + { + "epoch": 2.4874151018777466, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6226 + }, + { + "epoch": 2.4878146224530564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6227 + }, + { + "epoch": 2.488214143028366, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6228 + }, + { + "epoch": 2.4886136636036755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6229 + }, + { + "epoch": 2.4890131841789853, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6230 + }, + { + "epoch": 2.4894127047542947, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6231 + }, + { + "epoch": 2.4898122253296044, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6232 + }, + { + "epoch": 2.4902117459049142, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6233 + }, + { + "epoch": 2.4906112664802236, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6234 + }, + { + "epoch": 2.4910107870555334, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 6235 + }, + { + "epoch": 2.491410307630843, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6236 + }, + { + "epoch": 2.4918098282061525, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6237 + }, + { + "epoch": 2.4922093487814623, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6238 + }, + { + "epoch": 2.492608869356772, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6239 + }, + { + "epoch": 2.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6240 + }, + { + "epoch": 2.493407910507391, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6241 + }, + { + "epoch": 2.4938074310827005, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6242 + }, + { + "epoch": 2.4942069516580103, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 6243 + }, + { + "epoch": 2.49460647223332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6244 + }, + { + "epoch": 2.49500599280863, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6245 + }, + { + "epoch": 2.4954055133839392, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 6246 + }, + { + "epoch": 2.495805033959249, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6247 + }, + { + "epoch": 2.4962045545345584, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6248 + }, + { + "epoch": 2.496604075109868, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6249 + }, + { + "epoch": 2.497003595685178, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6250 + }, + { + "epoch": 2.4974031162604873, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6251 + }, + { + "epoch": 2.497802636835797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6252 + }, + { + "epoch": 2.498202157411107, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6253 + }, + { + "epoch": 2.498601677986416, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6254 + }, + { + "epoch": 2.499001198561726, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 6255 + }, + { + "epoch": 2.4994007191370358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6256 + }, + { + "epoch": 2.499800239712345, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6257 + }, + { + "epoch": 2.500199760287655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 6258 + }, + { + "epoch": 2.5005992808629642, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6259 + }, + { + "epoch": 2.500998801438274, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6260 + }, + { + "epoch": 2.501398322013584, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6261 + }, + { + "epoch": 2.5017978425888936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6262 + }, + { + "epoch": 2.502197363164203, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6263 + }, + { + "epoch": 2.5025968837395127, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6264 + }, + { + "epoch": 2.502996404314822, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6265 + }, + { + "epoch": 2.503395924890132, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6266 + }, + { + "epoch": 2.5037954454654416, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6267 + }, + { + "epoch": 2.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6268 + }, + { + "epoch": 2.5045944866160608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6269 + }, + { + "epoch": 2.50499400719137, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6270 + }, + { + "epoch": 2.50539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6271 + }, + { + "epoch": 2.5057930483419897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 6272 + }, + { + "epoch": 2.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6273 + }, + { + "epoch": 2.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6274 + }, + { + "epoch": 2.5069916100679186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6275 + }, + { + "epoch": 2.507391130643228, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 6276 + }, + { + "epoch": 2.5077906512185377, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6277 + }, + { + "epoch": 2.5081901717938475, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6278 + }, + { + "epoch": 2.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6279 + }, + { + "epoch": 2.5089892129444666, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6280 + }, + { + "epoch": 2.5093887335197764, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6281 + }, + { + "epoch": 2.5097882540950858, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6282 + }, + { + "epoch": 2.5101877746703956, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6283 + }, + { + "epoch": 2.5105872952457053, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6284 + }, + { + "epoch": 2.5109868158210147, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6285 + }, + { + "epoch": 2.5113863363963245, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6286 + }, + { + "epoch": 2.511785856971634, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6287 + }, + { + "epoch": 2.5121853775469436, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 6288 + }, + { + "epoch": 2.5125848981222534, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6289 + }, + { + "epoch": 2.512984418697563, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6290 + }, + { + "epoch": 2.5133839392728725, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6291 + }, + { + "epoch": 2.5137834598481823, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6292 + }, + { + "epoch": 2.5141829804234916, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6293 + }, + { + "epoch": 2.5145825009988014, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6294 + }, + { + "epoch": 2.514982021574111, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6295 + }, + { + "epoch": 2.5153815421494206, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6296 + }, + { + "epoch": 2.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6297 + }, + { + "epoch": 2.5161805833000397, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6298 + }, + { + "epoch": 2.5165801038753495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6299 + }, + { + "epoch": 2.5169796244506593, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6300 + }, + { + "epoch": 2.517379145025969, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6301 + }, + { + "epoch": 2.5177786656012784, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6302 + }, + { + "epoch": 2.518178186176588, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6303 + }, + { + "epoch": 2.5185777067518975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6304 + }, + { + "epoch": 2.5189772273272073, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6305 + }, + { + "epoch": 2.519376747902517, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6306 + }, + { + "epoch": 2.519776268477827, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6307 + }, + { + "epoch": 2.520175789053136, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6308 + }, + { + "epoch": 2.520575309628446, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6309 + }, + { + "epoch": 2.5209748302037553, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6310 + }, + { + "epoch": 2.521374350779065, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 6311 + }, + { + "epoch": 2.521773871354375, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6312 + }, + { + "epoch": 2.5221733919296843, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3441, + "step": 6313 + }, + { + "epoch": 2.522572912504994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6314 + }, + { + "epoch": 2.5229724330803034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6315 + }, + { + "epoch": 2.523371953655613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6316 + }, + { + "epoch": 2.523771474230923, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6317 + }, + { + "epoch": 2.5241709948062327, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6318 + }, + { + "epoch": 2.524570515381542, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6319 + }, + { + "epoch": 2.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6320 + }, + { + "epoch": 2.525369556532161, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6321 + }, + { + "epoch": 2.525769077107471, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6322 + }, + { + "epoch": 2.526168597682781, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 6323 + }, + { + "epoch": 2.52656811825809, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6324 + }, + { + "epoch": 2.5269676388334, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6325 + }, + { + "epoch": 2.5273671594087097, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6326 + }, + { + "epoch": 2.527766679984019, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 6327 + }, + { + "epoch": 2.528166200559329, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6328 + }, + { + "epoch": 2.5285657211346386, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6329 + }, + { + "epoch": 2.528965241709948, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6330 + }, + { + "epoch": 2.5293647622852578, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6331 + }, + { + "epoch": 2.529764282860567, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6332 + }, + { + "epoch": 2.530163803435877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6333 + }, + { + "epoch": 2.5305633240111867, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3395, + "step": 6334 + }, + { + "epoch": 2.5309628445864965, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3428, + "step": 6335 + }, + { + "epoch": 2.531362365161806, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6336 + }, + { + "epoch": 2.5317618857371156, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6337 + }, + { + "epoch": 2.532161406312425, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6338 + }, + { + "epoch": 2.5325609268877347, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6339 + }, + { + "epoch": 2.5329604474630445, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6340 + }, + { + "epoch": 2.533359968038354, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6341 + }, + { + "epoch": 2.5337594886136636, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6342 + }, + { + "epoch": 2.534159009188973, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6343 + }, + { + "epoch": 2.5345585297642828, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6344 + }, + { + "epoch": 2.5349580503395925, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6345 + }, + { + "epoch": 2.5353575709149023, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6346 + }, + { + "epoch": 2.5357570914902117, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 6347 + }, + { + "epoch": 2.5361566120655215, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6348 + }, + { + "epoch": 2.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6349 + }, + { + "epoch": 2.5369556532161406, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 6350 + }, + { + "epoch": 2.5373551737914504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6351 + }, + { + "epoch": 2.53775469436676, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6352 + }, + { + "epoch": 2.5381542149420695, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6353 + }, + { + "epoch": 2.5385537355173793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6354 + }, + { + "epoch": 2.5389532560926886, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6355 + }, + { + "epoch": 2.5393527766679984, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6356 + }, + { + "epoch": 2.539752297243308, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6357 + }, + { + "epoch": 2.5401518178186175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6358 + }, + { + "epoch": 2.5405513383939273, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 6359 + }, + { + "epoch": 2.5409508589692367, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6360 + }, + { + "epoch": 2.5413503795445465, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6361 + }, + { + "epoch": 2.5417499001198562, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 6362 + }, + { + "epoch": 2.542149420695166, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6363 + }, + { + "epoch": 2.5425489412704754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6364 + }, + { + "epoch": 2.542948461845785, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6365 + }, + { + "epoch": 2.5433479824210945, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 6366 + }, + { + "epoch": 2.5437475029964043, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6367 + }, + { + "epoch": 2.544147023571714, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6368 + }, + { + "epoch": 2.5445465441470234, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6369 + }, + { + "epoch": 2.544946064722333, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6370 + }, + { + "epoch": 2.545345585297643, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6371 + }, + { + "epoch": 2.5457451058729523, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6372 + }, + { + "epoch": 2.546144626448262, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6373 + }, + { + "epoch": 2.546544147023572, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6374 + }, + { + "epoch": 2.5469436675988812, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6375 + }, + { + "epoch": 2.547343188174191, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6376 + }, + { + "epoch": 2.5477427087495004, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6377 + }, + { + "epoch": 2.54814222932481, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6378 + }, + { + "epoch": 2.54854174990012, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6379 + }, + { + "epoch": 2.5489412704754297, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6380 + }, + { + "epoch": 2.549340791050739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6381 + }, + { + "epoch": 2.549740311626049, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6382 + }, + { + "epoch": 2.550139832201358, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6383 + }, + { + "epoch": 2.550539352776668, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6384 + }, + { + "epoch": 2.5509388733519778, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6385 + }, + { + "epoch": 2.551338393927287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 6386 + }, + { + "epoch": 2.551737914502597, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6387 + }, + { + "epoch": 2.5521374350779062, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 6388 + }, + { + "epoch": 2.552536955653216, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6389 + }, + { + "epoch": 2.552936476228526, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6390 + }, + { + "epoch": 2.5533359968038356, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6391 + }, + { + "epoch": 2.553735517379145, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6392 + }, + { + "epoch": 2.5541350379544547, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6393 + }, + { + "epoch": 2.554534558529764, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6394 + }, + { + "epoch": 2.554934079105074, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 6395 + }, + { + "epoch": 2.5553335996803836, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6396 + }, + { + "epoch": 2.5557331202556934, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6397 + }, + { + "epoch": 2.556132640831003, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6398 + }, + { + "epoch": 2.5565321614063126, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6399 + }, + { + "epoch": 2.556931681981622, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6400 + }, + { + "epoch": 2.5573312025569317, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6401 + }, + { + "epoch": 2.5577307231322415, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6402 + }, + { + "epoch": 2.558130243707551, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6403 + }, + { + "epoch": 2.5585297642828606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6404 + }, + { + "epoch": 2.55892928485817, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 6405 + }, + { + "epoch": 2.5593288054334797, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 6406 + }, + { + "epoch": 2.5597283260087895, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 6407 + }, + { + "epoch": 2.5601278465840993, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6408 + }, + { + "epoch": 2.5605273671594087, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6409 + }, + { + "epoch": 2.5609268877347184, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 6410 + }, + { + "epoch": 2.561326408310028, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6411 + }, + { + "epoch": 2.5617259288853376, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3492, + "step": 6412 + }, + { + "epoch": 2.5621254494606474, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6413 + }, + { + "epoch": 2.5625249700359567, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6414 + }, + { + "epoch": 2.5629244906112665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6415 + }, + { + "epoch": 2.5633240111865763, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6416 + }, + { + "epoch": 2.5637235317618856, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3467, + "step": 6417 + }, + { + "epoch": 2.5641230523371954, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6418 + }, + { + "epoch": 2.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 6419 + }, + { + "epoch": 2.5649220934878145, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6420 + }, + { + "epoch": 2.5653216140631243, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6421 + }, + { + "epoch": 2.5657211346384337, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6422 + }, + { + "epoch": 2.5661206552137434, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6423 + }, + { + "epoch": 2.5665201757890532, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6424 + }, + { + "epoch": 2.566919696364363, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6425 + }, + { + "epoch": 2.5673192169396724, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6426 + }, + { + "epoch": 2.567718737514982, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 6427 + }, + { + "epoch": 2.5681182580902915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6428 + }, + { + "epoch": 2.5685177786656013, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6429 + }, + { + "epoch": 2.568917299240911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6430 + }, + { + "epoch": 2.5693168198162204, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6431 + }, + { + "epoch": 2.56971634039153, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6432 + }, + { + "epoch": 2.5701158609668395, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6433 + }, + { + "epoch": 2.5705153815421493, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 6434 + }, + { + "epoch": 2.570914902117459, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6435 + }, + { + "epoch": 2.571314422692769, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6436 + }, + { + "epoch": 2.5717139432680782, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6437 + }, + { + "epoch": 2.572113463843388, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6438 + }, + { + "epoch": 2.5725129844186974, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6439 + }, + { + "epoch": 2.572912504994007, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6440 + }, + { + "epoch": 2.573312025569317, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6441 + }, + { + "epoch": 2.5737115461446267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6442 + }, + { + "epoch": 2.574111066719936, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6443 + }, + { + "epoch": 2.574510587295246, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 6444 + }, + { + "epoch": 2.574910107870555, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6445 + }, + { + "epoch": 2.575309628445865, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6446 + }, + { + "epoch": 2.5757091490211748, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6447 + }, + { + "epoch": 2.576108669596484, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6448 + }, + { + "epoch": 2.576508190171794, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6449 + }, + { + "epoch": 2.5769077107471032, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6450 + }, + { + "epoch": 2.577307231322413, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 6451 + }, + { + "epoch": 2.577706751897723, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6452 + }, + { + "epoch": 2.5781062724730326, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6453 + }, + { + "epoch": 2.578505793048342, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6454 + }, + { + "epoch": 2.5789053136236517, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6455 + }, + { + "epoch": 2.579304834198961, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6456 + }, + { + "epoch": 2.579704354774271, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3413, + "step": 6457 + }, + { + "epoch": 2.5801038753495806, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6458 + }, + { + "epoch": 2.58050339592489, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6459 + }, + { + "epoch": 2.5809029165001998, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6460 + }, + { + "epoch": 2.5813024370755095, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6461 + }, + { + "epoch": 2.581701957650819, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6462 + }, + { + "epoch": 2.5821014782261287, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6463 + }, + { + "epoch": 2.5825009988014385, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 6464 + }, + { + "epoch": 2.582900519376748, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6465 + }, + { + "epoch": 2.5833000399520576, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6466 + }, + { + "epoch": 2.583699560527367, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6467 + }, + { + "epoch": 2.5840990811026767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6468 + }, + { + "epoch": 2.5844986016779865, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 6469 + }, + { + "epoch": 2.5848981222532963, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6470 + }, + { + "epoch": 2.5852976428286056, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6471 + }, + { + "epoch": 2.5856971634039154, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6472 + }, + { + "epoch": 2.5860966839792248, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6473 + }, + { + "epoch": 2.5864962045545346, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6474 + }, + { + "epoch": 2.5868957251298443, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 6475 + }, + { + "epoch": 2.5872952457051537, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6476 + }, + { + "epoch": 2.5876947662804635, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6477 + }, + { + "epoch": 2.588094286855773, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6478 + }, + { + "epoch": 2.5884938074310826, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 6479 + }, + { + "epoch": 2.5888933280063924, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6480 + }, + { + "epoch": 2.589292848581702, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6481 + }, + { + "epoch": 2.5896923691570115, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6482 + }, + { + "epoch": 2.5900918897323213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 6483 + }, + { + "epoch": 2.5904914103076306, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 6484 + }, + { + "epoch": 2.5908909308829404, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6485 + }, + { + "epoch": 2.59129045145825, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6486 + }, + { + "epoch": 2.59168997203356, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 6487 + }, + { + "epoch": 2.5920894926088693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6488 + }, + { + "epoch": 2.592489013184179, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6489 + }, + { + "epoch": 2.5928885337594885, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6490 + }, + { + "epoch": 2.5932880543347983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 6491 + }, + { + "epoch": 2.593687574910108, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6492 + }, + { + "epoch": 2.5940870954854174, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3408, + "step": 6493 + }, + { + "epoch": 2.594486616060727, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 6494 + }, + { + "epoch": 2.5948861366360365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6495 + }, + { + "epoch": 2.5952856572113463, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 6496 + }, + { + "epoch": 2.595685177786656, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6497 + }, + { + "epoch": 2.596084698361966, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6498 + }, + { + "epoch": 2.596484218937275, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6499 + }, + { + "epoch": 2.596883739512585, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6500 + }, + { + "epoch": 2.5972832600878943, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6501 + }, + { + "epoch": 2.597682780663204, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6502 + }, + { + "epoch": 2.598082301238514, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 6503 + }, + { + "epoch": 2.5984818218138233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6504 + }, + { + "epoch": 2.598881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6505 + }, + { + "epoch": 2.599280862964443, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6506 + }, + { + "epoch": 2.599680383539752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6507 + }, + { + "epoch": 2.600079904115062, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3432, + "step": 6508 + }, + { + "epoch": 2.6004794246903717, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6509 + }, + { + "epoch": 2.600878945265681, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6510 + }, + { + "epoch": 2.601278465840991, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6511 + }, + { + "epoch": 2.6016779864163, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6512 + }, + { + "epoch": 2.60207750699161, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6513 + }, + { + "epoch": 2.60247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6514 + }, + { + "epoch": 2.6028765481422296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6515 + }, + { + "epoch": 2.603276068717539, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6516 + }, + { + "epoch": 2.6036755892928487, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 6517 + }, + { + "epoch": 2.604075109868158, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6518 + }, + { + "epoch": 2.604474630443468, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6519 + }, + { + "epoch": 2.6048741510187776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6520 + }, + { + "epoch": 2.605273671594087, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6521 + }, + { + "epoch": 2.6056731921693967, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6522 + }, + { + "epoch": 2.606072712744706, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3415, + "step": 6523 + }, + { + "epoch": 2.606472233320016, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 6524 + }, + { + "epoch": 2.6068717538953257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 6525 + }, + { + "epoch": 2.6072712744706354, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6526 + }, + { + "epoch": 2.607670795045945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6527 + }, + { + "epoch": 2.6080703156212546, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6528 + }, + { + "epoch": 2.608469836196564, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 6529 + }, + { + "epoch": 2.6088693567718737, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 6530 + }, + { + "epoch": 2.6092688773471835, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6531 + }, + { + "epoch": 2.6096683979224933, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6532 + }, + { + "epoch": 2.6100679184978026, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6533 + }, + { + "epoch": 2.6104674390731124, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6534 + }, + { + "epoch": 2.6108669596484217, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6535 + }, + { + "epoch": 2.6112664802237315, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 6536 + }, + { + "epoch": 2.6116660007990413, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6537 + }, + { + "epoch": 2.6120655213743507, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6538 + }, + { + "epoch": 2.6124650419496604, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6539 + }, + { + "epoch": 2.61286456252497, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6540 + }, + { + "epoch": 2.6132640831002796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 6541 + }, + { + "epoch": 2.6136636036755894, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6542 + }, + { + "epoch": 2.614063124250899, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6543 + }, + { + "epoch": 2.6144626448262085, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6544 + }, + { + "epoch": 2.6148621654015183, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 6545 + }, + { + "epoch": 2.6152616859768276, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6546 + }, + { + "epoch": 2.6156612065521374, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6547 + }, + { + "epoch": 2.616060727127447, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3414, + "step": 6548 + }, + { + "epoch": 2.6164602477027565, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6549 + }, + { + "epoch": 2.6168597682780663, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6550 + }, + { + "epoch": 2.617259288853376, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6551 + }, + { + "epoch": 2.6176588094286855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6552 + }, + { + "epoch": 2.6180583300039952, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6553 + }, + { + "epoch": 2.618457850579305, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 6554 + }, + { + "epoch": 2.6188573711546144, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6555 + }, + { + "epoch": 2.619256891729924, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6556 + }, + { + "epoch": 2.6196564123052335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6557 + }, + { + "epoch": 2.6200559328805433, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6558 + }, + { + "epoch": 2.620455453455853, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 6559 + }, + { + "epoch": 2.620854974031163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6560 + }, + { + "epoch": 2.621254494606472, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6561 + }, + { + "epoch": 2.621654015181782, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6562 + }, + { + "epoch": 2.6220535357570913, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 6563 + }, + { + "epoch": 2.622453056332401, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 6564 + }, + { + "epoch": 2.622852576907711, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6565 + }, + { + "epoch": 2.6232520974830202, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6566 + }, + { + "epoch": 2.62365161805833, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 6567 + }, + { + "epoch": 2.6240511386336394, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6568 + }, + { + "epoch": 2.624450659208949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6569 + }, + { + "epoch": 2.624850179784259, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 6570 + }, + { + "epoch": 2.6252497003595687, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6571 + }, + { + "epoch": 2.625649220934878, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6572 + }, + { + "epoch": 2.626048741510188, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6573 + }, + { + "epoch": 2.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 6574 + }, + { + "epoch": 2.626847782660807, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6575 + }, + { + "epoch": 2.6272473032361168, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6576 + }, + { + "epoch": 2.6276468238114266, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 6577 + }, + { + "epoch": 2.628046344386736, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6578 + }, + { + "epoch": 2.6284458649620457, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6579 + }, + { + "epoch": 2.628845385537355, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6580 + }, + { + "epoch": 2.629244906112665, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 6581 + }, + { + "epoch": 2.6296444266879746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6582 + }, + { + "epoch": 2.630043947263284, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6583 + }, + { + "epoch": 2.6304434678385937, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6584 + }, + { + "epoch": 2.630842988413903, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 6585 + }, + { + "epoch": 2.631242508989213, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6586 + }, + { + "epoch": 2.6316420295645226, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6587 + }, + { + "epoch": 2.6320415501398324, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6588 + }, + { + "epoch": 2.6324410707151418, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6589 + }, + { + "epoch": 2.6328405912904516, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 6590 + }, + { + "epoch": 2.633240111865761, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6591 + }, + { + "epoch": 2.6336396324410707, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6592 + }, + { + "epoch": 2.6340391530163805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6593 + }, + { + "epoch": 2.63443867359169, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6594 + }, + { + "epoch": 2.6348381941669996, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6595 + }, + { + "epoch": 2.6352377147423094, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6596 + }, + { + "epoch": 2.6356372353176187, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6597 + }, + { + "epoch": 2.6360367558929285, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 6598 + }, + { + "epoch": 2.6364362764682383, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6599 + }, + { + "epoch": 2.6368357970435476, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6600 + }, + { + "epoch": 2.6372353176188574, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6601 + }, + { + "epoch": 2.6376348381941668, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6602 + }, + { + "epoch": 2.6380343587694766, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6603 + }, + { + "epoch": 2.6384338793447863, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6604 + }, + { + "epoch": 2.638833399920096, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6605 + }, + { + "epoch": 2.6392329204954055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6606 + }, + { + "epoch": 2.6396324410707153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6607 + }, + { + "epoch": 2.6400319616460246, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6608 + }, + { + "epoch": 2.6404314822213344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 6609 + }, + { + "epoch": 2.640831002796644, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6610 + }, + { + "epoch": 2.6412305233719535, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6611 + }, + { + "epoch": 2.6416300439472633, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 6612 + }, + { + "epoch": 2.6420295645225726, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6613 + }, + { + "epoch": 2.6424290850978824, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6614 + }, + { + "epoch": 2.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6615 + }, + { + "epoch": 2.643228126248502, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6616 + }, + { + "epoch": 2.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6617 + }, + { + "epoch": 2.644027167399121, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6618 + }, + { + "epoch": 2.6444266879744305, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6619 + }, + { + "epoch": 2.6448262085497403, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6620 + }, + { + "epoch": 2.64522572912505, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6621 + }, + { + "epoch": 2.64562524970036, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6622 + }, + { + "epoch": 2.646024770275669, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6623 + }, + { + "epoch": 2.646424290850979, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6624 + }, + { + "epoch": 2.6468238114262883, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6625 + }, + { + "epoch": 2.647223332001598, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6626 + }, + { + "epoch": 2.647622852576908, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6627 + }, + { + "epoch": 2.6480223731522172, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6628 + }, + { + "epoch": 2.648421893727527, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6629 + }, + { + "epoch": 2.6488214143028364, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6630 + }, + { + "epoch": 2.649220934878146, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6631 + }, + { + "epoch": 2.649620455453456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 6632 + }, + { + "epoch": 2.6500199760287657, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6633 + }, + { + "epoch": 2.650419496604075, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 6634 + }, + { + "epoch": 2.650819017179385, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 6635 + }, + { + "epoch": 2.651218537754694, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 6636 + }, + { + "epoch": 2.651618058330004, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6637 + }, + { + "epoch": 2.6520175789053138, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 6638 + }, + { + "epoch": 2.652417099480623, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 6639 + }, + { + "epoch": 2.652816620055933, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6640 + }, + { + "epoch": 2.6532161406312427, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6641 + }, + { + "epoch": 2.653615661206552, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6642 + }, + { + "epoch": 2.654015181781862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6643 + }, + { + "epoch": 2.6544147023571716, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6644 + }, + { + "epoch": 2.654814222932481, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6645 + }, + { + "epoch": 2.6552137435077907, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6646 + }, + { + "epoch": 2.6556132640831, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 6647 + }, + { + "epoch": 2.65601278465841, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6648 + }, + { + "epoch": 2.6564123052337196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 6649 + }, + { + "epoch": 2.6568118258090294, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6650 + }, + { + "epoch": 2.6572113463843388, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6651 + }, + { + "epoch": 2.6576108669596485, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 6652 + }, + { + "epoch": 2.658010387534958, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6653 + }, + { + "epoch": 2.6584099081102677, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6654 + }, + { + "epoch": 2.6588094286855775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6655 + }, + { + "epoch": 2.659208949260887, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6656 + }, + { + "epoch": 2.6596084698361966, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 6657 + }, + { + "epoch": 2.660007990411506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6658 + }, + { + "epoch": 2.6604075109868157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6659 + }, + { + "epoch": 2.6608070315621255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6660 + }, + { + "epoch": 2.6612065521374353, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6661 + }, + { + "epoch": 2.6616060727127446, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6662 + }, + { + "epoch": 2.6620055932880544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6663 + }, + { + "epoch": 2.6624051138633638, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 6664 + }, + { + "epoch": 2.6628046344386735, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6665 + }, + { + "epoch": 2.6632041550139833, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 6666 + }, + { + "epoch": 2.663603675589293, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6667 + }, + { + "epoch": 2.6640031961646025, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6668 + }, + { + "epoch": 2.6644027167399122, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6669 + }, + { + "epoch": 2.6648022373152216, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 6670 + }, + { + "epoch": 2.6652017578905314, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6671 + }, + { + "epoch": 2.665601278465841, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6672 + }, + { + "epoch": 2.6660007990411505, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6673 + }, + { + "epoch": 2.6664003196164603, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6674 + }, + { + "epoch": 2.6667998401917696, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6675 + }, + { + "epoch": 2.6671993607670794, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6676 + }, + { + "epoch": 2.667598881342389, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6677 + }, + { + "epoch": 2.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6678 + }, + { + "epoch": 2.6683979224930083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6679 + }, + { + "epoch": 2.668797443068318, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6680 + }, + { + "epoch": 2.6691969636436275, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 6681 + }, + { + "epoch": 2.6695964842189372, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6682 + }, + { + "epoch": 2.669996004794247, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6683 + }, + { + "epoch": 2.6703955253695564, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6684 + }, + { + "epoch": 2.670795045944866, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6685 + }, + { + "epoch": 2.671194566520176, + "grad_norm": 0.16015625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6686 + }, + { + "epoch": 2.6715940870954853, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 6687 + }, + { + "epoch": 2.671993607670795, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6688 + }, + { + "epoch": 2.672393128246105, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6689 + }, + { + "epoch": 2.672792648821414, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6690 + }, + { + "epoch": 2.673192169396724, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6691 + }, + { + "epoch": 2.6735916899720333, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6692 + }, + { + "epoch": 2.673991210547343, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 6693 + }, + { + "epoch": 2.674390731122653, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 6694 + }, + { + "epoch": 2.6747902516979627, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6695 + }, + { + "epoch": 2.675189772273272, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6696 + }, + { + "epoch": 2.675589292848582, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6697 + }, + { + "epoch": 2.675988813423891, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 6698 + }, + { + "epoch": 2.676388333999201, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6699 + }, + { + "epoch": 2.6767878545745107, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 6700 + }, + { + "epoch": 2.67718737514982, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6701 + }, + { + "epoch": 2.67758689572513, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 6702 + }, + { + "epoch": 2.677986416300439, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 6703 + }, + { + "epoch": 2.678385936875749, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 6704 + }, + { + "epoch": 2.678785457451059, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6705 + }, + { + "epoch": 2.6791849780263686, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6706 + }, + { + "epoch": 2.679584498601678, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6707 + }, + { + "epoch": 2.6799840191769877, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6708 + }, + { + "epoch": 2.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6709 + }, + { + "epoch": 2.680783060327607, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6710 + }, + { + "epoch": 2.6811825809029166, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6711 + }, + { + "epoch": 2.6815821014782264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 6712 + }, + { + "epoch": 2.6819816220535357, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 6713 + }, + { + "epoch": 2.6823811426288455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6714 + }, + { + "epoch": 2.682780663204155, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 6715 + }, + { + "epoch": 2.6831801837794647, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6716 + }, + { + "epoch": 2.6835797043547744, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6717 + }, + { + "epoch": 2.683979224930084, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6718 + }, + { + "epoch": 2.6843787455053936, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 6719 + }, + { + "epoch": 2.684778266080703, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3471, + "step": 6720 + }, + { + "epoch": 2.6851777866560127, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6721 + }, + { + "epoch": 2.6855773072313225, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6722 + }, + { + "epoch": 2.6859768278066323, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6723 + }, + { + "epoch": 2.6863763483819416, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 6724 + }, + { + "epoch": 2.6867758689572514, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6725 + }, + { + "epoch": 2.6871753895325607, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6726 + }, + { + "epoch": 2.6875749101078705, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6727 + }, + { + "epoch": 2.6879744306831803, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3448, + "step": 6728 + }, + { + "epoch": 2.6883739512584897, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 6729 + }, + { + "epoch": 2.6887734718337994, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6730 + }, + { + "epoch": 2.6891729924091092, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6731 + }, + { + "epoch": 2.6895725129844186, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6732 + }, + { + "epoch": 2.6899720335597284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 6733 + }, + { + "epoch": 2.690371554135038, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6734 + }, + { + "epoch": 2.6907710747103475, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6735 + }, + { + "epoch": 2.6911705952856573, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 6736 + }, + { + "epoch": 2.6915701158609666, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6737 + }, + { + "epoch": 2.6919696364362764, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6738 + }, + { + "epoch": 2.692369157011586, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6739 + }, + { + "epoch": 2.692768677586896, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 6740 + }, + { + "epoch": 2.6931681981622053, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 6741 + }, + { + "epoch": 2.693567718737515, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 6742 + }, + { + "epoch": 2.6939672393128244, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6743 + }, + { + "epoch": 2.6943667598881342, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6744 + }, + { + "epoch": 2.694766280463444, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6745 + }, + { + "epoch": 2.6951658010387534, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 6746 + }, + { + "epoch": 2.695565321614063, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6747 + }, + { + "epoch": 2.6959648421893725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6748 + }, + { + "epoch": 2.6963643627646823, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 6749 + }, + { + "epoch": 2.696763883339992, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 6750 + }, + { + "epoch": 2.697163403915302, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6751 + }, + { + "epoch": 2.697562924490611, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 6752 + }, + { + "epoch": 2.697962445065921, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6753 + }, + { + "epoch": 2.6983619656412303, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 6754 + }, + { + "epoch": 2.69876148621654, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 6755 + }, + { + "epoch": 2.69916100679185, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 6756 + }, + { + "epoch": 2.6995605273671597, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6757 + }, + { + "epoch": 2.699960047942469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6758 + }, + { + "epoch": 2.700359568517779, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6759 + }, + { + "epoch": 2.700759089093088, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6760 + }, + { + "epoch": 2.701158609668398, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 6761 + }, + { + "epoch": 2.7015581302437077, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 6762 + }, + { + "epoch": 2.701957650819017, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6763 + }, + { + "epoch": 2.702357171394327, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6764 + }, + { + "epoch": 2.702756691969636, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 6765 + }, + { + "epoch": 2.703156212544946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 6766 + }, + { + "epoch": 2.7035557331202558, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6767 + }, + { + "epoch": 2.7039552536955656, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 6768 + }, + { + "epoch": 2.704354774270875, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6769 + }, + { + "epoch": 2.7047542948461847, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 6770 + }, + { + "epoch": 2.705153815421494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 6771 + }, + { + "epoch": 2.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6772 + }, + { + "epoch": 2.7059528565721136, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6773 + }, + { + "epoch": 2.706352377147423, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 6774 + }, + { + "epoch": 2.7067518977227327, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 6775 + }, + { + "epoch": 2.7071514182980425, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6776 + }, + { + "epoch": 2.707550938873352, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 6777 + }, + { + "epoch": 2.7079504594486616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6778 + }, + { + "epoch": 2.7083499800239714, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6779 + }, + { + "epoch": 2.7087495005992808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6780 + }, + { + "epoch": 2.7091490211745906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 6781 + }, + { + "epoch": 2.7095485417499, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6782 + }, + { + "epoch": 2.7099480623252097, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6783 + }, + { + "epoch": 2.7103475829005195, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 6784 + }, + { + "epoch": 2.7107471034758293, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 6785 + }, + { + "epoch": 2.7111466240511386, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 6786 + }, + { + "epoch": 2.7115461446264484, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 6787 + }, + { + "epoch": 2.7119456652017577, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6788 + }, + { + "epoch": 2.7123451857770675, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6789 + }, + { + "epoch": 2.7127447063523773, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6790 + }, + { + "epoch": 2.7131442269276866, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 6791 + }, + { + "epoch": 2.7135437475029964, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6792 + }, + { + "epoch": 2.7139432680783058, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6793 + }, + { + "epoch": 2.7143427886536156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6794 + }, + { + "epoch": 2.7147423092289253, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6795 + }, + { + "epoch": 2.715141829804235, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 6796 + }, + { + "epoch": 2.7155413503795445, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6797 + }, + { + "epoch": 2.7159408709548543, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 6798 + }, + { + "epoch": 2.7163403915301636, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 6799 + }, + { + "epoch": 2.7167399121054734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 6800 + }, + { + "epoch": 2.717139432680783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 6801 + }, + { + "epoch": 2.717538953256093, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6802 + }, + { + "epoch": 2.7179384738314023, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 6803 + }, + { + "epoch": 2.718337994406712, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 6804 + }, + { + "epoch": 2.7187375149820214, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6805 + }, + { + "epoch": 2.719137035557331, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6806 + }, + { + "epoch": 2.719536556132641, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6807 + }, + { + "epoch": 2.7199360767079503, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 6808 + }, + { + "epoch": 2.72033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 6809 + }, + { + "epoch": 2.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6810 + }, + { + "epoch": 2.7211346384338793, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6811 + }, + { + "epoch": 2.721534159009189, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 6812 + }, + { + "epoch": 2.721933679584499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6813 + }, + { + "epoch": 2.722333200159808, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 6814 + }, + { + "epoch": 2.722732720735118, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 6815 + }, + { + "epoch": 2.7231322413104273, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6816 + }, + { + "epoch": 2.723531761885737, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 6817 + }, + { + "epoch": 2.723931282461047, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 6818 + }, + { + "epoch": 2.724330803036356, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6819 + }, + { + "epoch": 2.724730323611666, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6820 + }, + { + "epoch": 2.725129844186976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 6821 + }, + { + "epoch": 2.725529364762285, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 6822 + }, + { + "epoch": 2.725928885337595, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6823 + }, + { + "epoch": 2.7263284059129047, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6824 + }, + { + "epoch": 2.726727926488214, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6825 + }, + { + "epoch": 2.727127447063524, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6826 + }, + { + "epoch": 2.727526967638833, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 6827 + }, + { + "epoch": 2.727926488214143, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 6828 + }, + { + "epoch": 2.7283260087894528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6829 + }, + { + "epoch": 2.7287255293647625, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6830 + }, + { + "epoch": 2.729125049940072, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 6831 + }, + { + "epoch": 2.7295245705153817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 6832 + }, + { + "epoch": 2.729924091090691, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 6833 + }, + { + "epoch": 2.730323611666001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 6834 + }, + { + "epoch": 2.7307231322413106, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 6835 + }, + { + "epoch": 2.73112265281662, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 6836 + }, + { + "epoch": 2.7315221733919297, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 6837 + }, + { + "epoch": 2.731921693967239, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 6838 + }, + { + "epoch": 2.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 6839 + }, + { + "epoch": 2.7327207351178586, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 6840 + }, + { + "epoch": 2.7331202556931684, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 6841 + }, + { + "epoch": 2.7335197762684778, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 6842 + }, + { + "epoch": 2.7339192968437875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 6843 + }, + { + "epoch": 2.734318817419097, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6844 + }, + { + "epoch": 2.7347183379944067, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6845 + }, + { + "epoch": 2.7351178585697165, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 6846 + }, + { + "epoch": 2.7355173791450262, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 6847 + }, + { + "epoch": 2.7359168997203356, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6848 + }, + { + "epoch": 2.7363164202956454, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 6849 + }, + { + "epoch": 2.7367159408709547, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 6850 + }, + { + "epoch": 2.7371154614462645, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6851 + }, + { + "epoch": 2.7375149820215743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6852 + }, + { + "epoch": 2.7379145025968836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6853 + }, + { + "epoch": 2.7383140231721934, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 6854 + }, + { + "epoch": 2.7387135437475028, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 6855 + }, + { + "epoch": 2.7391130643228125, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 6856 + }, + { + "epoch": 2.7395125848981223, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 6857 + }, + { + "epoch": 2.739912105473432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 6858 + }, + { + "epoch": 2.7403116260487415, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6859 + }, + { + "epoch": 2.7407111466240512, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 6860 + }, + { + "epoch": 2.7411106671993606, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 6861 + }, + { + "epoch": 2.7415101877746704, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6862 + }, + { + "epoch": 2.74190970834998, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6863 + }, + { + "epoch": 2.7423092289252895, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6864 + }, + { + "epoch": 2.7427087495005993, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6865 + }, + { + "epoch": 2.743108270075909, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 6866 + }, + { + "epoch": 2.7435077906512184, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 6867 + }, + { + "epoch": 2.743907311226528, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 6868 + }, + { + "epoch": 2.744306831801838, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 6869 + }, + { + "epoch": 2.7447063523771473, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6870 + }, + { + "epoch": 2.745105872952457, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6871 + }, + { + "epoch": 2.7455053935277665, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6872 + }, + { + "epoch": 2.7459049141030762, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 6873 + }, + { + "epoch": 2.746304434678386, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 6874 + }, + { + "epoch": 2.746703955253696, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 6875 + }, + { + "epoch": 2.747103475829005, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6876 + }, + { + "epoch": 2.747502996404315, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 6877 + }, + { + "epoch": 2.7479025169796243, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 6878 + }, + { + "epoch": 2.748302037554934, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 6879 + }, + { + "epoch": 2.748701558130244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6880 + }, + { + "epoch": 2.749101078705553, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6881 + }, + { + "epoch": 2.749500599280863, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6882 + }, + { + "epoch": 2.7499001198561723, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 6883 + }, + { + "epoch": 2.750299640431482, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 6884 + }, + { + "epoch": 2.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6885 + }, + { + "epoch": 2.7510986815821017, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6886 + }, + { + "epoch": 2.751498202157411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6887 + }, + { + "epoch": 2.751897722732721, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 6888 + }, + { + "epoch": 2.75229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 6889 + }, + { + "epoch": 2.75269676388334, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 6890 + }, + { + "epoch": 2.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6891 + }, + { + "epoch": 2.7534958050339595, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 6892 + }, + { + "epoch": 2.753895325609269, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 6893 + }, + { + "epoch": 2.7542948461845786, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 6894 + }, + { + "epoch": 2.754694366759888, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 6895 + }, + { + "epoch": 2.7550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 6896 + }, + { + "epoch": 2.7554934079105076, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6897 + }, + { + "epoch": 2.755892928485817, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6898 + }, + { + "epoch": 2.7562924490611267, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 6899 + }, + { + "epoch": 2.756691969636436, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 6900 + }, + { + "epoch": 2.757091490211746, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6901 + }, + { + "epoch": 2.7574910107870556, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 6902 + }, + { + "epoch": 2.7578905313623654, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 6903 + }, + { + "epoch": 2.7582900519376747, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 6904 + }, + { + "epoch": 2.7586895725129845, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6905 + }, + { + "epoch": 2.759089093088294, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 6906 + }, + { + "epoch": 2.7594886136636037, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6907 + }, + { + "epoch": 2.7598881342389134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 6908 + }, + { + "epoch": 2.760287654814223, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6909 + }, + { + "epoch": 2.7606871753895326, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6910 + }, + { + "epoch": 2.7610866959648424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 6911 + }, + { + "epoch": 2.7614862165401517, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 6912 + }, + { + "epoch": 2.7618857371154615, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 6913 + }, + { + "epoch": 2.7622852576907713, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 6914 + }, + { + "epoch": 2.7626847782660806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 6915 + }, + { + "epoch": 2.7630842988413904, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 6916 + }, + { + "epoch": 2.7634838194166997, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 6917 + }, + { + "epoch": 2.7638833399920095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6918 + }, + { + "epoch": 2.7642828605673193, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 6919 + }, + { + "epoch": 2.764682381142629, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 6920 + }, + { + "epoch": 2.7650819017179384, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 6921 + }, + { + "epoch": 2.7654814222932482, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 6922 + }, + { + "epoch": 2.7658809428685576, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 6923 + }, + { + "epoch": 2.7662804634438674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 6924 + }, + { + "epoch": 2.766679984019177, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 6925 + }, + { + "epoch": 2.7670795045944865, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6926 + }, + { + "epoch": 2.7674790251697963, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 6927 + }, + { + "epoch": 2.7678785457451056, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 6928 + }, + { + "epoch": 2.7682780663204154, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 6929 + }, + { + "epoch": 2.768677586895725, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 6930 + }, + { + "epoch": 2.769077107471035, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 6931 + }, + { + "epoch": 2.7694766280463443, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6932 + }, + { + "epoch": 2.769876148621654, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 6933 + }, + { + "epoch": 2.7702756691969634, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 6934 + }, + { + "epoch": 2.7706751897722732, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 6935 + }, + { + "epoch": 2.771074710347583, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 6936 + }, + { + "epoch": 2.771474230922893, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 6937 + }, + { + "epoch": 2.771873751498202, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 6938 + }, + { + "epoch": 2.772273272073512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6939 + }, + { + "epoch": 2.7726727926488213, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 6940 + }, + { + "epoch": 2.773072313224131, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 6941 + }, + { + "epoch": 2.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 6942 + }, + { + "epoch": 2.77387135437475, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6943 + }, + { + "epoch": 2.77427087495006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 6944 + }, + { + "epoch": 2.7746703955253693, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 6945 + }, + { + "epoch": 2.775069916100679, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6946 + }, + { + "epoch": 2.775469436675989, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 6947 + }, + { + "epoch": 2.7758689572512987, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 6948 + }, + { + "epoch": 2.776268477826608, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 6949 + }, + { + "epoch": 2.776667998401918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 6950 + }, + { + "epoch": 2.777067518977227, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 6951 + }, + { + "epoch": 2.777467039552537, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 6952 + }, + { + "epoch": 2.7778665601278467, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 6953 + }, + { + "epoch": 2.778266080703156, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 6954 + }, + { + "epoch": 2.778665601278466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 6955 + }, + { + "epoch": 2.7790651218537756, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 6956 + }, + { + "epoch": 2.779464642429085, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 6957 + }, + { + "epoch": 2.7798641630043948, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6958 + }, + { + "epoch": 2.7802636835797045, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 6959 + }, + { + "epoch": 2.780663204155014, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 6960 + }, + { + "epoch": 2.7810627247303237, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6961 + }, + { + "epoch": 2.781462245305633, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 6962 + }, + { + "epoch": 2.781861765880943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 6963 + }, + { + "epoch": 2.7822612864562526, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 6964 + }, + { + "epoch": 2.7826608070315624, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 6965 + }, + { + "epoch": 2.7830603276068717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 6966 + }, + { + "epoch": 2.7834598481821815, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 6967 + }, + { + "epoch": 2.783859368757491, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 6968 + }, + { + "epoch": 2.7842588893328006, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 6969 + }, + { + "epoch": 2.7846584099081104, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 6970 + }, + { + "epoch": 2.7850579304834198, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 6971 + }, + { + "epoch": 2.7854574510587295, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6972 + }, + { + "epoch": 2.785856971634039, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 6973 + }, + { + "epoch": 2.7862564922093487, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 6974 + }, + { + "epoch": 2.7866560127846585, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 6975 + }, + { + "epoch": 2.7870555333599683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 6976 + }, + { + "epoch": 2.7874550539352776, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 6977 + }, + { + "epoch": 2.7878545745105874, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3449, + "step": 6978 + }, + { + "epoch": 2.7882540950858967, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 6979 + }, + { + "epoch": 2.7886536156612065, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 6980 + }, + { + "epoch": 2.7890531362365163, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 6981 + }, + { + "epoch": 2.789452656811826, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 6982 + }, + { + "epoch": 2.7898521773871354, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6983 + }, + { + "epoch": 2.790251697962445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 6984 + }, + { + "epoch": 2.7906512185377546, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 6985 + }, + { + "epoch": 2.7910507391130643, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 6986 + }, + { + "epoch": 2.791450259688374, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 6987 + }, + { + "epoch": 2.7918497802636835, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 6988 + }, + { + "epoch": 2.7922493008389933, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 6989 + }, + { + "epoch": 2.7926488214143026, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 6990 + }, + { + "epoch": 2.7930483419896124, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6991 + }, + { + "epoch": 2.793447862564922, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 6992 + }, + { + "epoch": 2.793847383140232, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 6993 + }, + { + "epoch": 2.7942469037155413, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 6994 + }, + { + "epoch": 2.794646424290851, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 6995 + }, + { + "epoch": 2.7950459448661604, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 6996 + }, + { + "epoch": 2.79544546544147, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 6997 + }, + { + "epoch": 2.79584498601678, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 6998 + }, + { + "epoch": 2.7962445065920893, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 6999 + }, + { + "epoch": 2.796644027167399, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7000 + }, + { + "epoch": 2.797043547742709, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7001 + }, + { + "epoch": 2.7974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7002 + }, + { + "epoch": 2.797842588893328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7003 + }, + { + "epoch": 2.798242109468638, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7004 + }, + { + "epoch": 2.798641630043947, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7005 + }, + { + "epoch": 2.799041150619257, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7006 + }, + { + "epoch": 2.7994406711945663, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7007 + }, + { + "epoch": 2.799840191769876, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7008 + }, + { + "epoch": 2.800239712345186, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 7009 + }, + { + "epoch": 2.8006392329204957, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7010 + }, + { + "epoch": 2.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7011 + }, + { + "epoch": 2.801438274071115, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7012 + }, + { + "epoch": 2.801837794646424, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7013 + }, + { + "epoch": 2.802237315221734, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7014 + }, + { + "epoch": 2.8026368357970437, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7015 + }, + { + "epoch": 2.803036356372353, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7016 + }, + { + "epoch": 2.803435876947663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7017 + }, + { + "epoch": 2.803835397522972, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7018 + }, + { + "epoch": 2.804234918098282, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7019 + }, + { + "epoch": 2.8046344386735917, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7020 + }, + { + "epoch": 2.8050339592489015, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7021 + }, + { + "epoch": 2.805433479824211, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7022 + }, + { + "epoch": 2.8058330003995207, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7023 + }, + { + "epoch": 2.80623252097483, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 7024 + }, + { + "epoch": 2.80663204155014, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7025 + }, + { + "epoch": 2.8070315621254496, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7026 + }, + { + "epoch": 2.8074310827007594, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7027 + }, + { + "epoch": 2.8078306032760687, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7028 + }, + { + "epoch": 2.8082301238513785, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7029 + }, + { + "epoch": 2.808629644426688, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7030 + }, + { + "epoch": 2.8090291650019976, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7031 + }, + { + "epoch": 2.8094286855773074, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7032 + }, + { + "epoch": 2.8098282061526167, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7033 + }, + { + "epoch": 2.8102277267279265, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7034 + }, + { + "epoch": 2.810627247303236, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7035 + }, + { + "epoch": 2.8110267678785457, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7036 + }, + { + "epoch": 2.8114262884538554, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7037 + }, + { + "epoch": 2.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7038 + }, + { + "epoch": 2.8122253296044746, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7039 + }, + { + "epoch": 2.8126248501797844, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7040 + }, + { + "epoch": 2.8130243707550937, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7041 + }, + { + "epoch": 2.8134238913304035, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7042 + }, + { + "epoch": 2.8138234119057133, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7043 + }, + { + "epoch": 2.8142229324810226, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7044 + }, + { + "epoch": 2.8146224530563324, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7045 + }, + { + "epoch": 2.815021973631642, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 7046 + }, + { + "epoch": 2.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7047 + }, + { + "epoch": 2.8158210147822613, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 7048 + }, + { + "epoch": 2.816220535357571, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7049 + }, + { + "epoch": 2.8166200559328805, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7050 + }, + { + "epoch": 2.8170195765081902, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7051 + }, + { + "epoch": 2.8174190970834996, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7052 + }, + { + "epoch": 2.8178186176588094, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 7053 + }, + { + "epoch": 2.818218138234119, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7054 + }, + { + "epoch": 2.818617658809429, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7055 + }, + { + "epoch": 2.8190171793847383, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7056 + }, + { + "epoch": 2.819416699960048, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7057 + }, + { + "epoch": 2.8198162205353574, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7058 + }, + { + "epoch": 2.820215741110667, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7059 + }, + { + "epoch": 2.820615261685977, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 7060 + }, + { + "epoch": 2.8210147822612863, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7061 + }, + { + "epoch": 2.821414302836596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7062 + }, + { + "epoch": 2.8218138234119055, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7063 + }, + { + "epoch": 2.8222133439872152, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7064 + }, + { + "epoch": 2.822612864562525, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7065 + }, + { + "epoch": 2.823012385137835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7066 + }, + { + "epoch": 2.823411905713144, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7067 + }, + { + "epoch": 2.823811426288454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7068 + }, + { + "epoch": 2.8242109468637633, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7069 + }, + { + "epoch": 2.824610467439073, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7070 + }, + { + "epoch": 2.825009988014383, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7071 + }, + { + "epoch": 2.8254095085896926, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7072 + }, + { + "epoch": 2.825809029165002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7073 + }, + { + "epoch": 2.8262085497403118, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 7074 + }, + { + "epoch": 2.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7075 + }, + { + "epoch": 2.827007590890931, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3424, + "step": 7076 + }, + { + "epoch": 2.8274071114662407, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7077 + }, + { + "epoch": 2.82780663204155, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 7078 + }, + { + "epoch": 2.82820615261686, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7079 + }, + { + "epoch": 2.828605673192169, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7080 + }, + { + "epoch": 2.829005193767479, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7081 + }, + { + "epoch": 2.8294047143427887, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7082 + }, + { + "epoch": 2.8298042349180985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7083 + }, + { + "epoch": 2.830203755493408, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7084 + }, + { + "epoch": 2.8306032760687176, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 7085 + }, + { + "epoch": 2.831002796644027, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7086 + }, + { + "epoch": 2.8314023172193368, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7087 + }, + { + "epoch": 2.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7088 + }, + { + "epoch": 2.832201358369956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7089 + }, + { + "epoch": 2.8326008789452657, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7090 + }, + { + "epoch": 2.8330003995205755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 7091 + }, + { + "epoch": 2.833399920095885, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7092 + }, + { + "epoch": 2.8337994406711946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7093 + }, + { + "epoch": 2.8341989612465044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7094 + }, + { + "epoch": 2.8345984818218137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 7095 + }, + { + "epoch": 2.8349980023971235, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7096 + }, + { + "epoch": 2.835397522972433, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7097 + }, + { + "epoch": 2.8357970435477426, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7098 + }, + { + "epoch": 2.8361965641230524, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7099 + }, + { + "epoch": 2.836596084698362, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7100 + }, + { + "epoch": 2.8369956052736716, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7101 + }, + { + "epoch": 2.8373951258489813, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7102 + }, + { + "epoch": 2.8377946464242907, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7103 + }, + { + "epoch": 2.8381941669996005, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7104 + }, + { + "epoch": 2.8385936875749103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7105 + }, + { + "epoch": 2.8389932081502196, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7106 + }, + { + "epoch": 2.8393927287255294, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7107 + }, + { + "epoch": 2.8397922493008387, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7108 + }, + { + "epoch": 2.8401917698761485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7109 + }, + { + "epoch": 2.8405912904514583, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7110 + }, + { + "epoch": 2.840990811026768, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7111 + }, + { + "epoch": 2.8413903316020774, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7112 + }, + { + "epoch": 2.841789852177387, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7113 + }, + { + "epoch": 2.8421893727526966, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7114 + }, + { + "epoch": 2.8425888933280063, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7115 + }, + { + "epoch": 2.842988413903316, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 7116 + }, + { + "epoch": 2.843387934478626, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7117 + }, + { + "epoch": 2.8437874550539353, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7118 + }, + { + "epoch": 2.844186975629245, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7119 + }, + { + "epoch": 2.8445864962045544, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7120 + }, + { + "epoch": 2.844986016779864, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7121 + }, + { + "epoch": 2.845385537355174, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7122 + }, + { + "epoch": 2.8457850579304833, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7123 + }, + { + "epoch": 2.846184578505793, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3412, + "step": 7124 + }, + { + "epoch": 2.8465840990811024, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7125 + }, + { + "epoch": 2.8469836196564122, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7126 + }, + { + "epoch": 2.847383140231722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7127 + }, + { + "epoch": 2.847782660807032, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7128 + }, + { + "epoch": 2.848182181382341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 7129 + }, + { + "epoch": 2.848581701957651, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7130 + }, + { + "epoch": 2.8489812225329603, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7131 + }, + { + "epoch": 2.84938074310827, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7132 + }, + { + "epoch": 2.84978026368358, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7133 + }, + { + "epoch": 2.850179784258889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7134 + }, + { + "epoch": 2.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7135 + }, + { + "epoch": 2.8509788254095088, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7136 + }, + { + "epoch": 2.851378345984818, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 7137 + }, + { + "epoch": 2.851777866560128, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7138 + }, + { + "epoch": 2.8521773871354377, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7139 + }, + { + "epoch": 2.852576907710747, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7140 + }, + { + "epoch": 2.852976428286057, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7141 + }, + { + "epoch": 2.853375948861366, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7142 + }, + { + "epoch": 2.853775469436676, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 7143 + }, + { + "epoch": 2.8541749900119857, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7144 + }, + { + "epoch": 2.8545745105872955, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7145 + }, + { + "epoch": 2.854974031162605, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7146 + }, + { + "epoch": 2.8553735517379146, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7147 + }, + { + "epoch": 2.855773072313224, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7148 + }, + { + "epoch": 2.8561725928885338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7149 + }, + { + "epoch": 2.8565721134638435, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7150 + }, + { + "epoch": 2.856971634039153, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7151 + }, + { + "epoch": 2.8573711546144627, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7152 + }, + { + "epoch": 2.857770675189772, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7153 + }, + { + "epoch": 2.858170195765082, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7154 + }, + { + "epoch": 2.8585697163403916, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7155 + }, + { + "epoch": 2.8589692369157014, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7156 + }, + { + "epoch": 2.8593687574910107, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7157 + }, + { + "epoch": 2.8597682780663205, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7158 + }, + { + "epoch": 2.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7159 + }, + { + "epoch": 2.8605673192169396, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7160 + }, + { + "epoch": 2.8609668397922494, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7161 + }, + { + "epoch": 2.861366360367559, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7162 + }, + { + "epoch": 2.8617658809428685, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7163 + }, + { + "epoch": 2.8621654015181783, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7164 + }, + { + "epoch": 2.8625649220934877, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 7165 + }, + { + "epoch": 2.8629644426687975, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7166 + }, + { + "epoch": 2.8633639632441072, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7167 + }, + { + "epoch": 2.8637634838194166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7168 + }, + { + "epoch": 2.8641630043947264, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7169 + }, + { + "epoch": 2.8645625249700357, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7170 + }, + { + "epoch": 2.8649620455453455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7171 + }, + { + "epoch": 2.8653615661206553, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 7172 + }, + { + "epoch": 2.865761086695965, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7173 + }, + { + "epoch": 2.8661606072712744, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7174 + }, + { + "epoch": 2.866560127846584, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7175 + }, + { + "epoch": 2.8669596484218935, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7176 + }, + { + "epoch": 2.8673591689972033, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7177 + }, + { + "epoch": 2.867758689572513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7178 + }, + { + "epoch": 2.8681582101478225, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7179 + }, + { + "epoch": 2.8685577307231322, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7180 + }, + { + "epoch": 2.868957251298442, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7181 + }, + { + "epoch": 2.8693567718737514, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7182 + }, + { + "epoch": 2.869756292449061, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7183 + }, + { + "epoch": 2.870155813024371, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7184 + }, + { + "epoch": 2.8705553335996803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7185 + }, + { + "epoch": 2.87095485417499, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7186 + }, + { + "epoch": 2.8713543747502994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7187 + }, + { + "epoch": 2.871753895325609, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7188 + }, + { + "epoch": 2.872153415900919, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7189 + }, + { + "epoch": 2.872552936476229, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7190 + }, + { + "epoch": 2.872952457051538, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7191 + }, + { + "epoch": 2.873351977626848, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7192 + }, + { + "epoch": 2.8737514982021573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7193 + }, + { + "epoch": 2.874151018777467, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7194 + }, + { + "epoch": 2.874550539352777, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7195 + }, + { + "epoch": 2.874950059928086, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 7196 + }, + { + "epoch": 2.875349580503396, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7197 + }, + { + "epoch": 2.8757491010787053, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7198 + }, + { + "epoch": 2.876148621654015, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7199 + }, + { + "epoch": 2.876548142229325, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 7200 + }, + { + "epoch": 2.8769476628046347, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7201 + }, + { + "epoch": 2.877347183379944, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7202 + }, + { + "epoch": 2.877746703955254, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7203 + }, + { + "epoch": 2.878146224530563, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7204 + }, + { + "epoch": 2.878545745105873, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7205 + }, + { + "epoch": 2.8789452656811827, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 7206 + }, + { + "epoch": 2.8793447862564925, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7207 + }, + { + "epoch": 2.879744306831802, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7208 + }, + { + "epoch": 2.8801438274071116, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7209 + }, + { + "epoch": 2.880543347982421, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 7210 + }, + { + "epoch": 2.8809428685577307, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7211 + }, + { + "epoch": 2.8813423891330405, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7212 + }, + { + "epoch": 2.88174190970835, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 7213 + }, + { + "epoch": 2.8821414302836597, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7214 + }, + { + "epoch": 2.882540950858969, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7215 + }, + { + "epoch": 2.882940471434279, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7216 + }, + { + "epoch": 2.8833399920095886, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7217 + }, + { + "epoch": 2.8837395125848984, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 7218 + }, + { + "epoch": 2.8841390331602077, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7219 + }, + { + "epoch": 2.8845385537355175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7220 + }, + { + "epoch": 2.884938074310827, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7221 + }, + { + "epoch": 2.8853375948861366, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7222 + }, + { + "epoch": 2.8857371154614464, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7223 + }, + { + "epoch": 2.8861366360367557, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7224 + }, + { + "epoch": 2.8865361566120655, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7225 + }, + { + "epoch": 2.8869356771873753, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7226 + }, + { + "epoch": 2.8873351977626847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7227 + }, + { + "epoch": 2.8877347183379944, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7228 + }, + { + "epoch": 2.8881342389133042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7229 + }, + { + "epoch": 2.8885337594886136, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7230 + }, + { + "epoch": 2.8889332800639234, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7231 + }, + { + "epoch": 2.8893328006392327, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7232 + }, + { + "epoch": 2.8897323212145425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 7233 + }, + { + "epoch": 2.8901318417898523, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7234 + }, + { + "epoch": 2.890531362365162, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7235 + }, + { + "epoch": 2.8909308829404714, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7236 + }, + { + "epoch": 2.891330403515781, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7237 + }, + { + "epoch": 2.8917299240910905, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7238 + }, + { + "epoch": 2.8921294446664003, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7239 + }, + { + "epoch": 2.89252896524171, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7240 + }, + { + "epoch": 2.8929284858170194, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7241 + }, + { + "epoch": 2.8933280063923292, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7242 + }, + { + "epoch": 2.8937275269676386, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7243 + }, + { + "epoch": 2.8941270475429484, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7244 + }, + { + "epoch": 2.894526568118258, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3404, + "step": 7245 + }, + { + "epoch": 2.894926088693568, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7246 + }, + { + "epoch": 2.8953256092688773, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7247 + }, + { + "epoch": 2.895725129844187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7248 + }, + { + "epoch": 2.8961246504194964, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7249 + }, + { + "epoch": 2.896524170994806, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 7250 + }, + { + "epoch": 2.896923691570116, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 7251 + }, + { + "epoch": 2.8973232121454258, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7252 + }, + { + "epoch": 2.897722732720735, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7253 + }, + { + "epoch": 2.898122253296045, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7254 + }, + { + "epoch": 2.8985217738713542, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7255 + }, + { + "epoch": 2.898921294446664, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 7256 + }, + { + "epoch": 2.899320815021974, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7257 + }, + { + "epoch": 2.899720335597283, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7258 + }, + { + "epoch": 2.900119856172593, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7259 + }, + { + "epoch": 2.9005193767479023, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7260 + }, + { + "epoch": 2.900918897323212, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7261 + }, + { + "epoch": 2.901318417898522, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7262 + }, + { + "epoch": 2.9017179384738316, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7263 + }, + { + "epoch": 2.902117459049141, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7264 + }, + { + "epoch": 2.9025169796244508, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7265 + }, + { + "epoch": 2.90291650019976, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 7266 + }, + { + "epoch": 2.90331602077507, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7267 + }, + { + "epoch": 2.9037155413503797, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 7268 + }, + { + "epoch": 2.904115061925689, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7269 + }, + { + "epoch": 2.904514582500999, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7270 + }, + { + "epoch": 2.9049141030763086, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7271 + }, + { + "epoch": 2.905313623651618, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7272 + }, + { + "epoch": 2.9057131442269277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7273 + }, + { + "epoch": 2.9061126648022375, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7274 + }, + { + "epoch": 2.906512185377547, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7275 + }, + { + "epoch": 2.9069117059528566, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7276 + }, + { + "epoch": 2.907311226528166, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7277 + }, + { + "epoch": 2.9077107471034758, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7278 + }, + { + "epoch": 2.9081102676787856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7279 + }, + { + "epoch": 2.9085097882540953, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7280 + }, + { + "epoch": 2.9089093088294047, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 7281 + }, + { + "epoch": 2.9093088294047145, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7282 + }, + { + "epoch": 2.909708349980024, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7283 + }, + { + "epoch": 2.9101078705553336, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7284 + }, + { + "epoch": 2.9105073911306434, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7285 + }, + { + "epoch": 2.9109069117059527, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7286 + }, + { + "epoch": 2.9113064322812625, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7287 + }, + { + "epoch": 2.911705952856572, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3469, + "step": 7288 + }, + { + "epoch": 2.9121054734318816, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7289 + }, + { + "epoch": 2.9125049940071914, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7290 + }, + { + "epoch": 2.912904514582501, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7291 + }, + { + "epoch": 2.9133040351578106, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 7292 + }, + { + "epoch": 2.9137035557331203, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7293 + }, + { + "epoch": 2.9141030763084297, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7294 + }, + { + "epoch": 2.9145025968837395, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7295 + }, + { + "epoch": 2.9149021174590493, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7296 + }, + { + "epoch": 2.915301638034359, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7297 + }, + { + "epoch": 2.9157011586096684, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7298 + }, + { + "epoch": 2.916100679184978, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7299 + }, + { + "epoch": 2.9165001997602875, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7300 + }, + { + "epoch": 2.9168997203355973, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7301 + }, + { + "epoch": 2.917299240910907, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7302 + }, + { + "epoch": 2.9176987614862164, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7303 + }, + { + "epoch": 2.918098282061526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7304 + }, + { + "epoch": 2.9184978026368356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7305 + }, + { + "epoch": 2.9188973232121453, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7306 + }, + { + "epoch": 2.919296843787455, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7307 + }, + { + "epoch": 2.919696364362765, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 7308 + }, + { + "epoch": 2.9200958849380743, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7309 + }, + { + "epoch": 2.920495405513384, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7310 + }, + { + "epoch": 2.9208949260886934, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 7311 + }, + { + "epoch": 2.921294446664003, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7312 + }, + { + "epoch": 2.921693967239313, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7313 + }, + { + "epoch": 2.9220934878146223, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 7314 + }, + { + "epoch": 2.922493008389932, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7315 + }, + { + "epoch": 2.922892528965242, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7316 + }, + { + "epoch": 2.923292049540551, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7317 + }, + { + "epoch": 2.923691570115861, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7318 + }, + { + "epoch": 2.924091090691171, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7319 + }, + { + "epoch": 2.92449061126648, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7320 + }, + { + "epoch": 2.92489013184179, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 7321 + }, + { + "epoch": 2.9252896524170993, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7322 + }, + { + "epoch": 2.925689172992409, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7323 + }, + { + "epoch": 2.926088693567719, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7324 + }, + { + "epoch": 2.9264882141430286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7325 + }, + { + "epoch": 2.926887734718338, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7326 + }, + { + "epoch": 2.9272872552936477, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7327 + }, + { + "epoch": 2.927686775868957, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7328 + }, + { + "epoch": 2.928086296444267, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7329 + }, + { + "epoch": 2.9284858170195767, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7330 + }, + { + "epoch": 2.928885337594886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 7331 + }, + { + "epoch": 2.929284858170196, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7332 + }, + { + "epoch": 2.929684378745505, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7333 + }, + { + "epoch": 2.930083899320815, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7334 + }, + { + "epoch": 2.9304834198961247, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7335 + }, + { + "epoch": 2.9308829404714345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7336 + }, + { + "epoch": 2.931282461046744, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 7337 + }, + { + "epoch": 2.9316819816220536, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7338 + }, + { + "epoch": 2.932081502197363, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7339 + }, + { + "epoch": 2.9324810227726728, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7340 + }, + { + "epoch": 2.9328805433479825, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7341 + }, + { + "epoch": 2.9332800639232923, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7342 + }, + { + "epoch": 2.9336795844986017, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7343 + }, + { + "epoch": 2.9340791050739115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7344 + }, + { + "epoch": 2.934478625649221, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7345 + }, + { + "epoch": 2.9348781462245306, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7346 + }, + { + "epoch": 2.9352776667998404, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7347 + }, + { + "epoch": 2.9356771873751497, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7348 + }, + { + "epoch": 2.9360767079504595, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7349 + }, + { + "epoch": 2.936476228525769, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 7350 + }, + { + "epoch": 2.9368757491010786, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7351 + }, + { + "epoch": 2.9372752696763884, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7352 + }, + { + "epoch": 2.937674790251698, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7353 + }, + { + "epoch": 2.9380743108270075, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7354 + }, + { + "epoch": 2.9384738314023173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7355 + }, + { + "epoch": 2.9388733519776267, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7356 + }, + { + "epoch": 2.9392728725529365, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7357 + }, + { + "epoch": 2.9396723931282462, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 7358 + }, + { + "epoch": 2.9400719137035556, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7359 + }, + { + "epoch": 2.9404714342788654, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7360 + }, + { + "epoch": 2.9408709548541747, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7361 + }, + { + "epoch": 2.9412704754294845, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7362 + }, + { + "epoch": 2.9416699960047943, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7363 + }, + { + "epoch": 2.942069516580104, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7364 + }, + { + "epoch": 2.9424690371554134, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7365 + }, + { + "epoch": 2.942868557730723, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7366 + }, + { + "epoch": 2.9432680783060325, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7367 + }, + { + "epoch": 2.9436675988813423, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7368 + }, + { + "epoch": 2.944067119456652, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7369 + }, + { + "epoch": 2.944466640031962, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7370 + }, + { + "epoch": 2.9448661606072712, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7371 + }, + { + "epoch": 2.945265681182581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7372 + }, + { + "epoch": 2.9456652017578904, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7373 + }, + { + "epoch": 2.9460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7374 + }, + { + "epoch": 2.94646424290851, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7375 + }, + { + "epoch": 2.9468637634838193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7376 + }, + { + "epoch": 2.947263284059129, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7377 + }, + { + "epoch": 2.9476628046344384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7378 + }, + { + "epoch": 2.948062325209748, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7379 + }, + { + "epoch": 2.948461845785058, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7380 + }, + { + "epoch": 2.9488613663603678, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7381 + }, + { + "epoch": 2.949260886935677, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7382 + }, + { + "epoch": 2.949660407510987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 7383 + }, + { + "epoch": 2.9500599280862962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7384 + }, + { + "epoch": 2.950459448661606, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7385 + }, + { + "epoch": 2.950858969236916, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7386 + }, + { + "epoch": 2.9512584898122256, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7387 + }, + { + "epoch": 2.951658010387535, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7388 + }, + { + "epoch": 2.9520575309628447, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7389 + }, + { + "epoch": 2.952457051538154, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7390 + }, + { + "epoch": 2.952856572113464, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7391 + }, + { + "epoch": 2.9532560926887736, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7392 + }, + { + "epoch": 2.953655613264083, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7393 + }, + { + "epoch": 2.9540551338393928, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7394 + }, + { + "epoch": 2.954454654414702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7395 + }, + { + "epoch": 2.954854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7396 + }, + { + "epoch": 2.9552536955653217, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7397 + }, + { + "epoch": 2.9556532161406315, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7398 + }, + { + "epoch": 2.956052736715941, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7399 + }, + { + "epoch": 2.9564522572912506, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 7400 + }, + { + "epoch": 2.95685177786656, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7401 + }, + { + "epoch": 2.9572512984418697, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7402 + }, + { + "epoch": 2.9576508190171795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7403 + }, + { + "epoch": 2.958050339592489, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7404 + }, + { + "epoch": 2.9584498601677987, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7405 + }, + { + "epoch": 2.958849380743108, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7406 + }, + { + "epoch": 2.959248901318418, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 7407 + }, + { + "epoch": 2.9596484218937276, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7408 + }, + { + "epoch": 2.9600479424690374, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7409 + }, + { + "epoch": 2.9604474630443467, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7410 + }, + { + "epoch": 2.9608469836196565, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3446, + "step": 7411 + }, + { + "epoch": 2.961246504194966, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7412 + }, + { + "epoch": 2.9616460247702756, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7413 + }, + { + "epoch": 2.9620455453455854, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7414 + }, + { + "epoch": 2.962445065920895, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7415 + }, + { + "epoch": 2.9628445864962045, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7416 + }, + { + "epoch": 2.9632441070715143, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7417 + }, + { + "epoch": 2.9636436276468237, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7418 + }, + { + "epoch": 2.9640431482221334, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7419 + }, + { + "epoch": 2.9644426687974432, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7420 + }, + { + "epoch": 2.9648421893727526, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7421 + }, + { + "epoch": 2.9652417099480624, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7422 + }, + { + "epoch": 2.9656412305233717, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7423 + }, + { + "epoch": 2.9660407510986815, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7424 + }, + { + "epoch": 2.9664402716739913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7425 + }, + { + "epoch": 2.966839792249301, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7426 + }, + { + "epoch": 2.9672393128246104, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7427 + }, + { + "epoch": 2.96763883339992, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7428 + }, + { + "epoch": 2.9680383539752295, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7429 + }, + { + "epoch": 2.9684378745505393, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7430 + }, + { + "epoch": 2.968837395125849, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7431 + }, + { + "epoch": 2.969236915701159, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 7432 + }, + { + "epoch": 2.9696364362764682, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7433 + }, + { + "epoch": 2.970035956851778, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7434 + }, + { + "epoch": 2.9704354774270874, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7435 + }, + { + "epoch": 2.970834998002397, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7436 + }, + { + "epoch": 2.971234518577707, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 7437 + }, + { + "epoch": 2.9716340391530163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7438 + }, + { + "epoch": 2.972033559728326, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7439 + }, + { + "epoch": 2.9724330803036354, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7440 + }, + { + "epoch": 2.972832600878945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7441 + }, + { + "epoch": 2.973232121454255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 7442 + }, + { + "epoch": 2.9736316420295648, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7443 + }, + { + "epoch": 2.974031162604874, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7444 + }, + { + "epoch": 2.974430683180184, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7445 + }, + { + "epoch": 2.9748302037554932, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7446 + }, + { + "epoch": 2.975229724330803, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7447 + }, + { + "epoch": 2.975629244906113, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7448 + }, + { + "epoch": 2.976028765481422, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 7449 + }, + { + "epoch": 2.976428286056732, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7450 + }, + { + "epoch": 2.9768278066320413, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7451 + }, + { + "epoch": 2.977227327207351, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7452 + }, + { + "epoch": 2.977626847782661, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7453 + }, + { + "epoch": 2.9780263683579706, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7454 + }, + { + "epoch": 2.97842588893328, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7455 + }, + { + "epoch": 2.9788254095085898, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7456 + }, + { + "epoch": 2.979224930083899, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7457 + }, + { + "epoch": 2.979624450659209, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 7458 + }, + { + "epoch": 2.9800239712345187, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7459 + }, + { + "epoch": 2.9804234918098285, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7460 + }, + { + "epoch": 2.980823012385138, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7461 + }, + { + "epoch": 2.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7462 + }, + { + "epoch": 2.981622053535757, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 7463 + }, + { + "epoch": 2.9820215741110667, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7464 + }, + { + "epoch": 2.9824210946863765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7465 + }, + { + "epoch": 2.982820615261686, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7466 + }, + { + "epoch": 2.9832201358369956, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 7467 + }, + { + "epoch": 2.983619656412305, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7468 + }, + { + "epoch": 2.9840191769876148, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7469 + }, + { + "epoch": 2.9844186975629245, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7470 + }, + { + "epoch": 2.9848182181382343, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7471 + }, + { + "epoch": 2.9852177387135437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7472 + }, + { + "epoch": 2.9856172592888535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7473 + }, + { + "epoch": 2.986016779864163, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7474 + }, + { + "epoch": 2.9864163004394726, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7475 + }, + { + "epoch": 2.9868158210147824, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7476 + }, + { + "epoch": 2.987215341590092, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7477 + }, + { + "epoch": 2.9876148621654015, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7478 + }, + { + "epoch": 2.9880143827407113, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7479 + }, + { + "epoch": 2.9884139033160206, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 7480 + }, + { + "epoch": 2.9888134238913304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7481 + }, + { + "epoch": 2.98921294446664, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7482 + }, + { + "epoch": 2.9896124650419496, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7483 + }, + { + "epoch": 2.9900119856172593, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7484 + }, + { + "epoch": 2.9904115061925687, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7485 + }, + { + "epoch": 2.9908110267678785, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 7486 + }, + { + "epoch": 2.9912105473431883, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7487 + }, + { + "epoch": 2.991610067918498, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7488 + }, + { + "epoch": 2.9920095884938074, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 7489 + }, + { + "epoch": 2.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7490 + }, + { + "epoch": 2.9928086296444265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7491 + }, + { + "epoch": 2.9932081502197363, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7492 + }, + { + "epoch": 2.993607670795046, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 7493 + }, + { + "epoch": 2.9940071913703554, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7494 + }, + { + "epoch": 2.994406711945665, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7495 + }, + { + "epoch": 2.9948062325209746, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7496 + }, + { + "epoch": 2.9952057530962843, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7497 + }, + { + "epoch": 2.995605273671594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3391, + "step": 7498 + }, + { + "epoch": 2.996004794246904, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 7499 + }, + { + "epoch": 2.9964043148222133, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7500 + }, + { + "epoch": 2.996803835397523, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 7501 + }, + { + "epoch": 2.9972033559728324, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7502 + }, + { + "epoch": 2.997602876548142, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7503 + }, + { + "epoch": 2.998002397123452, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3383, + "step": 7504 + }, + { + "epoch": 2.9984019176987617, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 7505 + }, + { + "epoch": 2.998801438274071, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7506 + }, + { + "epoch": 2.999200958849381, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7507 + }, + { + "epoch": 2.99960047942469, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7508 + }, + { + "epoch": 3.0, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7509 + }, + { + "epoch": 3.00039952057531, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7510 + }, + { + "epoch": 3.000799041150619, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7511 + }, + { + "epoch": 3.001198561725929, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7512 + }, + { + "epoch": 3.0015980823012387, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7513 + }, + { + "epoch": 3.001997602876548, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7514 + }, + { + "epoch": 3.002397123451858, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 7515 + }, + { + "epoch": 3.0027966440271676, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7516 + }, + { + "epoch": 3.003196164602477, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7517 + }, + { + "epoch": 3.0035956851777867, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7518 + }, + { + "epoch": 3.003995205753096, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7519 + }, + { + "epoch": 3.004394726328406, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 7520 + }, + { + "epoch": 3.0047942469037157, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7521 + }, + { + "epoch": 3.005193767479025, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7522 + }, + { + "epoch": 3.005593288054335, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 7523 + }, + { + "epoch": 3.0059928086296446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7524 + }, + { + "epoch": 3.006392329204954, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3464, + "step": 7525 + }, + { + "epoch": 3.0067918497802637, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7526 + }, + { + "epoch": 3.0071913703555735, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7527 + }, + { + "epoch": 3.007590890930883, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7528 + }, + { + "epoch": 3.0079904115061926, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7529 + }, + { + "epoch": 3.0083899320815024, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7530 + }, + { + "epoch": 3.0087894526568117, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7531 + }, + { + "epoch": 3.0091889732321215, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7532 + }, + { + "epoch": 3.009588493807431, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7533 + }, + { + "epoch": 3.0099880143827407, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7534 + }, + { + "epoch": 3.0103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7535 + }, + { + "epoch": 3.01078705553336, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7536 + }, + { + "epoch": 3.0111865761086696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7537 + }, + { + "epoch": 3.0115860966839794, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7538 + }, + { + "epoch": 3.0119856172592887, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7539 + }, + { + "epoch": 3.0123851378345985, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7540 + }, + { + "epoch": 3.0127846584099083, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7541 + }, + { + "epoch": 3.0131841789852176, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7542 + }, + { + "epoch": 3.0135836995605274, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7543 + }, + { + "epoch": 3.013983220135837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7544 + }, + { + "epoch": 3.0143827407111465, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7545 + }, + { + "epoch": 3.0147822612864563, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 7546 + }, + { + "epoch": 3.0151817818617657, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 7547 + }, + { + "epoch": 3.0155813024370755, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7548 + }, + { + "epoch": 3.0159808230123852, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7549 + }, + { + "epoch": 3.0163803435876946, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7550 + }, + { + "epoch": 3.0167798641630044, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7551 + }, + { + "epoch": 3.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7552 + }, + { + "epoch": 3.0175789053136235, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7553 + }, + { + "epoch": 3.0179784258889333, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7554 + }, + { + "epoch": 3.018377946464243, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7555 + }, + { + "epoch": 3.0187774670395524, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7556 + }, + { + "epoch": 3.019176987614862, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7557 + }, + { + "epoch": 3.019576508190172, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7558 + }, + { + "epoch": 3.0199760287654813, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7559 + }, + { + "epoch": 3.020375549340791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 7560 + }, + { + "epoch": 3.020775069916101, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7561 + }, + { + "epoch": 3.0211745904914102, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7562 + }, + { + "epoch": 3.02157411106672, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 7563 + }, + { + "epoch": 3.0219736316420294, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7564 + }, + { + "epoch": 3.022373152217339, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7565 + }, + { + "epoch": 3.022772672792649, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 7566 + }, + { + "epoch": 3.0231721933679583, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 7567 + }, + { + "epoch": 3.023571713943268, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7568 + }, + { + "epoch": 3.023971234518578, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7569 + }, + { + "epoch": 3.024370755093887, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7570 + }, + { + "epoch": 3.024770275669197, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 7571 + }, + { + "epoch": 3.0251697962445068, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7572 + }, + { + "epoch": 3.025569316819816, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7573 + }, + { + "epoch": 3.025968837395126, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 7574 + }, + { + "epoch": 3.0263683579704357, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7575 + }, + { + "epoch": 3.026767878545745, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7576 + }, + { + "epoch": 3.027167399121055, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7577 + }, + { + "epoch": 3.027566919696364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7578 + }, + { + "epoch": 3.027966440271674, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7579 + }, + { + "epoch": 3.0283659608469837, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7580 + }, + { + "epoch": 3.028765481422293, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7581 + }, + { + "epoch": 3.029165001997603, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7582 + }, + { + "epoch": 3.0295645225729126, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3462, + "step": 7583 + }, + { + "epoch": 3.029964043148222, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7584 + }, + { + "epoch": 3.0303635637235318, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7585 + }, + { + "epoch": 3.0307630842988416, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 7586 + }, + { + "epoch": 3.031162604874151, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7587 + }, + { + "epoch": 3.0315621254494607, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7588 + }, + { + "epoch": 3.0319616460247705, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7589 + }, + { + "epoch": 3.03236116660008, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7590 + }, + { + "epoch": 3.0327606871753896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 7591 + }, + { + "epoch": 3.033160207750699, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 7592 + }, + { + "epoch": 3.0335597283260087, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7593 + }, + { + "epoch": 3.0339592489013185, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7594 + }, + { + "epoch": 3.034358769476628, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 7595 + }, + { + "epoch": 3.0347582900519376, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7596 + }, + { + "epoch": 3.0351578106272474, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7597 + }, + { + "epoch": 3.0355573312025568, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 7598 + }, + { + "epoch": 3.0359568517778666, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 7599 + }, + { + "epoch": 3.0363563723531763, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 7600 + }, + { + "epoch": 3.0367558929284857, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 7601 + }, + { + "epoch": 3.0371554135037955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7602 + }, + { + "epoch": 3.0375549340791053, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7603 + }, + { + "epoch": 3.0379544546544146, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7604 + }, + { + "epoch": 3.0383539752297244, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7605 + }, + { + "epoch": 3.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 7606 + }, + { + "epoch": 3.0391530163803435, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7607 + }, + { + "epoch": 3.0395525369556533, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7608 + }, + { + "epoch": 3.0399520575309626, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7609 + }, + { + "epoch": 3.0403515781062724, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7610 + }, + { + "epoch": 3.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7611 + }, + { + "epoch": 3.0411506192568916, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 7612 + }, + { + "epoch": 3.0415501398322013, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7613 + }, + { + "epoch": 3.041949660407511, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7614 + }, + { + "epoch": 3.0423491809828205, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7615 + }, + { + "epoch": 3.0427487015581303, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7616 + }, + { + "epoch": 3.04314822213344, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7617 + }, + { + "epoch": 3.0435477427087494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7618 + }, + { + "epoch": 3.043947263284059, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7619 + }, + { + "epoch": 3.044346783859369, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7620 + }, + { + "epoch": 3.0447463044346783, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7621 + }, + { + "epoch": 3.045145825009988, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 7622 + }, + { + "epoch": 3.0455453455852974, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7623 + }, + { + "epoch": 3.0459448661606072, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 7624 + }, + { + "epoch": 3.046344386735917, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3427, + "step": 7625 + }, + { + "epoch": 3.0467439073112264, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7626 + }, + { + "epoch": 3.047143427886536, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 7627 + }, + { + "epoch": 3.047542948461846, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7628 + }, + { + "epoch": 3.0479424690371553, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7629 + }, + { + "epoch": 3.048341989612465, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 7630 + }, + { + "epoch": 3.048741510187775, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7631 + }, + { + "epoch": 3.049141030763084, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7632 + }, + { + "epoch": 3.049540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7633 + }, + { + "epoch": 3.0499400719137038, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 7634 + }, + { + "epoch": 3.050339592489013, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7635 + }, + { + "epoch": 3.050739113064323, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7636 + }, + { + "epoch": 3.0511386336396322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7637 + }, + { + "epoch": 3.051538154214942, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7638 + }, + { + "epoch": 3.051937674790252, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7639 + }, + { + "epoch": 3.052337195365561, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7640 + }, + { + "epoch": 3.052736715940871, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7641 + }, + { + "epoch": 3.0531362365161807, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7642 + }, + { + "epoch": 3.05353575709149, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7643 + }, + { + "epoch": 3.0539352776668, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7644 + }, + { + "epoch": 3.0543347982421096, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7645 + }, + { + "epoch": 3.054734318817419, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7646 + }, + { + "epoch": 3.0551338393927288, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7647 + }, + { + "epoch": 3.0555333599680385, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7648 + }, + { + "epoch": 3.055932880543348, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7649 + }, + { + "epoch": 3.0563324011186577, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3419, + "step": 7650 + }, + { + "epoch": 3.0567319216939675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7651 + }, + { + "epoch": 3.057131442269277, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 7652 + }, + { + "epoch": 3.0575309628445866, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7653 + }, + { + "epoch": 3.057930483419896, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 7654 + }, + { + "epoch": 3.0583300039952057, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7655 + }, + { + "epoch": 3.0587295245705155, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7656 + }, + { + "epoch": 3.059129045145825, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7657 + }, + { + "epoch": 3.0595285657211346, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7658 + }, + { + "epoch": 3.0599280862964444, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7659 + }, + { + "epoch": 3.0603276068717538, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7660 + }, + { + "epoch": 3.0607271274470635, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 7661 + }, + { + "epoch": 3.0611266480223733, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7662 + }, + { + "epoch": 3.0615261685976827, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7663 + }, + { + "epoch": 3.0619256891729925, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 7664 + }, + { + "epoch": 3.0623252097483022, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7665 + }, + { + "epoch": 3.0627247303236116, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7666 + }, + { + "epoch": 3.0631242508989214, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7667 + }, + { + "epoch": 3.0635237714742307, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7668 + }, + { + "epoch": 3.0639232920495405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7669 + }, + { + "epoch": 3.0643228126248503, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 7670 + }, + { + "epoch": 3.0647223332001596, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 7671 + }, + { + "epoch": 3.0651218537754694, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 7672 + }, + { + "epoch": 3.065521374350779, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 7673 + }, + { + "epoch": 3.0659208949260885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7674 + }, + { + "epoch": 3.0663204155013983, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7675 + }, + { + "epoch": 3.066719936076708, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 7676 + }, + { + "epoch": 3.0671194566520175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 7677 + }, + { + "epoch": 3.0675189772273272, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7678 + }, + { + "epoch": 3.067918497802637, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7679 + }, + { + "epoch": 3.0683180183779464, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7680 + }, + { + "epoch": 3.068717538953256, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7681 + }, + { + "epoch": 3.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 7682 + }, + { + "epoch": 3.0695165801038753, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 7683 + }, + { + "epoch": 3.069916100679185, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 7684 + }, + { + "epoch": 3.0703156212544944, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 7685 + }, + { + "epoch": 3.070715141829804, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7686 + }, + { + "epoch": 3.071114662405114, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7687 + }, + { + "epoch": 3.0715141829804233, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7688 + }, + { + "epoch": 3.071913703555733, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 7689 + }, + { + "epoch": 3.072313224131043, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 7690 + }, + { + "epoch": 3.0727127447063523, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7691 + }, + { + "epoch": 3.073112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7692 + }, + { + "epoch": 3.073511785856972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7693 + }, + { + "epoch": 3.073911306432281, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7694 + }, + { + "epoch": 3.074310827007591, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 7695 + }, + { + "epoch": 3.0747103475829007, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7696 + }, + { + "epoch": 3.07510986815821, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7697 + }, + { + "epoch": 3.07550938873352, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 7698 + }, + { + "epoch": 3.075908909308829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7699 + }, + { + "epoch": 3.076308429884139, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7700 + }, + { + "epoch": 3.076707950459449, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7701 + }, + { + "epoch": 3.077107471034758, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7702 + }, + { + "epoch": 3.077506991610068, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 7703 + }, + { + "epoch": 3.0779065121853777, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 7704 + }, + { + "epoch": 3.078306032760687, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7705 + }, + { + "epoch": 3.078705553335997, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7706 + }, + { + "epoch": 3.0791050739113066, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 7707 + }, + { + "epoch": 3.079504594486616, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7708 + }, + { + "epoch": 3.0799041150619257, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7709 + }, + { + "epoch": 3.0803036356372355, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7710 + }, + { + "epoch": 3.080703156212545, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 7711 + }, + { + "epoch": 3.0811026767878547, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7712 + }, + { + "epoch": 3.081502197363164, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 7713 + }, + { + "epoch": 3.081901717938474, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7714 + }, + { + "epoch": 3.0823012385137836, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 7715 + }, + { + "epoch": 3.082700759089093, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7716 + }, + { + "epoch": 3.0831002796644027, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7717 + }, + { + "epoch": 3.0834998002397125, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 7718 + }, + { + "epoch": 3.083899320815022, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7719 + }, + { + "epoch": 3.0842988413903316, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 7720 + }, + { + "epoch": 3.0846983619656414, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7721 + }, + { + "epoch": 3.0850978825409507, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 7722 + }, + { + "epoch": 3.0854974031162605, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7723 + }, + { + "epoch": 3.0858969236915703, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 7724 + }, + { + "epoch": 3.0862964442668797, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7725 + }, + { + "epoch": 3.0866959648421894, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7726 + }, + { + "epoch": 3.087095485417499, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 7727 + }, + { + "epoch": 3.0874950059928086, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7728 + }, + { + "epoch": 3.0878945265681184, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 7729 + }, + { + "epoch": 3.0882940471434277, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7730 + }, + { + "epoch": 3.0886935677187375, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7731 + }, + { + "epoch": 3.0890930882940473, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 7732 + }, + { + "epoch": 3.0894926088693566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7733 + }, + { + "epoch": 3.0898921294446664, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7734 + }, + { + "epoch": 3.090291650019976, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 7735 + }, + { + "epoch": 3.0906911705952855, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7736 + }, + { + "epoch": 3.0910906911705953, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 7737 + }, + { + "epoch": 3.091490211745905, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 7738 + }, + { + "epoch": 3.0918897323212144, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 7739 + }, + { + "epoch": 3.0922892528965242, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7740 + }, + { + "epoch": 3.092688773471834, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7741 + }, + { + "epoch": 3.0930882940471434, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7742 + }, + { + "epoch": 3.093487814622453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7743 + }, + { + "epoch": 3.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 7744 + }, + { + "epoch": 3.0942868557730723, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7745 + }, + { + "epoch": 3.094686376348382, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 7746 + }, + { + "epoch": 3.0950858969236914, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 7747 + }, + { + "epoch": 3.095485417499001, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 7748 + }, + { + "epoch": 3.095884938074311, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7749 + }, + { + "epoch": 3.0962844586496203, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7750 + }, + { + "epoch": 3.09668397922493, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 7751 + }, + { + "epoch": 3.09708349980024, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7752 + }, + { + "epoch": 3.0974830203755492, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7753 + }, + { + "epoch": 3.097882540950859, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 7754 + }, + { + "epoch": 3.098282061526169, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 7755 + }, + { + "epoch": 3.098681582101478, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7756 + }, + { + "epoch": 3.099081102676788, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7757 + }, + { + "epoch": 3.0994806232520973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7758 + }, + { + "epoch": 3.099880143827407, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7759 + }, + { + "epoch": 3.100279664402717, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7760 + }, + { + "epoch": 3.100679184978026, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7761 + }, + { + "epoch": 3.101078705553336, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 7762 + }, + { + "epoch": 3.1014782261286458, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 7763 + }, + { + "epoch": 3.101877746703955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7764 + }, + { + "epoch": 3.102277267279265, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7765 + }, + { + "epoch": 3.1026767878545747, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 7766 + }, + { + "epoch": 3.103076308429884, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7767 + }, + { + "epoch": 3.103475829005194, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7768 + }, + { + "epoch": 3.1038753495805036, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7769 + }, + { + "epoch": 3.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 7770 + }, + { + "epoch": 3.1046743907311227, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7771 + }, + { + "epoch": 3.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 7772 + }, + { + "epoch": 3.105473431881742, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 7773 + }, + { + "epoch": 3.1058729524570516, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7774 + }, + { + "epoch": 3.106272473032361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7775 + }, + { + "epoch": 3.1066719936076708, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 7776 + }, + { + "epoch": 3.1070715141829806, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 7777 + }, + { + "epoch": 3.10747103475829, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7778 + }, + { + "epoch": 3.1078705553335997, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 7779 + }, + { + "epoch": 3.1082700759089095, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7780 + }, + { + "epoch": 3.108669596484219, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7781 + }, + { + "epoch": 3.1090691170595286, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 7782 + }, + { + "epoch": 3.1094686376348384, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 7783 + }, + { + "epoch": 3.1098681582101477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7784 + }, + { + "epoch": 3.1102676787854575, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 7785 + }, + { + "epoch": 3.1106671993607673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 7786 + }, + { + "epoch": 3.1110667199360766, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 7787 + }, + { + "epoch": 3.1114662405113864, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 7788 + }, + { + "epoch": 3.1118657610866958, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7789 + }, + { + "epoch": 3.1122652816620056, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7790 + }, + { + "epoch": 3.1126648022373153, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7791 + }, + { + "epoch": 3.1130643228126247, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7792 + }, + { + "epoch": 3.1134638433879345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 7793 + }, + { + "epoch": 3.1138633639632443, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7794 + }, + { + "epoch": 3.1142628845385536, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 7795 + }, + { + "epoch": 3.1146624051138634, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 7796 + }, + { + "epoch": 3.115061925689173, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7797 + }, + { + "epoch": 3.1154614462644825, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7798 + }, + { + "epoch": 3.1158609668397923, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 7799 + }, + { + "epoch": 3.116260487415102, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7800 + }, + { + "epoch": 3.1166600079904114, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7801 + }, + { + "epoch": 3.117059528565721, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 7802 + }, + { + "epoch": 3.1174590491410306, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7803 + }, + { + "epoch": 3.1178585697163403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7804 + }, + { + "epoch": 3.11825809029165, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7805 + }, + { + "epoch": 3.1186576108669595, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 7806 + }, + { + "epoch": 3.1190571314422693, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 7807 + }, + { + "epoch": 3.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7808 + }, + { + "epoch": 3.1198561725928884, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 7809 + }, + { + "epoch": 3.120255693168198, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7810 + }, + { + "epoch": 3.120655213743508, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 7811 + }, + { + "epoch": 3.1210547343188173, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7812 + }, + { + "epoch": 3.121454254894127, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7813 + }, + { + "epoch": 3.121853775469437, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7814 + }, + { + "epoch": 3.122253296044746, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7815 + }, + { + "epoch": 3.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7816 + }, + { + "epoch": 3.1230523371953653, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7817 + }, + { + "epoch": 3.123451857770675, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7818 + }, + { + "epoch": 3.123851378345985, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 7819 + }, + { + "epoch": 3.1242508989212943, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 7820 + }, + { + "epoch": 3.124650419496604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7821 + }, + { + "epoch": 3.125049940071914, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 7822 + }, + { + "epoch": 3.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7823 + }, + { + "epoch": 3.125848981222533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 7824 + }, + { + "epoch": 3.1262485017978427, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 7825 + }, + { + "epoch": 3.126648022373152, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7826 + }, + { + "epoch": 3.127047542948462, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 7827 + }, + { + "epoch": 3.1274470635237717, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 7828 + }, + { + "epoch": 3.127846584099081, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7829 + }, + { + "epoch": 3.128246104674391, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 7830 + }, + { + "epoch": 3.1286456252497006, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7831 + }, + { + "epoch": 3.12904514582501, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7832 + }, + { + "epoch": 3.1294446664003197, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 7833 + }, + { + "epoch": 3.129844186975629, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 7834 + }, + { + "epoch": 3.130243707550939, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 7835 + }, + { + "epoch": 3.1306432281262486, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 7836 + }, + { + "epoch": 3.131042748701558, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7837 + }, + { + "epoch": 3.1314422692768678, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 7838 + }, + { + "epoch": 3.1318417898521775, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 7839 + }, + { + "epoch": 3.132241310427487, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 7840 + }, + { + "epoch": 3.1326408310027967, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 7841 + }, + { + "epoch": 3.1330403515781065, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 7842 + }, + { + "epoch": 3.133439872153416, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7843 + }, + { + "epoch": 3.1338393927287256, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7844 + }, + { + "epoch": 3.134238913304035, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7845 + }, + { + "epoch": 3.1346384338793447, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 7846 + }, + { + "epoch": 3.1350379544546545, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7847 + }, + { + "epoch": 3.135437475029964, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7848 + }, + { + "epoch": 3.1358369956052736, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7849 + }, + { + "epoch": 3.1362365161805834, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 7850 + }, + { + "epoch": 3.1366360367558928, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7851 + }, + { + "epoch": 3.1370355573312025, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 7852 + }, + { + "epoch": 3.1374350779065123, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 7853 + }, + { + "epoch": 3.1378345984818217, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 7854 + }, + { + "epoch": 3.1382341190571315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 7855 + }, + { + "epoch": 3.1386336396324412, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 7856 + }, + { + "epoch": 3.1390331602077506, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7857 + }, + { + "epoch": 3.1394326807830604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7858 + }, + { + "epoch": 3.13983220135837, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7859 + }, + { + "epoch": 3.1402317219336795, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7860 + }, + { + "epoch": 3.1406312425089893, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7861 + }, + { + "epoch": 3.1410307630842986, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 7862 + }, + { + "epoch": 3.1414302836596084, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7863 + }, + { + "epoch": 3.141829804234918, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7864 + }, + { + "epoch": 3.1422293248102275, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 7865 + }, + { + "epoch": 3.1426288453855373, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 7866 + }, + { + "epoch": 3.143028365960847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7867 + }, + { + "epoch": 3.1434278865361565, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 7868 + }, + { + "epoch": 3.1438274071114662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 7869 + }, + { + "epoch": 3.144226927686776, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3455, + "step": 7870 + }, + { + "epoch": 3.1446264482620854, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7871 + }, + { + "epoch": 3.145025968837395, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7872 + }, + { + "epoch": 3.145425489412705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 7873 + }, + { + "epoch": 3.1458250099880143, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 7874 + }, + { + "epoch": 3.146224530563324, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 7875 + }, + { + "epoch": 3.146624051138634, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7876 + }, + { + "epoch": 3.147023571713943, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 7877 + }, + { + "epoch": 3.147423092289253, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3398, + "step": 7878 + }, + { + "epoch": 3.1478226128645623, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7879 + }, + { + "epoch": 3.148222133439872, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7880 + }, + { + "epoch": 3.148621654015182, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 7881 + }, + { + "epoch": 3.1490211745904912, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7882 + }, + { + "epoch": 3.149420695165801, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 7883 + }, + { + "epoch": 3.149820215741111, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7884 + }, + { + "epoch": 3.15021973631642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 7885 + }, + { + "epoch": 3.15061925689173, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7886 + }, + { + "epoch": 3.1510187774670397, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 7887 + }, + { + "epoch": 3.151418298042349, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 7888 + }, + { + "epoch": 3.151817818617659, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7889 + }, + { + "epoch": 3.152217339192968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 7890 + }, + { + "epoch": 3.152616859768278, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 7891 + }, + { + "epoch": 3.1530163803435878, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 7892 + }, + { + "epoch": 3.153415900918897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7893 + }, + { + "epoch": 3.153815421494207, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7894 + }, + { + "epoch": 3.1542149420695167, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7895 + }, + { + "epoch": 3.154614462644826, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 7896 + }, + { + "epoch": 3.155013983220136, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7897 + }, + { + "epoch": 3.1554135037954456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 7898 + }, + { + "epoch": 3.155813024370755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7899 + }, + { + "epoch": 3.1562125449460647, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 7900 + }, + { + "epoch": 3.1566120655213745, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 7901 + }, + { + "epoch": 3.157011586096684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 7902 + }, + { + "epoch": 3.1574111066719937, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 7903 + }, + { + "epoch": 3.1578106272473034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 7904 + }, + { + "epoch": 3.158210147822613, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 7905 + }, + { + "epoch": 3.1586096683979226, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 7906 + }, + { + "epoch": 3.159009188973232, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 7907 + }, + { + "epoch": 3.1594087095485417, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 7908 + }, + { + "epoch": 3.1598082301238515, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3461, + "step": 7909 + }, + { + "epoch": 3.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 7910 + }, + { + "epoch": 3.1606072712744706, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7911 + }, + { + "epoch": 3.1610067918497804, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7912 + }, + { + "epoch": 3.1614063124250897, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 7913 + }, + { + "epoch": 3.1618058330003995, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 7914 + }, + { + "epoch": 3.1622053535757093, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 7915 + }, + { + "epoch": 3.1626048741510187, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7916 + }, + { + "epoch": 3.1630043947263284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 7917 + }, + { + "epoch": 3.1634039153016382, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7918 + }, + { + "epoch": 3.1638034358769476, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 7919 + }, + { + "epoch": 3.1642029564522574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 7920 + }, + { + "epoch": 3.164602477027567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 7921 + }, + { + "epoch": 3.1650019976028765, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7922 + }, + { + "epoch": 3.1654015181781863, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 7923 + }, + { + "epoch": 3.1658010387534956, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 7924 + }, + { + "epoch": 3.1662005593288054, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 7925 + }, + { + "epoch": 3.166600079904115, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 7926 + }, + { + "epoch": 3.1669996004794245, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 7927 + }, + { + "epoch": 3.1673991210547343, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7928 + }, + { + "epoch": 3.167798641630044, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 7929 + }, + { + "epoch": 3.1681981622053534, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 7930 + }, + { + "epoch": 3.1685976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 7931 + }, + { + "epoch": 3.168997203355973, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 7932 + }, + { + "epoch": 3.1693967239312824, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 7933 + }, + { + "epoch": 3.169796244506592, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 7934 + }, + { + "epoch": 3.1701957650819015, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 7935 + }, + { + "epoch": 3.1705952856572113, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 7936 + }, + { + "epoch": 3.170994806232521, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 7937 + }, + { + "epoch": 3.1713943268078304, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 7938 + }, + { + "epoch": 3.17179384738314, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 7939 + }, + { + "epoch": 3.17219336795845, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 7940 + }, + { + "epoch": 3.1725928885337593, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 7941 + }, + { + "epoch": 3.172992409109069, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 7942 + }, + { + "epoch": 3.173391929684379, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7943 + }, + { + "epoch": 3.1737914502596882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 7944 + }, + { + "epoch": 3.174190970834998, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 7945 + }, + { + "epoch": 3.174590491410308, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7946 + }, + { + "epoch": 3.174990011985617, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 7947 + }, + { + "epoch": 3.175389532560927, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 7948 + }, + { + "epoch": 3.1757890531362367, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 7949 + }, + { + "epoch": 3.176188573711546, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3436, + "step": 7950 + }, + { + "epoch": 3.176588094286856, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 7951 + }, + { + "epoch": 3.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7952 + }, + { + "epoch": 3.177387135437475, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 7953 + }, + { + "epoch": 3.1777866560127848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 7954 + }, + { + "epoch": 3.178186176588094, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 7955 + }, + { + "epoch": 3.178585697163404, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 7956 + }, + { + "epoch": 3.1789852177387137, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 7957 + }, + { + "epoch": 3.179384738314023, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 7958 + }, + { + "epoch": 3.179784258889333, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 7959 + }, + { + "epoch": 3.1801837794646426, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 7960 + }, + { + "epoch": 3.180583300039952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 7961 + }, + { + "epoch": 3.1809828206152617, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7962 + }, + { + "epoch": 3.1813823411905715, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 7963 + }, + { + "epoch": 3.181781861765881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 7964 + }, + { + "epoch": 3.1821813823411906, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7965 + }, + { + "epoch": 3.1825809029165004, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 7966 + }, + { + "epoch": 3.1829804234918098, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 7967 + }, + { + "epoch": 3.1833799440671195, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7968 + }, + { + "epoch": 3.183779464642429, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 7969 + }, + { + "epoch": 3.1841789852177387, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7970 + }, + { + "epoch": 3.1845785057930485, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 7971 + }, + { + "epoch": 3.184978026368358, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 7972 + }, + { + "epoch": 3.1853775469436676, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 7973 + }, + { + "epoch": 3.1857770675189774, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 7974 + }, + { + "epoch": 3.1861765880942867, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 7975 + }, + { + "epoch": 3.1865761086695965, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 7976 + }, + { + "epoch": 3.1869756292449063, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 7977 + }, + { + "epoch": 3.1873751498202156, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 7978 + }, + { + "epoch": 3.1877746703955254, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 7979 + }, + { + "epoch": 3.1881741909708348, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 7980 + }, + { + "epoch": 3.1885737115461446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 7981 + }, + { + "epoch": 3.1889732321214543, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 7982 + }, + { + "epoch": 3.1893727526967637, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 7983 + }, + { + "epoch": 3.1897722732720735, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 7984 + }, + { + "epoch": 3.1901717938473833, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 7985 + }, + { + "epoch": 3.1905713144226926, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 7986 + }, + { + "epoch": 3.1909708349980024, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 7987 + }, + { + "epoch": 3.191370355573312, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 7988 + }, + { + "epoch": 3.1917698761486215, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 7989 + }, + { + "epoch": 3.1921693967239313, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 7990 + }, + { + "epoch": 3.192568917299241, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 7991 + }, + { + "epoch": 3.1929684378745504, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 7992 + }, + { + "epoch": 3.19336795844986, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 7993 + }, + { + "epoch": 3.19376747902517, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 7994 + }, + { + "epoch": 3.1941669996004793, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 7995 + }, + { + "epoch": 3.194566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 7996 + }, + { + "epoch": 3.1949660407510985, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 7997 + }, + { + "epoch": 3.1953655613264083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 7998 + }, + { + "epoch": 3.195765081901718, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 7999 + }, + { + "epoch": 3.1961646024770274, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8000 + }, + { + "epoch": 3.196564123052337, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8001 + }, + { + "epoch": 3.196963643627647, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8002 + }, + { + "epoch": 3.1973631642029563, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8003 + }, + { + "epoch": 3.197762684778266, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8004 + }, + { + "epoch": 3.198162205353576, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8005 + }, + { + "epoch": 3.198561725928885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 8006 + }, + { + "epoch": 3.198961246504195, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8007 + }, + { + "epoch": 3.199360767079505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8008 + }, + { + "epoch": 3.199760287654814, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8009 + }, + { + "epoch": 3.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 8010 + }, + { + "epoch": 3.2005593288054337, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8011 + }, + { + "epoch": 3.200958849380743, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8012 + }, + { + "epoch": 3.201358369956053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8013 + }, + { + "epoch": 3.201757890531362, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8014 + }, + { + "epoch": 3.202157411106672, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8015 + }, + { + "epoch": 3.2025569316819817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8016 + }, + { + "epoch": 3.202956452257291, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8017 + }, + { + "epoch": 3.203355972832601, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8018 + }, + { + "epoch": 3.2037554934079107, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8019 + }, + { + "epoch": 3.20415501398322, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 8020 + }, + { + "epoch": 3.20455453455853, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8021 + }, + { + "epoch": 3.2049540551338396, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 8022 + }, + { + "epoch": 3.205353575709149, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8023 + }, + { + "epoch": 3.2057530962844587, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8024 + }, + { + "epoch": 3.206152616859768, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8025 + }, + { + "epoch": 3.206552137435078, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8026 + }, + { + "epoch": 3.2069516580103876, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8027 + }, + { + "epoch": 3.207351178585697, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8028 + }, + { + "epoch": 3.2077506991610067, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8029 + }, + { + "epoch": 3.2081502197363165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 8030 + }, + { + "epoch": 3.208549740311626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8031 + }, + { + "epoch": 3.2089492608869357, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8032 + }, + { + "epoch": 3.2093487814622454, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8033 + }, + { + "epoch": 3.209748302037555, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8034 + }, + { + "epoch": 3.2101478226128646, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8035 + }, + { + "epoch": 3.2105473431881744, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8036 + }, + { + "epoch": 3.2109468637634837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 8037 + }, + { + "epoch": 3.2113463843387935, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8038 + }, + { + "epoch": 3.2117459049141033, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8039 + }, + { + "epoch": 3.2121454254894126, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8040 + }, + { + "epoch": 3.2125449460647224, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8041 + }, + { + "epoch": 3.2129444666400317, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8042 + }, + { + "epoch": 3.2133439872153415, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8043 + }, + { + "epoch": 3.2137435077906513, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8044 + }, + { + "epoch": 3.2141430283659607, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8045 + }, + { + "epoch": 3.2145425489412705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8046 + }, + { + "epoch": 3.2149420695165802, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8047 + }, + { + "epoch": 3.2153415900918896, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8048 + }, + { + "epoch": 3.2157411106671994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8049 + }, + { + "epoch": 3.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8050 + }, + { + "epoch": 3.2165401518178185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8051 + }, + { + "epoch": 3.2169396723931283, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8052 + }, + { + "epoch": 3.217339192968438, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8053 + }, + { + "epoch": 3.2177387135437474, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8054 + }, + { + "epoch": 3.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8055 + }, + { + "epoch": 3.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8056 + }, + { + "epoch": 3.2189372752696763, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8057 + }, + { + "epoch": 3.219336795844986, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8058 + }, + { + "epoch": 3.2197363164202955, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8059 + }, + { + "epoch": 3.2201358369956052, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3389, + "step": 8060 + }, + { + "epoch": 3.220535357570915, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8061 + }, + { + "epoch": 3.2209348781462244, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8062 + }, + { + "epoch": 3.221334398721534, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8063 + }, + { + "epoch": 3.221733919296844, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8064 + }, + { + "epoch": 3.2221334398721533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8065 + }, + { + "epoch": 3.222532960447463, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 8066 + }, + { + "epoch": 3.222932481022773, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8067 + }, + { + "epoch": 3.223332001598082, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3422, + "step": 8068 + }, + { + "epoch": 3.223731522173392, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8069 + }, + { + "epoch": 3.2241310427487013, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8070 + }, + { + "epoch": 3.224530563324011, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8071 + }, + { + "epoch": 3.224930083899321, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8072 + }, + { + "epoch": 3.2253296044746302, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8073 + }, + { + "epoch": 3.22572912504994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8074 + }, + { + "epoch": 3.22612864562525, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8075 + }, + { + "epoch": 3.226528166200559, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8076 + }, + { + "epoch": 3.226927686775869, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8077 + }, + { + "epoch": 3.2273272073511787, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8078 + }, + { + "epoch": 3.227726727926488, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8079 + }, + { + "epoch": 3.228126248501798, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8080 + }, + { + "epoch": 3.2285257690771076, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8081 + }, + { + "epoch": 3.228925289652417, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8082 + }, + { + "epoch": 3.2293248102277268, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8083 + }, + { + "epoch": 3.2297243308030366, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8084 + }, + { + "epoch": 3.230123851378346, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8085 + }, + { + "epoch": 3.2305233719536557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8086 + }, + { + "epoch": 3.230922892528965, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8087 + }, + { + "epoch": 3.231322413104275, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8088 + }, + { + "epoch": 3.2317219336795846, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8089 + }, + { + "epoch": 3.232121454254894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8090 + }, + { + "epoch": 3.2325209748302037, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8091 + }, + { + "epoch": 3.2329204954055135, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8092 + }, + { + "epoch": 3.233320015980823, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8093 + }, + { + "epoch": 3.2337195365561326, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8094 + }, + { + "epoch": 3.2341190571314424, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8095 + }, + { + "epoch": 3.2345185777067518, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8096 + }, + { + "epoch": 3.2349180982820616, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 8097 + }, + { + "epoch": 3.2353176188573713, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 8098 + }, + { + "epoch": 3.2357171394326807, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8099 + }, + { + "epoch": 3.2361166600079905, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3407, + "step": 8100 + }, + { + "epoch": 3.2365161805833003, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8101 + }, + { + "epoch": 3.2369157011586096, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 8102 + }, + { + "epoch": 3.2373152217339194, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8103 + }, + { + "epoch": 3.2377147423092287, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8104 + }, + { + "epoch": 3.2381142628845385, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8105 + }, + { + "epoch": 3.2385137834598483, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8106 + }, + { + "epoch": 3.2389133040351576, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3425, + "step": 8107 + }, + { + "epoch": 3.2393128246104674, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8108 + }, + { + "epoch": 3.239712345185777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8109 + }, + { + "epoch": 3.2401118657610866, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8110 + }, + { + "epoch": 3.2405113863363963, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 8111 + }, + { + "epoch": 3.240910906911706, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8112 + }, + { + "epoch": 3.2413104274870155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8113 + }, + { + "epoch": 3.2417099480623253, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8114 + }, + { + "epoch": 3.2421094686376346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8115 + }, + { + "epoch": 3.2425089892129444, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8116 + }, + { + "epoch": 3.242908509788254, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8117 + }, + { + "epoch": 3.2433080303635635, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8118 + }, + { + "epoch": 3.2437075509388733, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3417, + "step": 8119 + }, + { + "epoch": 3.244107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8120 + }, + { + "epoch": 3.2445065920894924, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8121 + }, + { + "epoch": 3.2449061126648022, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8122 + }, + { + "epoch": 3.245305633240112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8123 + }, + { + "epoch": 3.2457051538154214, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8124 + }, + { + "epoch": 3.246104674390731, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8125 + }, + { + "epoch": 3.246504194966041, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8126 + }, + { + "epoch": 3.2469037155413503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8127 + }, + { + "epoch": 3.24730323611666, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8128 + }, + { + "epoch": 3.24770275669197, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8129 + }, + { + "epoch": 3.248102277267279, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8130 + }, + { + "epoch": 3.248501797842589, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8131 + }, + { + "epoch": 3.2489013184178983, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8132 + }, + { + "epoch": 3.249300838993208, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8133 + }, + { + "epoch": 3.249700359568518, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8134 + }, + { + "epoch": 3.2500998801438272, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8135 + }, + { + "epoch": 3.250499400719137, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 8136 + }, + { + "epoch": 3.250898921294447, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8137 + }, + { + "epoch": 3.251298441869756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8138 + }, + { + "epoch": 3.251697962445066, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 8139 + }, + { + "epoch": 3.2520974830203757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8140 + }, + { + "epoch": 3.252497003595685, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8141 + }, + { + "epoch": 3.252896524170995, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 8142 + }, + { + "epoch": 3.2532960447463046, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8143 + }, + { + "epoch": 3.253695565321614, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8144 + }, + { + "epoch": 3.2540950858969238, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8145 + }, + { + "epoch": 3.2544946064722335, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8146 + }, + { + "epoch": 3.254894127047543, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8147 + }, + { + "epoch": 3.2552936476228527, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8148 + }, + { + "epoch": 3.255693168198162, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8149 + }, + { + "epoch": 3.256092688773472, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8150 + }, + { + "epoch": 3.2564922093487816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8151 + }, + { + "epoch": 3.256891729924091, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8152 + }, + { + "epoch": 3.2572912504994007, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8153 + }, + { + "epoch": 3.2576907710747105, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8154 + }, + { + "epoch": 3.25809029165002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8155 + }, + { + "epoch": 3.2584898122253296, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8156 + }, + { + "epoch": 3.2588893328006394, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8157 + }, + { + "epoch": 3.2592888533759488, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8158 + }, + { + "epoch": 3.2596883739512585, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8159 + }, + { + "epoch": 3.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8160 + }, + { + "epoch": 3.2604874151018777, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8161 + }, + { + "epoch": 3.2608869356771875, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8162 + }, + { + "epoch": 3.2612864562524972, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8163 + }, + { + "epoch": 3.2616859768278066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8164 + }, + { + "epoch": 3.2620854974031164, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 8165 + }, + { + "epoch": 3.2624850179784257, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8166 + }, + { + "epoch": 3.2628845385537355, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8167 + }, + { + "epoch": 3.2632840591290453, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8168 + }, + { + "epoch": 3.2636835797043546, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8169 + }, + { + "epoch": 3.2640831002796644, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8170 + }, + { + "epoch": 3.264482620854974, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8171 + }, + { + "epoch": 3.2648821414302835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8172 + }, + { + "epoch": 3.2652816620055933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8173 + }, + { + "epoch": 3.265681182580903, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8174 + }, + { + "epoch": 3.2660807031562125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8175 + }, + { + "epoch": 3.2664802237315222, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8176 + }, + { + "epoch": 3.2668797443068316, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8177 + }, + { + "epoch": 3.2672792648821414, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8178 + }, + { + "epoch": 3.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8179 + }, + { + "epoch": 3.2680783060327605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8180 + }, + { + "epoch": 3.2684778266080703, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8181 + }, + { + "epoch": 3.26887734718338, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8182 + }, + { + "epoch": 3.2692768677586894, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8183 + }, + { + "epoch": 3.269676388333999, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8184 + }, + { + "epoch": 3.270075908909309, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8185 + }, + { + "epoch": 3.2704754294846183, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8186 + }, + { + "epoch": 3.270874950059928, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3457, + "step": 8187 + }, + { + "epoch": 3.271274470635238, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8188 + }, + { + "epoch": 3.2716739912105472, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8189 + }, + { + "epoch": 3.272073511785857, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8190 + }, + { + "epoch": 3.272473032361167, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8191 + }, + { + "epoch": 3.272872552936476, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 8192 + }, + { + "epoch": 3.273272073511786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8193 + }, + { + "epoch": 3.2736715940870953, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8194 + }, + { + "epoch": 3.274071114662405, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8195 + }, + { + "epoch": 3.274470635237715, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8196 + }, + { + "epoch": 3.274870155813024, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 8197 + }, + { + "epoch": 3.275269676388334, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8198 + }, + { + "epoch": 3.275669196963644, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8199 + }, + { + "epoch": 3.276068717538953, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 8200 + }, + { + "epoch": 3.276468238114263, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8201 + }, + { + "epoch": 3.2768677586895727, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8202 + }, + { + "epoch": 3.277267279264882, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8203 + }, + { + "epoch": 3.277666799840192, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8204 + }, + { + "epoch": 3.278066320415501, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8205 + }, + { + "epoch": 3.278465840990811, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8206 + }, + { + "epoch": 3.2788653615661207, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8207 + }, + { + "epoch": 3.2792648821414305, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8208 + }, + { + "epoch": 3.27966440271674, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8209 + }, + { + "epoch": 3.2800639232920497, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8210 + }, + { + "epoch": 3.280463443867359, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 8211 + }, + { + "epoch": 3.280862964442669, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8212 + }, + { + "epoch": 3.2812624850179786, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8213 + }, + { + "epoch": 3.281662005593288, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 8214 + }, + { + "epoch": 3.2820615261685977, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8215 + }, + { + "epoch": 3.2824610467439075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8216 + }, + { + "epoch": 3.282860567319217, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8217 + }, + { + "epoch": 3.2832600878945266, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8218 + }, + { + "epoch": 3.2836596084698364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8219 + }, + { + "epoch": 3.2840591290451457, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8220 + }, + { + "epoch": 3.2844586496204555, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8221 + }, + { + "epoch": 3.284858170195765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8222 + }, + { + "epoch": 3.2852576907710747, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8223 + }, + { + "epoch": 3.2856572113463844, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8224 + }, + { + "epoch": 3.286056731921694, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8225 + }, + { + "epoch": 3.2864562524970036, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8226 + }, + { + "epoch": 3.2868557730723134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8227 + }, + { + "epoch": 3.2872552936476227, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 8228 + }, + { + "epoch": 3.2876548142229325, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8229 + }, + { + "epoch": 3.2880543347982423, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8230 + }, + { + "epoch": 3.2884538553735516, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8231 + }, + { + "epoch": 3.2888533759488614, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8232 + }, + { + "epoch": 3.289252896524171, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8233 + }, + { + "epoch": 3.2896524170994805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8234 + }, + { + "epoch": 3.2900519376747903, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8235 + }, + { + "epoch": 3.2904514582501, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8236 + }, + { + "epoch": 3.2908509788254094, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8237 + }, + { + "epoch": 3.2912504994007192, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8238 + }, + { + "epoch": 3.2916500199760286, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8239 + }, + { + "epoch": 3.2920495405513384, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8240 + }, + { + "epoch": 3.292449061126648, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8241 + }, + { + "epoch": 3.2928485817019575, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8242 + }, + { + "epoch": 3.2932481022772673, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8243 + }, + { + "epoch": 3.293647622852577, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8244 + }, + { + "epoch": 3.2940471434278864, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8245 + }, + { + "epoch": 3.294446664003196, + "grad_norm": 0.046630859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8246 + }, + { + "epoch": 3.294846184578506, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8247 + }, + { + "epoch": 3.2952457051538153, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8248 + }, + { + "epoch": 3.295645225729125, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8249 + }, + { + "epoch": 3.2960447463044344, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8250 + }, + { + "epoch": 3.2964442668797442, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8251 + }, + { + "epoch": 3.296843787455054, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8252 + }, + { + "epoch": 3.297243308030364, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8253 + }, + { + "epoch": 3.297642828605673, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8254 + }, + { + "epoch": 3.298042349180983, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8255 + }, + { + "epoch": 3.2984418697562923, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8256 + }, + { + "epoch": 3.298841390331602, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8257 + }, + { + "epoch": 3.299240910906912, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8258 + }, + { + "epoch": 3.299640431482221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8259 + }, + { + "epoch": 3.300039952057531, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8260 + }, + { + "epoch": 3.3004394726328408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8261 + }, + { + "epoch": 3.30083899320815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8262 + }, + { + "epoch": 3.30123851378346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8263 + }, + { + "epoch": 3.3016380343587697, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8264 + }, + { + "epoch": 3.302037554934079, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8265 + }, + { + "epoch": 3.302437075509389, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8266 + }, + { + "epoch": 3.302836596084698, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8267 + }, + { + "epoch": 3.303236116660008, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8268 + }, + { + "epoch": 3.3036356372353177, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8269 + }, + { + "epoch": 3.304035157810627, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8270 + }, + { + "epoch": 3.304434678385937, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8271 + }, + { + "epoch": 3.3048341989612466, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8272 + }, + { + "epoch": 3.305233719536556, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8273 + }, + { + "epoch": 3.3056332401118658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8274 + }, + { + "epoch": 3.3060327606871756, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8275 + }, + { + "epoch": 3.306432281262485, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8276 + }, + { + "epoch": 3.3068318018377947, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8277 + }, + { + "epoch": 3.3072313224131045, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8278 + }, + { + "epoch": 3.307630842988414, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3451, + "step": 8279 + }, + { + "epoch": 3.3080303635637236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8280 + }, + { + "epoch": 3.3084298841390334, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8281 + }, + { + "epoch": 3.3088294047143427, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8282 + }, + { + "epoch": 3.3092289252896525, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8283 + }, + { + "epoch": 3.309628445864962, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8284 + }, + { + "epoch": 3.3100279664402716, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8285 + }, + { + "epoch": 3.3104274870155814, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8286 + }, + { + "epoch": 3.3108270075908908, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8287 + }, + { + "epoch": 3.3112265281662006, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8288 + }, + { + "epoch": 3.3116260487415103, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8289 + }, + { + "epoch": 3.3120255693168197, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8290 + }, + { + "epoch": 3.3124250898921295, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8291 + }, + { + "epoch": 3.3128246104674393, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8292 + }, + { + "epoch": 3.3132241310427486, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8293 + }, + { + "epoch": 3.3136236516180584, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8294 + }, + { + "epoch": 3.3140231721933677, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8295 + }, + { + "epoch": 3.3144226927686775, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8296 + }, + { + "epoch": 3.3148222133439873, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8297 + }, + { + "epoch": 3.315221733919297, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8298 + }, + { + "epoch": 3.3156212544946064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8299 + }, + { + "epoch": 3.316020775069916, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 8300 + }, + { + "epoch": 3.3164202956452256, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8301 + }, + { + "epoch": 3.3168198162205353, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8302 + }, + { + "epoch": 3.317219336795845, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8303 + }, + { + "epoch": 3.3176188573711545, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8304 + }, + { + "epoch": 3.3180183779464643, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8305 + }, + { + "epoch": 3.318417898521774, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8306 + }, + { + "epoch": 3.3188174190970834, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8307 + }, + { + "epoch": 3.319216939672393, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 8308 + }, + { + "epoch": 3.319616460247703, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8309 + }, + { + "epoch": 3.3200159808230123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8310 + }, + { + "epoch": 3.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8311 + }, + { + "epoch": 3.3208150219736314, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8312 + }, + { + "epoch": 3.321214542548941, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8313 + }, + { + "epoch": 3.321614063124251, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8314 + }, + { + "epoch": 3.3220135836995603, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8315 + }, + { + "epoch": 3.32241310427487, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8316 + }, + { + "epoch": 3.32281262485018, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8317 + }, + { + "epoch": 3.3232121454254893, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8318 + }, + { + "epoch": 3.323611666000799, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8319 + }, + { + "epoch": 3.324011186576109, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8320 + }, + { + "epoch": 3.324410707151418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 8321 + }, + { + "epoch": 3.324810227726728, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8322 + }, + { + "epoch": 3.3252097483020377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8323 + }, + { + "epoch": 3.325609268877347, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8324 + }, + { + "epoch": 3.326008789452657, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8325 + }, + { + "epoch": 3.3264083100279667, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 8326 + }, + { + "epoch": 3.326807830603276, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8327 + }, + { + "epoch": 3.327207351178586, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8328 + }, + { + "epoch": 3.327606871753895, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8329 + }, + { + "epoch": 3.328006392329205, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8330 + }, + { + "epoch": 3.3284059129045147, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 8331 + }, + { + "epoch": 3.328805433479824, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8332 + }, + { + "epoch": 3.329204954055134, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8333 + }, + { + "epoch": 3.3296044746304436, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8334 + }, + { + "epoch": 3.330003995205753, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8335 + }, + { + "epoch": 3.3304035157810628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8336 + }, + { + "epoch": 3.3308030363563725, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8337 + }, + { + "epoch": 3.331202556931682, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8338 + }, + { + "epoch": 3.3316020775069917, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8339 + }, + { + "epoch": 3.332001598082301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8340 + }, + { + "epoch": 3.332401118657611, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 8341 + }, + { + "epoch": 3.3328006392329206, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8342 + }, + { + "epoch": 3.3332001598082304, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8343 + }, + { + "epoch": 3.3335996803835397, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8344 + }, + { + "epoch": 3.3339992009588495, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8345 + }, + { + "epoch": 3.334398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8346 + }, + { + "epoch": 3.3347982421094686, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8347 + }, + { + "epoch": 3.3351977626847784, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8348 + }, + { + "epoch": 3.3355972832600878, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8349 + }, + { + "epoch": 3.3359968038353975, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8350 + }, + { + "epoch": 3.3363963244107073, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8351 + }, + { + "epoch": 3.3367958449860167, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8352 + }, + { + "epoch": 3.3371953655613265, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8353 + }, + { + "epoch": 3.3375948861366362, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8354 + }, + { + "epoch": 3.3379944067119456, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8355 + }, + { + "epoch": 3.3383939272872554, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 8356 + }, + { + "epoch": 3.3387934478625647, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 8357 + }, + { + "epoch": 3.3391929684378745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8358 + }, + { + "epoch": 3.3395924890131843, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 8359 + }, + { + "epoch": 3.3399920095884936, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8360 + }, + { + "epoch": 3.3403915301638034, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8361 + }, + { + "epoch": 3.340791050739113, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8362 + }, + { + "epoch": 3.3411905713144225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 8363 + }, + { + "epoch": 3.3415900918897323, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 8364 + }, + { + "epoch": 3.341989612465042, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 8365 + }, + { + "epoch": 3.3423891330403515, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8366 + }, + { + "epoch": 3.3427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8367 + }, + { + "epoch": 3.343188174190971, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8368 + }, + { + "epoch": 3.3435876947662804, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8369 + }, + { + "epoch": 3.34398721534159, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8370 + }, + { + "epoch": 3.3443867359169, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8371 + }, + { + "epoch": 3.3447862564922093, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8372 + }, + { + "epoch": 3.345185777067519, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8373 + }, + { + "epoch": 3.3455852976428284, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 8374 + }, + { + "epoch": 3.345984818218138, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8375 + }, + { + "epoch": 3.346384338793448, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8376 + }, + { + "epoch": 3.3467838593687573, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8377 + }, + { + "epoch": 3.347183379944067, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8378 + }, + { + "epoch": 3.347582900519377, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8379 + }, + { + "epoch": 3.3479824210946862, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8380 + }, + { + "epoch": 3.348381941669996, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8381 + }, + { + "epoch": 3.348781462245306, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8382 + }, + { + "epoch": 3.349180982820615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8383 + }, + { + "epoch": 3.349580503395925, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 8384 + }, + { + "epoch": 3.3499800239712343, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8385 + }, + { + "epoch": 3.350379544546544, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8386 + }, + { + "epoch": 3.350779065121854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8387 + }, + { + "epoch": 3.3511785856971636, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8388 + }, + { + "epoch": 3.351578106272473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8389 + }, + { + "epoch": 3.3519776268477828, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8390 + }, + { + "epoch": 3.352377147423092, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8391 + }, + { + "epoch": 3.352776667998402, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8392 + }, + { + "epoch": 3.3531761885737117, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8393 + }, + { + "epoch": 3.353575709149021, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8394 + }, + { + "epoch": 3.353975229724331, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 8395 + }, + { + "epoch": 3.3543747502996406, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8396 + }, + { + "epoch": 3.35477427087495, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8397 + }, + { + "epoch": 3.3551737914502597, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8398 + }, + { + "epoch": 3.3555733120255695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8399 + }, + { + "epoch": 3.355972832600879, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 8400 + }, + { + "epoch": 3.3563723531761886, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 8401 + }, + { + "epoch": 3.356771873751498, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8402 + }, + { + "epoch": 3.357171394326808, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8403 + }, + { + "epoch": 3.3575709149021176, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 8404 + }, + { + "epoch": 3.357970435477427, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8405 + }, + { + "epoch": 3.3583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 8406 + }, + { + "epoch": 3.3587694766280465, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8407 + }, + { + "epoch": 3.359168997203356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3355, + "step": 8408 + }, + { + "epoch": 3.3595685177786656, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8409 + }, + { + "epoch": 3.3599680383539754, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8410 + }, + { + "epoch": 3.3603675589292847, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8411 + }, + { + "epoch": 3.3607670795045945, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8412 + }, + { + "epoch": 3.361166600079904, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 8413 + }, + { + "epoch": 3.3615661206552137, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8414 + }, + { + "epoch": 3.3619656412305234, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8415 + }, + { + "epoch": 3.3623651618058332, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 8416 + }, + { + "epoch": 3.3627646823811426, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8417 + }, + { + "epoch": 3.3631642029564524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8418 + }, + { + "epoch": 3.3635637235317617, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8419 + }, + { + "epoch": 3.3639632441070715, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3396, + "step": 8420 + }, + { + "epoch": 3.3643627646823813, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8421 + }, + { + "epoch": 3.3647622852576906, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8422 + }, + { + "epoch": 3.3651618058330004, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 8423 + }, + { + "epoch": 3.36556132640831, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8424 + }, + { + "epoch": 3.3659608469836195, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8425 + }, + { + "epoch": 3.3663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8426 + }, + { + "epoch": 3.366759888134239, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8427 + }, + { + "epoch": 3.3671594087095484, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8428 + }, + { + "epoch": 3.3675589292848582, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8429 + }, + { + "epoch": 3.3679584498601676, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8430 + }, + { + "epoch": 3.3683579704354774, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 8431 + }, + { + "epoch": 3.368757491010787, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8432 + }, + { + "epoch": 3.369157011586097, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8433 + }, + { + "epoch": 3.3695565321614063, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8434 + }, + { + "epoch": 3.369956052736716, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 8435 + }, + { + "epoch": 3.3703555733120254, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8436 + }, + { + "epoch": 3.370755093887335, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 8437 + }, + { + "epoch": 3.371154614462645, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8438 + }, + { + "epoch": 3.3715541350379543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8439 + }, + { + "epoch": 3.371953655613264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8440 + }, + { + "epoch": 3.372353176188574, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8441 + }, + { + "epoch": 3.3727526967638832, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 8442 + }, + { + "epoch": 3.373152217339193, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8443 + }, + { + "epoch": 3.373551737914503, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8444 + }, + { + "epoch": 3.373951258489812, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8445 + }, + { + "epoch": 3.374350779065122, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8446 + }, + { + "epoch": 3.3747502996404313, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8447 + }, + { + "epoch": 3.375149820215741, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8448 + }, + { + "epoch": 3.375549340791051, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8449 + }, + { + "epoch": 3.37594886136636, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8450 + }, + { + "epoch": 3.37634838194167, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8451 + }, + { + "epoch": 3.3767479025169798, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8452 + }, + { + "epoch": 3.377147423092289, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 8453 + }, + { + "epoch": 3.377546943667599, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8454 + }, + { + "epoch": 3.3779464642429087, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8455 + }, + { + "epoch": 3.378345984818218, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8456 + }, + { + "epoch": 3.378745505393528, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8457 + }, + { + "epoch": 3.379145025968837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8458 + }, + { + "epoch": 3.379544546544147, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8459 + }, + { + "epoch": 3.3799440671194567, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8460 + }, + { + "epoch": 3.3803435876947665, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8461 + }, + { + "epoch": 3.380743108270076, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8462 + }, + { + "epoch": 3.3811426288453856, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3397, + "step": 8463 + }, + { + "epoch": 3.381542149420695, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8464 + }, + { + "epoch": 3.3819416699960048, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8465 + }, + { + "epoch": 3.3823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 8466 + }, + { + "epoch": 3.382740711146624, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 8467 + }, + { + "epoch": 3.3831402317219337, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8468 + }, + { + "epoch": 3.3835397522972435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8469 + }, + { + "epoch": 3.383939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8470 + }, + { + "epoch": 3.3843387934478626, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8471 + }, + { + "epoch": 3.3847383140231724, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8472 + }, + { + "epoch": 3.3851378345984817, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8473 + }, + { + "epoch": 3.3855373551737915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 8474 + }, + { + "epoch": 3.385936875749101, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8475 + }, + { + "epoch": 3.3863363963244106, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8476 + }, + { + "epoch": 3.3867359168997204, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8477 + }, + { + "epoch": 3.38713543747503, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8478 + }, + { + "epoch": 3.3875349580503396, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8479 + }, + { + "epoch": 3.3879344786256493, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8480 + }, + { + "epoch": 3.3883339992009587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8481 + }, + { + "epoch": 3.3887335197762685, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8482 + }, + { + "epoch": 3.3891330403515783, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8483 + }, + { + "epoch": 3.3895325609268876, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8484 + }, + { + "epoch": 3.3899320815021974, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8485 + }, + { + "epoch": 3.390331602077507, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8486 + }, + { + "epoch": 3.3907311226528165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8487 + }, + { + "epoch": 3.3911306432281263, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 8488 + }, + { + "epoch": 3.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8489 + }, + { + "epoch": 3.3919296843787454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8490 + }, + { + "epoch": 3.392329204954055, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8491 + }, + { + "epoch": 3.3927287255293646, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8492 + }, + { + "epoch": 3.3931282461046743, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8493 + }, + { + "epoch": 3.393527766679984, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8494 + }, + { + "epoch": 3.3939272872552935, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 8495 + }, + { + "epoch": 3.3943268078306033, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8496 + }, + { + "epoch": 3.394726328405913, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8497 + }, + { + "epoch": 3.3951258489812224, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8498 + }, + { + "epoch": 3.395525369556532, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8499 + }, + { + "epoch": 3.395924890131842, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 8500 + }, + { + "epoch": 3.3963244107071513, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8501 + }, + { + "epoch": 3.396723931282461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 8502 + }, + { + "epoch": 3.3971234518577704, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8503 + }, + { + "epoch": 3.39752297243308, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8504 + }, + { + "epoch": 3.39792249300839, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8505 + }, + { + "epoch": 3.3983220135837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 8506 + }, + { + "epoch": 3.398721534159009, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8507 + }, + { + "epoch": 3.399121054734319, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8508 + }, + { + "epoch": 3.3995205753096283, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8509 + }, + { + "epoch": 3.399920095884938, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 8510 + }, + { + "epoch": 3.400319616460248, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8511 + }, + { + "epoch": 3.400719137035557, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8512 + }, + { + "epoch": 3.401118657610867, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8513 + }, + { + "epoch": 3.4015181781861767, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8514 + }, + { + "epoch": 3.401917698761486, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8515 + }, + { + "epoch": 3.402317219336796, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8516 + }, + { + "epoch": 3.4027167399121057, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8517 + }, + { + "epoch": 3.403116260487415, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8518 + }, + { + "epoch": 3.403515781062725, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 8519 + }, + { + "epoch": 3.403915301638034, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8520 + }, + { + "epoch": 3.404314822213344, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 8521 + }, + { + "epoch": 3.4047143427886537, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8522 + }, + { + "epoch": 3.4051138633639635, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8523 + }, + { + "epoch": 3.405513383939273, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8524 + }, + { + "epoch": 3.4059129045145826, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8525 + }, + { + "epoch": 3.406312425089892, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8526 + }, + { + "epoch": 3.4067119456652017, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8527 + }, + { + "epoch": 3.4071114662405115, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8528 + }, + { + "epoch": 3.407510986815821, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8529 + }, + { + "epoch": 3.4079105073911307, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8530 + }, + { + "epoch": 3.4083100279664404, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 8531 + }, + { + "epoch": 3.40870954854175, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8532 + }, + { + "epoch": 3.4091090691170596, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8533 + }, + { + "epoch": 3.4095085896923694, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8534 + }, + { + "epoch": 3.4099081102676787, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8535 + }, + { + "epoch": 3.4103076308429885, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8536 + }, + { + "epoch": 3.410707151418298, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8537 + }, + { + "epoch": 3.4111066719936076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 8538 + }, + { + "epoch": 3.4115061925689174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 8539 + }, + { + "epoch": 3.4119057131442267, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8540 + }, + { + "epoch": 3.4123052337195365, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8541 + }, + { + "epoch": 3.4127047542948463, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8542 + }, + { + "epoch": 3.4131042748701557, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 8543 + }, + { + "epoch": 3.4135037954454654, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8544 + }, + { + "epoch": 3.4139033160207752, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8545 + }, + { + "epoch": 3.4143028365960846, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8546 + }, + { + "epoch": 3.4147023571713944, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8547 + }, + { + "epoch": 3.4151018777467037, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8548 + }, + { + "epoch": 3.4155013983220135, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8549 + }, + { + "epoch": 3.4159009188973233, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8550 + }, + { + "epoch": 3.416300439472633, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8551 + }, + { + "epoch": 3.4166999600479424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8552 + }, + { + "epoch": 3.417099480623252, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 8553 + }, + { + "epoch": 3.4174990011985615, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8554 + }, + { + "epoch": 3.4178985217738713, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8555 + }, + { + "epoch": 3.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8556 + }, + { + "epoch": 3.4186975629244905, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 8557 + }, + { + "epoch": 3.4190970834998002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8558 + }, + { + "epoch": 3.41949660407511, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8559 + }, + { + "epoch": 3.4198961246504194, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 8560 + }, + { + "epoch": 3.420295645225729, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 8561 + }, + { + "epoch": 3.420695165801039, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8562 + }, + { + "epoch": 3.4210946863763483, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8563 + }, + { + "epoch": 3.421494206951658, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 8564 + }, + { + "epoch": 3.4218937275269674, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8565 + }, + { + "epoch": 3.422293248102277, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8566 + }, + { + "epoch": 3.422692768677587, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 8567 + }, + { + "epoch": 3.4230922892528968, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 8568 + }, + { + "epoch": 3.423491809828206, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8569 + }, + { + "epoch": 3.423891330403516, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 8570 + }, + { + "epoch": 3.4242908509788252, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8571 + }, + { + "epoch": 3.424690371554135, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8572 + }, + { + "epoch": 3.425089892129445, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 8573 + }, + { + "epoch": 3.425489412704754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8574 + }, + { + "epoch": 3.425888933280064, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 8575 + }, + { + "epoch": 3.4262884538553737, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 8576 + }, + { + "epoch": 3.426687974430683, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 8577 + }, + { + "epoch": 3.427087495005993, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8578 + }, + { + "epoch": 3.4274870155813026, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8579 + }, + { + "epoch": 3.427886536156612, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8580 + }, + { + "epoch": 3.4282860567319218, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 8581 + }, + { + "epoch": 3.428685577307231, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8582 + }, + { + "epoch": 3.429085097882541, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8583 + }, + { + "epoch": 3.4294846184578507, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8584 + }, + { + "epoch": 3.42988413903316, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8585 + }, + { + "epoch": 3.43028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8586 + }, + { + "epoch": 3.4306831801837796, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8587 + }, + { + "epoch": 3.431082700759089, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8588 + }, + { + "epoch": 3.4314822213343987, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8589 + }, + { + "epoch": 3.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 8590 + }, + { + "epoch": 3.432281262485018, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8591 + }, + { + "epoch": 3.4326807830603276, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3444, + "step": 8592 + }, + { + "epoch": 3.433080303635637, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8593 + }, + { + "epoch": 3.4334798242109468, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 8594 + }, + { + "epoch": 3.4338793447862566, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8595 + }, + { + "epoch": 3.4342788653615663, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8596 + }, + { + "epoch": 3.4346783859368757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8597 + }, + { + "epoch": 3.4350779065121855, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 8598 + }, + { + "epoch": 3.435477427087495, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8599 + }, + { + "epoch": 3.4358769476628046, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8600 + }, + { + "epoch": 3.4362764682381144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8601 + }, + { + "epoch": 3.4366759888134237, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8602 + }, + { + "epoch": 3.4370755093887335, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8603 + }, + { + "epoch": 3.4374750299640433, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 8604 + }, + { + "epoch": 3.4378745505393526, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3463, + "step": 8605 + }, + { + "epoch": 3.4382740711146624, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 8606 + }, + { + "epoch": 3.438673591689972, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8607 + }, + { + "epoch": 3.4390731122652816, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 8608 + }, + { + "epoch": 3.4394726328405913, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8609 + }, + { + "epoch": 3.4398721534159007, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8610 + }, + { + "epoch": 3.4402716739912105, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3411, + "step": 8611 + }, + { + "epoch": 3.4406711945665203, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8612 + }, + { + "epoch": 3.44107071514183, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 8613 + }, + { + "epoch": 3.4414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8614 + }, + { + "epoch": 3.441869756292449, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8615 + }, + { + "epoch": 3.4422692768677585, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8616 + }, + { + "epoch": 3.4426687974430683, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8617 + }, + { + "epoch": 3.443068318018378, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8618 + }, + { + "epoch": 3.4434678385936874, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8619 + }, + { + "epoch": 3.443867359168997, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8620 + }, + { + "epoch": 3.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8621 + }, + { + "epoch": 3.4446664003196164, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8622 + }, + { + "epoch": 3.445065920894926, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8623 + }, + { + "epoch": 3.445465441470236, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 8624 + }, + { + "epoch": 3.4458649620455453, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8625 + }, + { + "epoch": 3.446264482620855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8626 + }, + { + "epoch": 3.4466640031961644, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8627 + }, + { + "epoch": 3.447063523771474, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 8628 + }, + { + "epoch": 3.447463044346784, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8629 + }, + { + "epoch": 3.4478625649220933, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 8630 + }, + { + "epoch": 3.448262085497403, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8631 + }, + { + "epoch": 3.448661606072713, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8632 + }, + { + "epoch": 3.4490611266480222, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8633 + }, + { + "epoch": 3.449460647223332, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8634 + }, + { + "epoch": 3.449860167798642, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 8635 + }, + { + "epoch": 3.450259688373951, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8636 + }, + { + "epoch": 3.450659208949261, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8637 + }, + { + "epoch": 3.4510587295245703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8638 + }, + { + "epoch": 3.45145825009988, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8639 + }, + { + "epoch": 3.45185777067519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 8640 + }, + { + "epoch": 3.4522572912504996, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8641 + }, + { + "epoch": 3.452656811825809, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8642 + }, + { + "epoch": 3.4530563324011188, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8643 + }, + { + "epoch": 3.453455852976428, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8644 + }, + { + "epoch": 3.453855373551738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8645 + }, + { + "epoch": 3.4542548941270477, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8646 + }, + { + "epoch": 3.454654414702357, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8647 + }, + { + "epoch": 3.455053935277667, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8648 + }, + { + "epoch": 3.4554534558529766, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8649 + }, + { + "epoch": 3.455852976428286, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8650 + }, + { + "epoch": 3.4562524970035957, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 8651 + }, + { + "epoch": 3.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 8652 + }, + { + "epoch": 3.457051538154215, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8653 + }, + { + "epoch": 3.4574510587295246, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 8654 + }, + { + "epoch": 3.457850579304834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8655 + }, + { + "epoch": 3.4582500998801438, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8656 + }, + { + "epoch": 3.4586496204554535, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8657 + }, + { + "epoch": 3.4590491410307633, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8658 + }, + { + "epoch": 3.4594486616060727, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8659 + }, + { + "epoch": 3.4598481821813825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 8660 + }, + { + "epoch": 3.460247702756692, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 8661 + }, + { + "epoch": 3.4606472233320016, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8662 + }, + { + "epoch": 3.4610467439073114, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8663 + }, + { + "epoch": 3.4614462644826207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8664 + }, + { + "epoch": 3.4618457850579305, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8665 + }, + { + "epoch": 3.4622453056332403, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 8666 + }, + { + "epoch": 3.4626448262085496, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 8667 + }, + { + "epoch": 3.4630443467838594, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 8668 + }, + { + "epoch": 3.463443867359169, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8669 + }, + { + "epoch": 3.4638433879344785, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8670 + }, + { + "epoch": 3.4642429085097883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 8671 + }, + { + "epoch": 3.4646424290850977, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8672 + }, + { + "epoch": 3.4650419496604075, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 8673 + }, + { + "epoch": 3.4654414702357172, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8674 + }, + { + "epoch": 3.4658409908110266, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 8675 + }, + { + "epoch": 3.4662405113863364, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 8676 + }, + { + "epoch": 3.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8677 + }, + { + "epoch": 3.4670395525369555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8678 + }, + { + "epoch": 3.4674390731122653, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 8679 + }, + { + "epoch": 3.467838593687575, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8680 + }, + { + "epoch": 3.4682381142628844, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 8681 + }, + { + "epoch": 3.468637634838194, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 8682 + }, + { + "epoch": 3.4690371554135035, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 8683 + }, + { + "epoch": 3.4694366759888133, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8684 + }, + { + "epoch": 3.469836196564123, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 8685 + }, + { + "epoch": 3.470235717139433, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8686 + }, + { + "epoch": 3.4706352377147422, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8687 + }, + { + "epoch": 3.471034758290052, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8688 + }, + { + "epoch": 3.4714342788653614, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8689 + }, + { + "epoch": 3.471833799440671, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8690 + }, + { + "epoch": 3.472233320015981, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8691 + }, + { + "epoch": 3.4726328405912903, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8692 + }, + { + "epoch": 3.4730323611666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 8693 + }, + { + "epoch": 3.47343188174191, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8694 + }, + { + "epoch": 3.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8695 + }, + { + "epoch": 3.474230922892529, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 8696 + }, + { + "epoch": 3.474630443467839, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8697 + }, + { + "epoch": 3.475029964043148, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8698 + }, + { + "epoch": 3.475429484618458, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 8699 + }, + { + "epoch": 3.4758290051937673, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8700 + }, + { + "epoch": 3.476228525769077, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8701 + }, + { + "epoch": 3.476628046344387, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8702 + }, + { + "epoch": 3.4770275669196966, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8703 + }, + { + "epoch": 3.477427087495006, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8704 + }, + { + "epoch": 3.4778266080703157, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8705 + }, + { + "epoch": 3.478226128645625, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8706 + }, + { + "epoch": 3.478625649220935, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 8707 + }, + { + "epoch": 3.4790251697962447, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 8708 + }, + { + "epoch": 3.479424690371554, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8709 + }, + { + "epoch": 3.479824210946864, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 8710 + }, + { + "epoch": 3.4802237315221736, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 8711 + }, + { + "epoch": 3.480623252097483, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8712 + }, + { + "epoch": 3.4810227726727927, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8713 + }, + { + "epoch": 3.4814222932481025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8714 + }, + { + "epoch": 3.481821813823412, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8715 + }, + { + "epoch": 3.4822213343987216, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 8716 + }, + { + "epoch": 3.482620854974031, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8717 + }, + { + "epoch": 3.4830203755493407, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 8718 + }, + { + "epoch": 3.4834198961246505, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 8719 + }, + { + "epoch": 3.48381941669996, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 8720 + }, + { + "epoch": 3.4842189372752697, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 8721 + }, + { + "epoch": 3.4846184578505794, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8722 + }, + { + "epoch": 3.485017978425889, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8723 + }, + { + "epoch": 3.4854174990011986, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 8724 + }, + { + "epoch": 3.4858170195765084, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8725 + }, + { + "epoch": 3.4862165401518177, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8726 + }, + { + "epoch": 3.4866160607271275, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 8727 + }, + { + "epoch": 3.487015581302437, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8728 + }, + { + "epoch": 3.4874151018777466, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8729 + }, + { + "epoch": 3.4878146224530564, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8730 + }, + { + "epoch": 3.488214143028366, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8731 + }, + { + "epoch": 3.4886136636036755, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 8732 + }, + { + "epoch": 3.4890131841789853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8733 + }, + { + "epoch": 3.4894127047542947, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8734 + }, + { + "epoch": 3.4898122253296044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8735 + }, + { + "epoch": 3.4902117459049142, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8736 + }, + { + "epoch": 3.4906112664802236, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 8737 + }, + { + "epoch": 3.4910107870555334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8738 + }, + { + "epoch": 3.491410307630843, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 8739 + }, + { + "epoch": 3.4918098282061525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 8740 + }, + { + "epoch": 3.4922093487814623, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8741 + }, + { + "epoch": 3.492608869356772, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 8742 + }, + { + "epoch": 3.4930083899320814, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 8743 + }, + { + "epoch": 3.493407910507391, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8744 + }, + { + "epoch": 3.4938074310827005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8745 + }, + { + "epoch": 3.4942069516580103, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8746 + }, + { + "epoch": 3.49460647223332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 8747 + }, + { + "epoch": 3.49500599280863, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8748 + }, + { + "epoch": 3.4954055133839392, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8749 + }, + { + "epoch": 3.495805033959249, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8750 + }, + { + "epoch": 3.4962045545345584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 8751 + }, + { + "epoch": 3.496604075109868, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8752 + }, + { + "epoch": 3.497003595685178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8753 + }, + { + "epoch": 3.4974031162604873, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 8754 + }, + { + "epoch": 3.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8755 + }, + { + "epoch": 3.498202157411107, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8756 + }, + { + "epoch": 3.498601677986416, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 8757 + }, + { + "epoch": 3.499001198561726, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 8758 + }, + { + "epoch": 3.4994007191370358, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8759 + }, + { + "epoch": 3.499800239712345, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8760 + }, + { + "epoch": 3.500199760287655, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8761 + }, + { + "epoch": 3.5005992808629642, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8762 + }, + { + "epoch": 3.500998801438274, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8763 + }, + { + "epoch": 3.501398322013584, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 8764 + }, + { + "epoch": 3.5017978425888936, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8765 + }, + { + "epoch": 3.502197363164203, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 8766 + }, + { + "epoch": 3.5025968837395127, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8767 + }, + { + "epoch": 3.502996404314822, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8768 + }, + { + "epoch": 3.503395924890132, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3365, + "step": 8769 + }, + { + "epoch": 3.5037954454654416, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 8770 + }, + { + "epoch": 3.504194966040751, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8771 + }, + { + "epoch": 3.5045944866160608, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 8772 + }, + { + "epoch": 3.50499400719137, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8773 + }, + { + "epoch": 3.50539352776668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8774 + }, + { + "epoch": 3.5057930483419897, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 8775 + }, + { + "epoch": 3.5061925689172995, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8776 + }, + { + "epoch": 3.506592089492609, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8777 + }, + { + "epoch": 3.5069916100679186, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8778 + }, + { + "epoch": 3.507391130643228, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 8779 + }, + { + "epoch": 3.5077906512185377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 8780 + }, + { + "epoch": 3.5081901717938475, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 8781 + }, + { + "epoch": 3.508589692369157, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8782 + }, + { + "epoch": 3.5089892129444666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 8783 + }, + { + "epoch": 3.5093887335197764, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 8784 + }, + { + "epoch": 3.5097882540950858, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8785 + }, + { + "epoch": 3.5101877746703956, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8786 + }, + { + "epoch": 3.5105872952457053, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8787 + }, + { + "epoch": 3.5109868158210147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 8788 + }, + { + "epoch": 3.5113863363963245, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 8789 + }, + { + "epoch": 3.511785856971634, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8790 + }, + { + "epoch": 3.5121853775469436, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8791 + }, + { + "epoch": 3.5125848981222534, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8792 + }, + { + "epoch": 3.512984418697563, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8793 + }, + { + "epoch": 3.5133839392728725, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8794 + }, + { + "epoch": 3.5137834598481823, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8795 + }, + { + "epoch": 3.5141829804234916, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8796 + }, + { + "epoch": 3.5145825009988014, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 8797 + }, + { + "epoch": 3.514982021574111, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 8798 + }, + { + "epoch": 3.5153815421494206, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 8799 + }, + { + "epoch": 3.5157810627247303, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8800 + }, + { + "epoch": 3.5161805833000397, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8801 + }, + { + "epoch": 3.5165801038753495, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8802 + }, + { + "epoch": 3.5169796244506593, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8803 + }, + { + "epoch": 3.517379145025969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8804 + }, + { + "epoch": 3.5177786656012784, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8805 + }, + { + "epoch": 3.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 8806 + }, + { + "epoch": 3.5185777067518975, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8807 + }, + { + "epoch": 3.5189772273272073, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8808 + }, + { + "epoch": 3.519376747902517, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 8809 + }, + { + "epoch": 3.519776268477827, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8810 + }, + { + "epoch": 3.520175789053136, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 8811 + }, + { + "epoch": 3.520575309628446, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 8812 + }, + { + "epoch": 3.5209748302037553, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 8813 + }, + { + "epoch": 3.521374350779065, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8814 + }, + { + "epoch": 3.521773871354375, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8815 + }, + { + "epoch": 3.5221733919296843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 8816 + }, + { + "epoch": 3.522572912504994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 8817 + }, + { + "epoch": 3.5229724330803034, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8818 + }, + { + "epoch": 3.523371953655613, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 8819 + }, + { + "epoch": 3.523771474230923, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 8820 + }, + { + "epoch": 3.5241709948062327, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8821 + }, + { + "epoch": 3.524570515381542, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 8822 + }, + { + "epoch": 3.524970035956852, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 8823 + }, + { + "epoch": 3.525369556532161, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8824 + }, + { + "epoch": 3.525769077107471, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8825 + }, + { + "epoch": 3.526168597682781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8826 + }, + { + "epoch": 3.52656811825809, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 8827 + }, + { + "epoch": 3.5269676388334, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 8828 + }, + { + "epoch": 3.5273671594087097, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8829 + }, + { + "epoch": 3.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 8830 + }, + { + "epoch": 3.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8831 + }, + { + "epoch": 3.5285657211346386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8832 + }, + { + "epoch": 3.528965241709948, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8833 + }, + { + "epoch": 3.5293647622852578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8834 + }, + { + "epoch": 3.529764282860567, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8835 + }, + { + "epoch": 3.530163803435877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8836 + }, + { + "epoch": 3.5305633240111867, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 8837 + }, + { + "epoch": 3.5309628445864965, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 8838 + }, + { + "epoch": 3.531362365161806, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8839 + }, + { + "epoch": 3.5317618857371156, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 8840 + }, + { + "epoch": 3.532161406312425, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8841 + }, + { + "epoch": 3.5325609268877347, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 8842 + }, + { + "epoch": 3.5329604474630445, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8843 + }, + { + "epoch": 3.533359968038354, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8844 + }, + { + "epoch": 3.5337594886136636, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 8845 + }, + { + "epoch": 3.534159009188973, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 8846 + }, + { + "epoch": 3.5345585297642828, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 8847 + }, + { + "epoch": 3.5349580503395925, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8848 + }, + { + "epoch": 3.5353575709149023, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 8849 + }, + { + "epoch": 3.5357570914902117, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 8850 + }, + { + "epoch": 3.5361566120655215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8851 + }, + { + "epoch": 3.536556132640831, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8852 + }, + { + "epoch": 3.5369556532161406, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 8853 + }, + { + "epoch": 3.5373551737914504, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 8854 + }, + { + "epoch": 3.53775469436676, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8855 + }, + { + "epoch": 3.5381542149420695, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8856 + }, + { + "epoch": 3.5385537355173793, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 8857 + }, + { + "epoch": 3.5389532560926886, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 8858 + }, + { + "epoch": 3.5393527766679984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 8859 + }, + { + "epoch": 3.539752297243308, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 8860 + }, + { + "epoch": 3.5401518178186175, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 8861 + }, + { + "epoch": 3.5405513383939273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 8862 + }, + { + "epoch": 3.5409508589692367, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 8863 + }, + { + "epoch": 3.5413503795445465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 8864 + }, + { + "epoch": 3.5417499001198562, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3416, + "step": 8865 + }, + { + "epoch": 3.542149420695166, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8866 + }, + { + "epoch": 3.5425489412704754, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 8867 + }, + { + "epoch": 3.542948461845785, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8868 + }, + { + "epoch": 3.5433479824210945, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 8869 + }, + { + "epoch": 3.5437475029964043, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 8870 + }, + { + "epoch": 3.544147023571714, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 8871 + }, + { + "epoch": 3.5445465441470234, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8872 + }, + { + "epoch": 3.544946064722333, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 8873 + }, + { + "epoch": 3.545345585297643, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8874 + }, + { + "epoch": 3.5457451058729523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 8875 + }, + { + "epoch": 3.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8876 + }, + { + "epoch": 3.546544147023572, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8877 + }, + { + "epoch": 3.5469436675988812, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 8878 + }, + { + "epoch": 3.547343188174191, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 8879 + }, + { + "epoch": 3.5477427087495004, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 8880 + }, + { + "epoch": 3.54814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 8881 + }, + { + "epoch": 3.54854174990012, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 8882 + }, + { + "epoch": 3.5489412704754297, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8883 + }, + { + "epoch": 3.549340791050739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8884 + }, + { + "epoch": 3.549740311626049, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 8885 + }, + { + "epoch": 3.550139832201358, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 8886 + }, + { + "epoch": 3.550539352776668, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8887 + }, + { + "epoch": 3.5509388733519778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 8888 + }, + { + "epoch": 3.551338393927287, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 8889 + }, + { + "epoch": 3.551737914502597, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 8890 + }, + { + "epoch": 3.5521374350779062, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 8891 + }, + { + "epoch": 3.552536955653216, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 8892 + }, + { + "epoch": 3.552936476228526, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8893 + }, + { + "epoch": 3.5533359968038356, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8894 + }, + { + "epoch": 3.553735517379145, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8895 + }, + { + "epoch": 3.5541350379544547, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 8896 + }, + { + "epoch": 3.554534558529764, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 8897 + }, + { + "epoch": 3.554934079105074, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 8898 + }, + { + "epoch": 3.5553335996803836, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 8899 + }, + { + "epoch": 3.5557331202556934, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 8900 + }, + { + "epoch": 3.556132640831003, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8901 + }, + { + "epoch": 3.5565321614063126, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8902 + }, + { + "epoch": 3.556931681981622, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 8903 + }, + { + "epoch": 3.5573312025569317, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8904 + }, + { + "epoch": 3.5577307231322415, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8905 + }, + { + "epoch": 3.558130243707551, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 8906 + }, + { + "epoch": 3.5585297642828606, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 8907 + }, + { + "epoch": 3.55892928485817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 8908 + }, + { + "epoch": 3.5593288054334797, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 8909 + }, + { + "epoch": 3.5597283260087895, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8910 + }, + { + "epoch": 3.5601278465840993, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8911 + }, + { + "epoch": 3.5605273671594087, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 8912 + }, + { + "epoch": 3.5609268877347184, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 8913 + }, + { + "epoch": 3.561326408310028, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 8914 + }, + { + "epoch": 3.5617259288853376, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8915 + }, + { + "epoch": 3.5621254494606474, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 8916 + }, + { + "epoch": 3.5625249700359567, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8917 + }, + { + "epoch": 3.5629244906112665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 8918 + }, + { + "epoch": 3.5633240111865763, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 8919 + }, + { + "epoch": 3.5637235317618856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 8920 + }, + { + "epoch": 3.5641230523371954, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 8921 + }, + { + "epoch": 3.564522572912505, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 8922 + }, + { + "epoch": 3.5649220934878145, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8923 + }, + { + "epoch": 3.5653216140631243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 8924 + }, + { + "epoch": 3.5657211346384337, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 8925 + }, + { + "epoch": 3.5661206552137434, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8926 + }, + { + "epoch": 3.5665201757890532, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8927 + }, + { + "epoch": 3.566919696364363, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 8928 + }, + { + "epoch": 3.5673192169396724, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 8929 + }, + { + "epoch": 3.567718737514982, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 8930 + }, + { + "epoch": 3.5681182580902915, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 8931 + }, + { + "epoch": 3.5685177786656013, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 8932 + }, + { + "epoch": 3.568917299240911, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 8933 + }, + { + "epoch": 3.5693168198162204, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 8934 + }, + { + "epoch": 3.56971634039153, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8935 + }, + { + "epoch": 3.5701158609668395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8936 + }, + { + "epoch": 3.5705153815421493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 8937 + }, + { + "epoch": 3.570914902117459, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 8938 + }, + { + "epoch": 3.571314422692769, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8939 + }, + { + "epoch": 3.5717139432680782, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8940 + }, + { + "epoch": 3.572113463843388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 8941 + }, + { + "epoch": 3.5725129844186974, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 8942 + }, + { + "epoch": 3.572912504994007, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 8943 + }, + { + "epoch": 3.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 8944 + }, + { + "epoch": 3.5737115461446267, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8945 + }, + { + "epoch": 3.574111066719936, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 8946 + }, + { + "epoch": 3.574510587295246, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 8947 + }, + { + "epoch": 3.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 8948 + }, + { + "epoch": 3.575309628445865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 8949 + }, + { + "epoch": 3.5757091490211748, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 8950 + }, + { + "epoch": 3.576108669596484, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 8951 + }, + { + "epoch": 3.576508190171794, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8952 + }, + { + "epoch": 3.5769077107471032, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8953 + }, + { + "epoch": 3.577307231322413, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 8954 + }, + { + "epoch": 3.577706751897723, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8955 + }, + { + "epoch": 3.5781062724730326, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 8956 + }, + { + "epoch": 3.578505793048342, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 8957 + }, + { + "epoch": 3.5789053136236517, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8958 + }, + { + "epoch": 3.579304834198961, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 8959 + }, + { + "epoch": 3.579704354774271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 8960 + }, + { + "epoch": 3.5801038753495806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 8961 + }, + { + "epoch": 3.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 8962 + }, + { + "epoch": 3.5809029165001998, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 8963 + }, + { + "epoch": 3.5813024370755095, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 8964 + }, + { + "epoch": 3.581701957650819, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 8965 + }, + { + "epoch": 3.5821014782261287, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 8966 + }, + { + "epoch": 3.5825009988014385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 8967 + }, + { + "epoch": 3.582900519376748, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 8968 + }, + { + "epoch": 3.5833000399520576, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 8969 + }, + { + "epoch": 3.583699560527367, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 8970 + }, + { + "epoch": 3.5840990811026767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 8971 + }, + { + "epoch": 3.5844986016779865, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 8972 + }, + { + "epoch": 3.5848981222532963, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 8973 + }, + { + "epoch": 3.5852976428286056, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 8974 + }, + { + "epoch": 3.5856971634039154, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 8975 + }, + { + "epoch": 3.5860966839792248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 8976 + }, + { + "epoch": 3.5864962045545346, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 8977 + }, + { + "epoch": 3.5868957251298443, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 8978 + }, + { + "epoch": 3.5872952457051537, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 8979 + }, + { + "epoch": 3.5876947662804635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 8980 + }, + { + "epoch": 3.588094286855773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 8981 + }, + { + "epoch": 3.5884938074310826, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 8982 + }, + { + "epoch": 3.5888933280063924, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 8983 + }, + { + "epoch": 3.589292848581702, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8984 + }, + { + "epoch": 3.5896923691570115, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 8985 + }, + { + "epoch": 3.5900918897323213, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 8986 + }, + { + "epoch": 3.5904914103076306, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 8987 + }, + { + "epoch": 3.5908909308829404, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 8988 + }, + { + "epoch": 3.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 8989 + }, + { + "epoch": 3.59168997203356, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 8990 + }, + { + "epoch": 3.5920894926088693, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8991 + }, + { + "epoch": 3.592489013184179, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 8992 + }, + { + "epoch": 3.5928885337594885, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 8993 + }, + { + "epoch": 3.5932880543347983, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 8994 + }, + { + "epoch": 3.593687574910108, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 8995 + }, + { + "epoch": 3.5940870954854174, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 8996 + }, + { + "epoch": 3.594486616060727, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 8997 + }, + { + "epoch": 3.5948861366360365, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 8998 + }, + { + "epoch": 3.5952856572113463, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 8999 + }, + { + "epoch": 3.595685177786656, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9000 + }, + { + "epoch": 3.596084698361966, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9001 + }, + { + "epoch": 3.596484218937275, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9002 + }, + { + "epoch": 3.596883739512585, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 9003 + }, + { + "epoch": 3.5972832600878943, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9004 + }, + { + "epoch": 3.597682780663204, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9005 + }, + { + "epoch": 3.598082301238514, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9006 + }, + { + "epoch": 3.5984818218138233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9007 + }, + { + "epoch": 3.598881342389133, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9008 + }, + { + "epoch": 3.599280862964443, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 9009 + }, + { + "epoch": 3.599680383539752, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9010 + }, + { + "epoch": 3.600079904115062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9011 + }, + { + "epoch": 3.6004794246903717, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9012 + }, + { + "epoch": 3.600878945265681, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9013 + }, + { + "epoch": 3.601278465840991, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9014 + }, + { + "epoch": 3.6016779864163, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9015 + }, + { + "epoch": 3.60207750699161, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9016 + }, + { + "epoch": 3.60247702756692, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9017 + }, + { + "epoch": 3.6028765481422296, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9018 + }, + { + "epoch": 3.603276068717539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9019 + }, + { + "epoch": 3.6036755892928487, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9020 + }, + { + "epoch": 3.604075109868158, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9021 + }, + { + "epoch": 3.604474630443468, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9022 + }, + { + "epoch": 3.6048741510187776, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9023 + }, + { + "epoch": 3.605273671594087, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9024 + }, + { + "epoch": 3.6056731921693967, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9025 + }, + { + "epoch": 3.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9026 + }, + { + "epoch": 3.606472233320016, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9027 + }, + { + "epoch": 3.6068717538953257, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9028 + }, + { + "epoch": 3.6072712744706354, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9029 + }, + { + "epoch": 3.607670795045945, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9030 + }, + { + "epoch": 3.6080703156212546, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9031 + }, + { + "epoch": 3.608469836196564, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9032 + }, + { + "epoch": 3.6088693567718737, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9033 + }, + { + "epoch": 3.6092688773471835, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 9034 + }, + { + "epoch": 3.6096683979224933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9035 + }, + { + "epoch": 3.6100679184978026, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9036 + }, + { + "epoch": 3.6104674390731124, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9037 + }, + { + "epoch": 3.6108669596484217, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9038 + }, + { + "epoch": 3.6112664802237315, + "grad_norm": 0.18359375, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 9039 + }, + { + "epoch": 3.6116660007990413, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9040 + }, + { + "epoch": 3.6120655213743507, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9041 + }, + { + "epoch": 3.6124650419496604, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9042 + }, + { + "epoch": 3.61286456252497, + "grad_norm": 0.2275390625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9043 + }, + { + "epoch": 3.6132640831002796, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9044 + }, + { + "epoch": 3.6136636036755894, + "grad_norm": 0.2041015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9045 + }, + { + "epoch": 3.614063124250899, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9046 + }, + { + "epoch": 3.6144626448262085, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 9047 + }, + { + "epoch": 3.6148621654015183, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9048 + }, + { + "epoch": 3.6152616859768276, + "grad_norm": 0.259765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9049 + }, + { + "epoch": 3.6156612065521374, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9050 + }, + { + "epoch": 3.616060727127447, + "grad_norm": 0.3046875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9051 + }, + { + "epoch": 3.6164602477027565, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9052 + }, + { + "epoch": 3.6168597682780663, + "grad_norm": 0.302734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9053 + }, + { + "epoch": 3.617259288853376, + "grad_norm": 0.1806640625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9054 + }, + { + "epoch": 3.6176588094286855, + "grad_norm": 0.47265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9055 + }, + { + "epoch": 3.6180583300039952, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9056 + }, + { + "epoch": 3.618457850579305, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9057 + }, + { + "epoch": 3.6188573711546144, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9058 + }, + { + "epoch": 3.619256891729924, + "grad_norm": 0.1904296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9059 + }, + { + "epoch": 3.6196564123052335, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9060 + }, + { + "epoch": 3.6200559328805433, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9061 + }, + { + "epoch": 3.620455453455853, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9062 + }, + { + "epoch": 3.620854974031163, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3403, + "step": 9063 + }, + { + "epoch": 3.621254494606472, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9064 + }, + { + "epoch": 3.621654015181782, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9065 + }, + { + "epoch": 3.6220535357570913, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9066 + }, + { + "epoch": 3.622453056332401, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9067 + }, + { + "epoch": 3.622852576907711, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9068 + }, + { + "epoch": 3.6232520974830202, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9069 + }, + { + "epoch": 3.62365161805833, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9070 + }, + { + "epoch": 3.6240511386336394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9071 + }, + { + "epoch": 3.624450659208949, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9072 + }, + { + "epoch": 3.624850179784259, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9073 + }, + { + "epoch": 3.6252497003595687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9074 + }, + { + "epoch": 3.625649220934878, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 9075 + }, + { + "epoch": 3.626048741510188, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9076 + }, + { + "epoch": 3.626448262085497, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9077 + }, + { + "epoch": 3.626847782660807, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9078 + }, + { + "epoch": 3.6272473032361168, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9079 + }, + { + "epoch": 3.6276468238114266, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9080 + }, + { + "epoch": 3.628046344386736, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9081 + }, + { + "epoch": 3.6284458649620457, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9082 + }, + { + "epoch": 3.628845385537355, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9083 + }, + { + "epoch": 3.629244906112665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9084 + }, + { + "epoch": 3.6296444266879746, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9085 + }, + { + "epoch": 3.630043947263284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9086 + }, + { + "epoch": 3.6304434678385937, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9087 + }, + { + "epoch": 3.630842988413903, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9088 + }, + { + "epoch": 3.631242508989213, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9089 + }, + { + "epoch": 3.6316420295645226, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9090 + }, + { + "epoch": 3.6320415501398324, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9091 + }, + { + "epoch": 3.6324410707151418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3456, + "step": 9092 + }, + { + "epoch": 3.6328405912904516, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9093 + }, + { + "epoch": 3.633240111865761, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9094 + }, + { + "epoch": 3.6336396324410707, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 9095 + }, + { + "epoch": 3.6340391530163805, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9096 + }, + { + "epoch": 3.63443867359169, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9097 + }, + { + "epoch": 3.6348381941669996, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9098 + }, + { + "epoch": 3.6352377147423094, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9099 + }, + { + "epoch": 3.6356372353176187, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9100 + }, + { + "epoch": 3.6360367558929285, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9101 + }, + { + "epoch": 3.6364362764682383, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9102 + }, + { + "epoch": 3.6368357970435476, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9103 + }, + { + "epoch": 3.6372353176188574, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9104 + }, + { + "epoch": 3.6376348381941668, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9105 + }, + { + "epoch": 3.6380343587694766, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9106 + }, + { + "epoch": 3.6384338793447863, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9107 + }, + { + "epoch": 3.638833399920096, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9108 + }, + { + "epoch": 3.6392329204954055, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9109 + }, + { + "epoch": 3.6396324410707153, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9110 + }, + { + "epoch": 3.6400319616460246, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9111 + }, + { + "epoch": 3.6404314822213344, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9112 + }, + { + "epoch": 3.640831002796644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9113 + }, + { + "epoch": 3.6412305233719535, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9114 + }, + { + "epoch": 3.6416300439472633, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9115 + }, + { + "epoch": 3.6420295645225726, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 9116 + }, + { + "epoch": 3.6424290850978824, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9117 + }, + { + "epoch": 3.642828605673192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9118 + }, + { + "epoch": 3.643228126248502, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9119 + }, + { + "epoch": 3.6436276468238114, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9120 + }, + { + "epoch": 3.644027167399121, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9121 + }, + { + "epoch": 3.6444266879744305, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9122 + }, + { + "epoch": 3.6448262085497403, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9123 + }, + { + "epoch": 3.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9124 + }, + { + "epoch": 3.64562524970036, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9125 + }, + { + "epoch": 3.646024770275669, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9126 + }, + { + "epoch": 3.646424290850979, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9127 + }, + { + "epoch": 3.6468238114262883, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9128 + }, + { + "epoch": 3.647223332001598, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 9129 + }, + { + "epoch": 3.647622852576908, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9130 + }, + { + "epoch": 3.6480223731522172, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9131 + }, + { + "epoch": 3.648421893727527, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9132 + }, + { + "epoch": 3.6488214143028364, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 9133 + }, + { + "epoch": 3.649220934878146, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9134 + }, + { + "epoch": 3.649620455453456, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9135 + }, + { + "epoch": 3.6500199760287657, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9136 + }, + { + "epoch": 3.650419496604075, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9137 + }, + { + "epoch": 3.650819017179385, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9138 + }, + { + "epoch": 3.651218537754694, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 9139 + }, + { + "epoch": 3.651618058330004, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 9140 + }, + { + "epoch": 3.6520175789053138, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9141 + }, + { + "epoch": 3.652417099480623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9142 + }, + { + "epoch": 3.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9143 + }, + { + "epoch": 3.6532161406312427, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9144 + }, + { + "epoch": 3.653615661206552, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3405, + "step": 9145 + }, + { + "epoch": 3.654015181781862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9146 + }, + { + "epoch": 3.6544147023571716, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9147 + }, + { + "epoch": 3.654814222932481, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9148 + }, + { + "epoch": 3.6552137435077907, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9149 + }, + { + "epoch": 3.6556132640831, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9150 + }, + { + "epoch": 3.65601278465841, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9151 + }, + { + "epoch": 3.6564123052337196, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9152 + }, + { + "epoch": 3.6568118258090294, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9153 + }, + { + "epoch": 3.6572113463843388, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9154 + }, + { + "epoch": 3.6576108669596485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 9155 + }, + { + "epoch": 3.658010387534958, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9156 + }, + { + "epoch": 3.6584099081102677, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9157 + }, + { + "epoch": 3.6588094286855775, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9158 + }, + { + "epoch": 3.659208949260887, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9159 + }, + { + "epoch": 3.6596084698361966, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9160 + }, + { + "epoch": 3.660007990411506, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9161 + }, + { + "epoch": 3.6604075109868157, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9162 + }, + { + "epoch": 3.6608070315621255, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9163 + }, + { + "epoch": 3.6612065521374353, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9164 + }, + { + "epoch": 3.6616060727127446, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9165 + }, + { + "epoch": 3.6620055932880544, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9166 + }, + { + "epoch": 3.6624051138633638, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9167 + }, + { + "epoch": 3.6628046344386735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9168 + }, + { + "epoch": 3.6632041550139833, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9169 + }, + { + "epoch": 3.663603675589293, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9170 + }, + { + "epoch": 3.6640031961646025, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9171 + }, + { + "epoch": 3.6644027167399122, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9172 + }, + { + "epoch": 3.6648022373152216, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9173 + }, + { + "epoch": 3.6652017578905314, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9174 + }, + { + "epoch": 3.665601278465841, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9175 + }, + { + "epoch": 3.6660007990411505, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9176 + }, + { + "epoch": 3.6664003196164603, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9177 + }, + { + "epoch": 3.6667998401917696, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9178 + }, + { + "epoch": 3.6671993607670794, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9179 + }, + { + "epoch": 3.667598881342389, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9180 + }, + { + "epoch": 3.667998401917699, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9181 + }, + { + "epoch": 3.6683979224930083, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9182 + }, + { + "epoch": 3.668797443068318, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 9183 + }, + { + "epoch": 3.6691969636436275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9184 + }, + { + "epoch": 3.6695964842189372, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9185 + }, + { + "epoch": 3.669996004794247, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9186 + }, + { + "epoch": 3.6703955253695564, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9187 + }, + { + "epoch": 3.670795045944866, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9188 + }, + { + "epoch": 3.671194566520176, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9189 + }, + { + "epoch": 3.6715940870954853, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9190 + }, + { + "epoch": 3.671993607670795, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9191 + }, + { + "epoch": 3.672393128246105, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9192 + }, + { + "epoch": 3.672792648821414, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9193 + }, + { + "epoch": 3.673192169396724, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9194 + }, + { + "epoch": 3.6735916899720333, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9195 + }, + { + "epoch": 3.673991210547343, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9196 + }, + { + "epoch": 3.674390731122653, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9197 + }, + { + "epoch": 3.6747902516979627, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9198 + }, + { + "epoch": 3.675189772273272, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9199 + }, + { + "epoch": 3.675589292848582, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9200 + }, + { + "epoch": 3.675988813423891, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 9201 + }, + { + "epoch": 3.676388333999201, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9202 + }, + { + "epoch": 3.6767878545745107, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9203 + }, + { + "epoch": 3.67718737514982, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9204 + }, + { + "epoch": 3.67758689572513, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9205 + }, + { + "epoch": 3.677986416300439, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9206 + }, + { + "epoch": 3.678385936875749, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9207 + }, + { + "epoch": 3.678785457451059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9208 + }, + { + "epoch": 3.6791849780263686, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9209 + }, + { + "epoch": 3.679584498601678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9210 + }, + { + "epoch": 3.6799840191769877, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9211 + }, + { + "epoch": 3.680383539752297, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9212 + }, + { + "epoch": 3.680783060327607, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9213 + }, + { + "epoch": 3.6811825809029166, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9214 + }, + { + "epoch": 3.6815821014782264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9215 + }, + { + "epoch": 3.6819816220535357, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9216 + }, + { + "epoch": 3.6823811426288455, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9217 + }, + { + "epoch": 3.682780663204155, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9218 + }, + { + "epoch": 3.6831801837794647, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9219 + }, + { + "epoch": 3.6835797043547744, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9220 + }, + { + "epoch": 3.683979224930084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9221 + }, + { + "epoch": 3.6843787455053936, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9222 + }, + { + "epoch": 3.684778266080703, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9223 + }, + { + "epoch": 3.6851777866560127, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9224 + }, + { + "epoch": 3.6855773072313225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9225 + }, + { + "epoch": 3.6859768278066323, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9226 + }, + { + "epoch": 3.6863763483819416, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9227 + }, + { + "epoch": 3.6867758689572514, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9228 + }, + { + "epoch": 3.6871753895325607, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9229 + }, + { + "epoch": 3.6875749101078705, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9230 + }, + { + "epoch": 3.6879744306831803, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9231 + }, + { + "epoch": 3.6883739512584897, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 9232 + }, + { + "epoch": 3.6887734718337994, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 9233 + }, + { + "epoch": 3.6891729924091092, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9234 + }, + { + "epoch": 3.6895725129844186, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9235 + }, + { + "epoch": 3.6899720335597284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9236 + }, + { + "epoch": 3.690371554135038, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9237 + }, + { + "epoch": 3.6907710747103475, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9238 + }, + { + "epoch": 3.6911705952856573, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9239 + }, + { + "epoch": 3.6915701158609666, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9240 + }, + { + "epoch": 3.6919696364362764, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9241 + }, + { + "epoch": 3.692369157011586, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9242 + }, + { + "epoch": 3.692768677586896, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9243 + }, + { + "epoch": 3.6931681981622053, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9244 + }, + { + "epoch": 3.693567718737515, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9245 + }, + { + "epoch": 3.6939672393128244, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9246 + }, + { + "epoch": 3.6943667598881342, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9247 + }, + { + "epoch": 3.694766280463444, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9248 + }, + { + "epoch": 3.6951658010387534, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 9249 + }, + { + "epoch": 3.695565321614063, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9250 + }, + { + "epoch": 3.6959648421893725, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3358, + "step": 9251 + }, + { + "epoch": 3.6963643627646823, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 9252 + }, + { + "epoch": 3.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9253 + }, + { + "epoch": 3.697163403915302, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9254 + }, + { + "epoch": 3.697562924490611, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9255 + }, + { + "epoch": 3.697962445065921, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9256 + }, + { + "epoch": 3.6983619656412303, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9257 + }, + { + "epoch": 3.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9258 + }, + { + "epoch": 3.69916100679185, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9259 + }, + { + "epoch": 3.6995605273671597, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9260 + }, + { + "epoch": 3.699960047942469, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9261 + }, + { + "epoch": 3.700359568517779, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 9262 + }, + { + "epoch": 3.700759089093088, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9263 + }, + { + "epoch": 3.701158609668398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9264 + }, + { + "epoch": 3.7015581302437077, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9265 + }, + { + "epoch": 3.701957650819017, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9266 + }, + { + "epoch": 3.702357171394327, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9267 + }, + { + "epoch": 3.702756691969636, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9268 + }, + { + "epoch": 3.703156212544946, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9269 + }, + { + "epoch": 3.7035557331202558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9270 + }, + { + "epoch": 3.7039552536955656, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9271 + }, + { + "epoch": 3.704354774270875, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9272 + }, + { + "epoch": 3.7047542948461847, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9273 + }, + { + "epoch": 3.705153815421494, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9274 + }, + { + "epoch": 3.705553335996804, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9275 + }, + { + "epoch": 3.7059528565721136, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9276 + }, + { + "epoch": 3.706352377147423, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9277 + }, + { + "epoch": 3.7067518977227327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9278 + }, + { + "epoch": 3.7071514182980425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9279 + }, + { + "epoch": 3.707550938873352, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9280 + }, + { + "epoch": 3.7079504594486616, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9281 + }, + { + "epoch": 3.7083499800239714, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9282 + }, + { + "epoch": 3.7087495005992808, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9283 + }, + { + "epoch": 3.7091490211745906, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9284 + }, + { + "epoch": 3.7095485417499, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9285 + }, + { + "epoch": 3.7099480623252097, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9286 + }, + { + "epoch": 3.7103475829005195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9287 + }, + { + "epoch": 3.7107471034758293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9288 + }, + { + "epoch": 3.7111466240511386, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9289 + }, + { + "epoch": 3.7115461446264484, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9290 + }, + { + "epoch": 3.7119456652017577, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9291 + }, + { + "epoch": 3.7123451857770675, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9292 + }, + { + "epoch": 3.7127447063523773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9293 + }, + { + "epoch": 3.7131442269276866, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9294 + }, + { + "epoch": 3.7135437475029964, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9295 + }, + { + "epoch": 3.7139432680783058, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 9296 + }, + { + "epoch": 3.7143427886536156, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9297 + }, + { + "epoch": 3.7147423092289253, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9298 + }, + { + "epoch": 3.715141829804235, + "grad_norm": 0.1376953125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9299 + }, + { + "epoch": 3.7155413503795445, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9300 + }, + { + "epoch": 3.7159408709548543, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.339, + "step": 9301 + }, + { + "epoch": 3.7163403915301636, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9302 + }, + { + "epoch": 3.7167399121054734, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3418, + "step": 9303 + }, + { + "epoch": 3.717139432680783, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9304 + }, + { + "epoch": 3.717538953256093, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9305 + }, + { + "epoch": 3.7179384738314023, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9306 + }, + { + "epoch": 3.718337994406712, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3378, + "step": 9307 + }, + { + "epoch": 3.7187375149820214, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9308 + }, + { + "epoch": 3.719137035557331, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9309 + }, + { + "epoch": 3.719536556132641, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9310 + }, + { + "epoch": 3.7199360767079503, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9311 + }, + { + "epoch": 3.72033559728326, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 9312 + }, + { + "epoch": 3.7207351178585695, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9313 + }, + { + "epoch": 3.7211346384338793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9314 + }, + { + "epoch": 3.721534159009189, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9315 + }, + { + "epoch": 3.721933679584499, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9316 + }, + { + "epoch": 3.722333200159808, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9317 + }, + { + "epoch": 3.722732720735118, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 9318 + }, + { + "epoch": 3.7231322413104273, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9319 + }, + { + "epoch": 3.723531761885737, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 9320 + }, + { + "epoch": 3.723931282461047, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9321 + }, + { + "epoch": 3.724330803036356, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9322 + }, + { + "epoch": 3.724730323611666, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 9323 + }, + { + "epoch": 3.725129844186976, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9324 + }, + { + "epoch": 3.725529364762285, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 9325 + }, + { + "epoch": 3.725928885337595, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9326 + }, + { + "epoch": 3.7263284059129047, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9327 + }, + { + "epoch": 3.726727926488214, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9328 + }, + { + "epoch": 3.727127447063524, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9329 + }, + { + "epoch": 3.727526967638833, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 9330 + }, + { + "epoch": 3.727926488214143, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9331 + }, + { + "epoch": 3.7283260087894528, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9332 + }, + { + "epoch": 3.7287255293647625, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 9333 + }, + { + "epoch": 3.729125049940072, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 9334 + }, + { + "epoch": 3.7295245705153817, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 9335 + }, + { + "epoch": 3.729924091090691, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9336 + }, + { + "epoch": 3.730323611666001, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9337 + }, + { + "epoch": 3.7307231322413106, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9338 + }, + { + "epoch": 3.73112265281662, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9339 + }, + { + "epoch": 3.7315221733919297, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9340 + }, + { + "epoch": 3.731921693967239, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9341 + }, + { + "epoch": 3.732321214542549, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9342 + }, + { + "epoch": 3.7327207351178586, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9343 + }, + { + "epoch": 3.7331202556931684, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9344 + }, + { + "epoch": 3.7335197762684778, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9345 + }, + { + "epoch": 3.7339192968437875, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9346 + }, + { + "epoch": 3.734318817419097, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9347 + }, + { + "epoch": 3.7347183379944067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9348 + }, + { + "epoch": 3.7351178585697165, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9349 + }, + { + "epoch": 3.7355173791450262, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 9350 + }, + { + "epoch": 3.7359168997203356, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9351 + }, + { + "epoch": 3.7363164202956454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9352 + }, + { + "epoch": 3.7367159408709547, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9353 + }, + { + "epoch": 3.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 9354 + }, + { + "epoch": 3.7375149820215743, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9355 + }, + { + "epoch": 3.7379145025968836, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9356 + }, + { + "epoch": 3.7383140231721934, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9357 + }, + { + "epoch": 3.7387135437475028, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 9358 + }, + { + "epoch": 3.7391130643228125, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 9359 + }, + { + "epoch": 3.7395125848981223, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9360 + }, + { + "epoch": 3.739912105473432, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9361 + }, + { + "epoch": 3.7403116260487415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9362 + }, + { + "epoch": 3.7407111466240512, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 9363 + }, + { + "epoch": 3.7411106671993606, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 9364 + }, + { + "epoch": 3.7415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9365 + }, + { + "epoch": 3.74190970834998, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9366 + }, + { + "epoch": 3.7423092289252895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9367 + }, + { + "epoch": 3.7427087495005993, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9368 + }, + { + "epoch": 3.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 9369 + }, + { + "epoch": 3.7435077906512184, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 9370 + }, + { + "epoch": 3.743907311226528, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9371 + }, + { + "epoch": 3.744306831801838, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9372 + }, + { + "epoch": 3.7447063523771473, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 9373 + }, + { + "epoch": 3.745105872952457, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9374 + }, + { + "epoch": 3.7455053935277665, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9375 + }, + { + "epoch": 3.7459049141030762, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9376 + }, + { + "epoch": 3.746304434678386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9377 + }, + { + "epoch": 3.746703955253696, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9378 + }, + { + "epoch": 3.747103475829005, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9379 + }, + { + "epoch": 3.747502996404315, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9380 + }, + { + "epoch": 3.7479025169796243, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9381 + }, + { + "epoch": 3.748302037554934, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9382 + }, + { + "epoch": 3.748701558130244, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9383 + }, + { + "epoch": 3.749101078705553, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9384 + }, + { + "epoch": 3.749500599280863, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9385 + }, + { + "epoch": 3.7499001198561723, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9386 + }, + { + "epoch": 3.750299640431482, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9387 + }, + { + "epoch": 3.750699161006792, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9388 + }, + { + "epoch": 3.7510986815821017, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9389 + }, + { + "epoch": 3.751498202157411, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9390 + }, + { + "epoch": 3.751897722732721, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9391 + }, + { + "epoch": 3.75229724330803, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9392 + }, + { + "epoch": 3.75269676388334, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9393 + }, + { + "epoch": 3.7530962844586497, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9394 + }, + { + "epoch": 3.7534958050339595, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9395 + }, + { + "epoch": 3.753895325609269, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 9396 + }, + { + "epoch": 3.7542948461845786, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9397 + }, + { + "epoch": 3.754694366759888, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9398 + }, + { + "epoch": 3.7550938873351978, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9399 + }, + { + "epoch": 3.7554934079105076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9400 + }, + { + "epoch": 3.755892928485817, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9401 + }, + { + "epoch": 3.7562924490611267, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9402 + }, + { + "epoch": 3.756691969636436, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9403 + }, + { + "epoch": 3.757091490211746, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9404 + }, + { + "epoch": 3.7574910107870556, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9405 + }, + { + "epoch": 3.7578905313623654, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9406 + }, + { + "epoch": 3.7582900519376747, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9407 + }, + { + "epoch": 3.7586895725129845, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 9408 + }, + { + "epoch": 3.759089093088294, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9409 + }, + { + "epoch": 3.7594886136636037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9410 + }, + { + "epoch": 3.7598881342389134, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9411 + }, + { + "epoch": 3.760287654814223, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9412 + }, + { + "epoch": 3.7606871753895326, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9413 + }, + { + "epoch": 3.7610866959648424, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9414 + }, + { + "epoch": 3.7614862165401517, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9415 + }, + { + "epoch": 3.7618857371154615, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9416 + }, + { + "epoch": 3.7622852576907713, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9417 + }, + { + "epoch": 3.7626847782660806, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9418 + }, + { + "epoch": 3.7630842988413904, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 9419 + }, + { + "epoch": 3.7634838194166997, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9420 + }, + { + "epoch": 3.7638833399920095, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9421 + }, + { + "epoch": 3.7642828605673193, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9422 + }, + { + "epoch": 3.764682381142629, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9423 + }, + { + "epoch": 3.7650819017179384, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9424 + }, + { + "epoch": 3.7654814222932482, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9425 + }, + { + "epoch": 3.7658809428685576, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9426 + }, + { + "epoch": 3.7662804634438674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9427 + }, + { + "epoch": 3.766679984019177, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9428 + }, + { + "epoch": 3.7670795045944865, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9429 + }, + { + "epoch": 3.7674790251697963, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9430 + }, + { + "epoch": 3.7678785457451056, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9431 + }, + { + "epoch": 3.7682780663204154, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9432 + }, + { + "epoch": 3.768677586895725, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9433 + }, + { + "epoch": 3.769077107471035, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9434 + }, + { + "epoch": 3.7694766280463443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9435 + }, + { + "epoch": 3.769876148621654, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 9436 + }, + { + "epoch": 3.7702756691969634, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9437 + }, + { + "epoch": 3.7706751897722732, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9438 + }, + { + "epoch": 3.771074710347583, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9439 + }, + { + "epoch": 3.771474230922893, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 9440 + }, + { + "epoch": 3.771873751498202, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9441 + }, + { + "epoch": 3.772273272073512, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 9442 + }, + { + "epoch": 3.7726727926488213, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 9443 + }, + { + "epoch": 3.773072313224131, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9444 + }, + { + "epoch": 3.773471833799441, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9445 + }, + { + "epoch": 3.77387135437475, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9446 + }, + { + "epoch": 3.77427087495006, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 9447 + }, + { + "epoch": 3.7746703955253693, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9448 + }, + { + "epoch": 3.775069916100679, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 9449 + }, + { + "epoch": 3.775469436675989, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9450 + }, + { + "epoch": 3.7758689572512987, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9451 + }, + { + "epoch": 3.776268477826608, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9452 + }, + { + "epoch": 3.776667998401918, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9453 + }, + { + "epoch": 3.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 9454 + }, + { + "epoch": 3.777467039552537, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9455 + }, + { + "epoch": 3.7778665601278467, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9456 + }, + { + "epoch": 3.778266080703156, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9457 + }, + { + "epoch": 3.778665601278466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9458 + }, + { + "epoch": 3.7790651218537756, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9459 + }, + { + "epoch": 3.779464642429085, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9460 + }, + { + "epoch": 3.7798641630043948, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9461 + }, + { + "epoch": 3.7802636835797045, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9462 + }, + { + "epoch": 3.780663204155014, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9463 + }, + { + "epoch": 3.7810627247303237, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9464 + }, + { + "epoch": 3.781462245305633, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 9465 + }, + { + "epoch": 3.781861765880943, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9466 + }, + { + "epoch": 3.7822612864562526, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9467 + }, + { + "epoch": 3.7826608070315624, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9468 + }, + { + "epoch": 3.7830603276068717, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9469 + }, + { + "epoch": 3.7834598481821815, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9470 + }, + { + "epoch": 3.783859368757491, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 9471 + }, + { + "epoch": 3.7842588893328006, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9472 + }, + { + "epoch": 3.7846584099081104, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9473 + }, + { + "epoch": 3.7850579304834198, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9474 + }, + { + "epoch": 3.7854574510587295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9475 + }, + { + "epoch": 3.785856971634039, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9476 + }, + { + "epoch": 3.7862564922093487, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9477 + }, + { + "epoch": 3.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9478 + }, + { + "epoch": 3.7870555333599683, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 9479 + }, + { + "epoch": 3.7874550539352776, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9480 + }, + { + "epoch": 3.7878545745105874, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9481 + }, + { + "epoch": 3.7882540950858967, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 9482 + }, + { + "epoch": 3.7886536156612065, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9483 + }, + { + "epoch": 3.7890531362365163, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9484 + }, + { + "epoch": 3.789452656811826, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9485 + }, + { + "epoch": 3.7898521773871354, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3384, + "step": 9486 + }, + { + "epoch": 3.790251697962445, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9487 + }, + { + "epoch": 3.7906512185377546, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9488 + }, + { + "epoch": 3.7910507391130643, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 9489 + }, + { + "epoch": 3.791450259688374, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 9490 + }, + { + "epoch": 3.7918497802636835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9491 + }, + { + "epoch": 3.7922493008389933, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9492 + }, + { + "epoch": 3.7926488214143026, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9493 + }, + { + "epoch": 3.7930483419896124, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9494 + }, + { + "epoch": 3.793447862564922, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9495 + }, + { + "epoch": 3.793847383140232, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9496 + }, + { + "epoch": 3.7942469037155413, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9497 + }, + { + "epoch": 3.794646424290851, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9498 + }, + { + "epoch": 3.7950459448661604, + "grad_norm": 0.1455078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9499 + }, + { + "epoch": 3.79544546544147, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9500 + }, + { + "epoch": 3.79584498601678, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9501 + }, + { + "epoch": 3.7962445065920893, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9502 + }, + { + "epoch": 3.796644027167399, + "grad_norm": 0.12060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9503 + }, + { + "epoch": 3.797043547742709, + "grad_norm": 0.17578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9504 + }, + { + "epoch": 3.7974430683180183, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 9505 + }, + { + "epoch": 3.797842588893328, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9506 + }, + { + "epoch": 3.798242109468638, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9507 + }, + { + "epoch": 3.798641630043947, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9508 + }, + { + "epoch": 3.799041150619257, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 9509 + }, + { + "epoch": 3.7994406711945663, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 9510 + }, + { + "epoch": 3.799840191769876, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9511 + }, + { + "epoch": 3.800239712345186, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9512 + }, + { + "epoch": 3.8006392329204957, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9513 + }, + { + "epoch": 3.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9514 + }, + { + "epoch": 3.801438274071115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9515 + }, + { + "epoch": 3.801837794646424, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9516 + }, + { + "epoch": 3.802237315221734, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9517 + }, + { + "epoch": 3.8026368357970437, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9518 + }, + { + "epoch": 3.803036356372353, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9519 + }, + { + "epoch": 3.803435876947663, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9520 + }, + { + "epoch": 3.803835397522972, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9521 + }, + { + "epoch": 3.804234918098282, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9522 + }, + { + "epoch": 3.8046344386735917, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9523 + }, + { + "epoch": 3.8050339592489015, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 9524 + }, + { + "epoch": 3.805433479824211, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9525 + }, + { + "epoch": 3.8058330003995207, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 9526 + }, + { + "epoch": 3.80623252097483, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9527 + }, + { + "epoch": 3.80663204155014, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9528 + }, + { + "epoch": 3.8070315621254496, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9529 + }, + { + "epoch": 3.8074310827007594, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9530 + }, + { + "epoch": 3.8078306032760687, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9531 + }, + { + "epoch": 3.8082301238513785, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9532 + }, + { + "epoch": 3.808629644426688, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9533 + }, + { + "epoch": 3.8090291650019976, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9534 + }, + { + "epoch": 3.8094286855773074, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 9535 + }, + { + "epoch": 3.8098282061526167, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9536 + }, + { + "epoch": 3.8102277267279265, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9537 + }, + { + "epoch": 3.810627247303236, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 9538 + }, + { + "epoch": 3.8110267678785457, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9539 + }, + { + "epoch": 3.8114262884538554, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 9540 + }, + { + "epoch": 3.8118258090291652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9541 + }, + { + "epoch": 3.8122253296044746, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9542 + }, + { + "epoch": 3.8126248501797844, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 9543 + }, + { + "epoch": 3.8130243707550937, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9544 + }, + { + "epoch": 3.8134238913304035, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 9545 + }, + { + "epoch": 3.8138234119057133, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9546 + }, + { + "epoch": 3.8142229324810226, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9547 + }, + { + "epoch": 3.8146224530563324, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9548 + }, + { + "epoch": 3.815021973631642, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9549 + }, + { + "epoch": 3.8154214942069515, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9550 + }, + { + "epoch": 3.8158210147822613, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 9551 + }, + { + "epoch": 3.816220535357571, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9552 + }, + { + "epoch": 3.8166200559328805, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9553 + }, + { + "epoch": 3.8170195765081902, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 9554 + }, + { + "epoch": 3.8174190970834996, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9555 + }, + { + "epoch": 3.8178186176588094, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9556 + }, + { + "epoch": 3.818218138234119, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 9557 + }, + { + "epoch": 3.818617658809429, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9558 + }, + { + "epoch": 3.8190171793847383, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 9559 + }, + { + "epoch": 3.819416699960048, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9560 + }, + { + "epoch": 3.8198162205353574, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9561 + }, + { + "epoch": 3.820215741110667, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9562 + }, + { + "epoch": 3.820615261685977, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9563 + }, + { + "epoch": 3.8210147822612863, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9564 + }, + { + "epoch": 3.821414302836596, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9565 + }, + { + "epoch": 3.8218138234119055, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9566 + }, + { + "epoch": 3.8222133439872152, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9567 + }, + { + "epoch": 3.822612864562525, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9568 + }, + { + "epoch": 3.823012385137835, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 9569 + }, + { + "epoch": 3.823411905713144, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 9570 + }, + { + "epoch": 3.823811426288454, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9571 + }, + { + "epoch": 3.8242109468637633, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9572 + }, + { + "epoch": 3.824610467439073, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9573 + }, + { + "epoch": 3.825009988014383, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9574 + }, + { + "epoch": 3.8254095085896926, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9575 + }, + { + "epoch": 3.825809029165002, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 9576 + }, + { + "epoch": 3.8262085497403118, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 9577 + }, + { + "epoch": 3.826608070315621, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9578 + }, + { + "epoch": 3.827007590890931, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9579 + }, + { + "epoch": 3.8274071114662407, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9580 + }, + { + "epoch": 3.82780663204155, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9581 + }, + { + "epoch": 3.82820615261686, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9582 + }, + { + "epoch": 3.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9583 + }, + { + "epoch": 3.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 9584 + }, + { + "epoch": 3.8294047143427887, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9585 + }, + { + "epoch": 3.8298042349180985, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9586 + }, + { + "epoch": 3.830203755493408, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9587 + }, + { + "epoch": 3.8306032760687176, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9588 + }, + { + "epoch": 3.831002796644027, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9589 + }, + { + "epoch": 3.8314023172193368, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9590 + }, + { + "epoch": 3.8318018377946466, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9591 + }, + { + "epoch": 3.832201358369956, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9592 + }, + { + "epoch": 3.8326008789452657, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9593 + }, + { + "epoch": 3.8330003995205755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 9594 + }, + { + "epoch": 3.833399920095885, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9595 + }, + { + "epoch": 3.8337994406711946, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9596 + }, + { + "epoch": 3.8341989612465044, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 9597 + }, + { + "epoch": 3.8345984818218137, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 9598 + }, + { + "epoch": 3.8349980023971235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9599 + }, + { + "epoch": 3.835397522972433, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9600 + }, + { + "epoch": 3.8357970435477426, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 9601 + }, + { + "epoch": 3.8361965641230524, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9602 + }, + { + "epoch": 3.836596084698362, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9603 + }, + { + "epoch": 3.8369956052736716, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9604 + }, + { + "epoch": 3.8373951258489813, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9605 + }, + { + "epoch": 3.8377946464242907, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9606 + }, + { + "epoch": 3.8381941669996005, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9607 + }, + { + "epoch": 3.8385936875749103, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9608 + }, + { + "epoch": 3.8389932081502196, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9609 + }, + { + "epoch": 3.8393927287255294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9610 + }, + { + "epoch": 3.8397922493008387, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 9611 + }, + { + "epoch": 3.8401917698761485, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9612 + }, + { + "epoch": 3.8405912904514583, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9613 + }, + { + "epoch": 3.840990811026768, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 9614 + }, + { + "epoch": 3.8413903316020774, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9615 + }, + { + "epoch": 3.841789852177387, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9616 + }, + { + "epoch": 3.8421893727526966, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9617 + }, + { + "epoch": 3.8425888933280063, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 9618 + }, + { + "epoch": 3.842988413903316, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 9619 + }, + { + "epoch": 3.843387934478626, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 9620 + }, + { + "epoch": 3.8437874550539353, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 9621 + }, + { + "epoch": 3.844186975629245, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 9622 + }, + { + "epoch": 3.8445864962045544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9623 + }, + { + "epoch": 3.844986016779864, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 9624 + }, + { + "epoch": 3.845385537355174, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9625 + }, + { + "epoch": 3.8457850579304833, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 9626 + }, + { + "epoch": 3.846184578505793, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9627 + }, + { + "epoch": 3.8465840990811024, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9628 + }, + { + "epoch": 3.8469836196564122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9629 + }, + { + "epoch": 3.847383140231722, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9630 + }, + { + "epoch": 3.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9631 + }, + { + "epoch": 3.848182181382341, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9632 + }, + { + "epoch": 3.848581701957651, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9633 + }, + { + "epoch": 3.8489812225329603, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 9634 + }, + { + "epoch": 3.84938074310827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9635 + }, + { + "epoch": 3.84978026368358, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 9636 + }, + { + "epoch": 3.850179784258889, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9637 + }, + { + "epoch": 3.850579304834199, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 9638 + }, + { + "epoch": 3.8509788254095088, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9639 + }, + { + "epoch": 3.851378345984818, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9640 + }, + { + "epoch": 3.851777866560128, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9641 + }, + { + "epoch": 3.8521773871354377, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 9642 + }, + { + "epoch": 3.852576907710747, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9643 + }, + { + "epoch": 3.852976428286057, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9644 + }, + { + "epoch": 3.853375948861366, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 9645 + }, + { + "epoch": 3.853775469436676, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 9646 + }, + { + "epoch": 3.8541749900119857, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 9647 + }, + { + "epoch": 3.8545745105872955, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9648 + }, + { + "epoch": 3.854974031162605, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 9649 + }, + { + "epoch": 3.8553735517379146, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9650 + }, + { + "epoch": 3.855773072313224, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9651 + }, + { + "epoch": 3.8561725928885338, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9652 + }, + { + "epoch": 3.8565721134638435, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9653 + }, + { + "epoch": 3.856971634039153, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 9654 + }, + { + "epoch": 3.8573711546144627, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9655 + }, + { + "epoch": 3.857770675189772, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 9656 + }, + { + "epoch": 3.858170195765082, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9657 + }, + { + "epoch": 3.8585697163403916, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 9658 + }, + { + "epoch": 3.8589692369157014, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 9659 + }, + { + "epoch": 3.8593687574910107, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9660 + }, + { + "epoch": 3.8597682780663205, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9661 + }, + { + "epoch": 3.86016779864163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 9662 + }, + { + "epoch": 3.8605673192169396, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9663 + }, + { + "epoch": 3.8609668397922494, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9664 + }, + { + "epoch": 3.861366360367559, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9665 + }, + { + "epoch": 3.8617658809428685, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9666 + }, + { + "epoch": 3.8621654015181783, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9667 + }, + { + "epoch": 3.8625649220934877, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9668 + }, + { + "epoch": 3.8629644426687975, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9669 + }, + { + "epoch": 3.8633639632441072, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9670 + }, + { + "epoch": 3.8637634838194166, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9671 + }, + { + "epoch": 3.8641630043947264, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9672 + }, + { + "epoch": 3.8645625249700357, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 9673 + }, + { + "epoch": 3.8649620455453455, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9674 + }, + { + "epoch": 3.8653615661206553, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 9675 + }, + { + "epoch": 3.865761086695965, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 9676 + }, + { + "epoch": 3.8661606072712744, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 9677 + }, + { + "epoch": 3.866560127846584, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9678 + }, + { + "epoch": 3.8669596484218935, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 9679 + }, + { + "epoch": 3.8673591689972033, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9680 + }, + { + "epoch": 3.867758689572513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 9681 + }, + { + "epoch": 3.8681582101478225, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 9682 + }, + { + "epoch": 3.8685577307231322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9683 + }, + { + "epoch": 3.868957251298442, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9684 + }, + { + "epoch": 3.8693567718737514, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9685 + }, + { + "epoch": 3.869756292449061, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9686 + }, + { + "epoch": 3.870155813024371, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9687 + }, + { + "epoch": 3.8705553335996803, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 9688 + }, + { + "epoch": 3.87095485417499, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 9689 + }, + { + "epoch": 3.8713543747502994, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9690 + }, + { + "epoch": 3.871753895325609, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9691 + }, + { + "epoch": 3.872153415900919, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9692 + }, + { + "epoch": 3.872552936476229, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9693 + }, + { + "epoch": 3.872952457051538, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9694 + }, + { + "epoch": 3.873351977626848, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9695 + }, + { + "epoch": 3.8737514982021573, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9696 + }, + { + "epoch": 3.874151018777467, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9697 + }, + { + "epoch": 3.874550539352777, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 9698 + }, + { + "epoch": 3.874950059928086, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9699 + }, + { + "epoch": 3.875349580503396, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9700 + }, + { + "epoch": 3.8757491010787053, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9701 + }, + { + "epoch": 3.876148621654015, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9702 + }, + { + "epoch": 3.876548142229325, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9703 + }, + { + "epoch": 3.8769476628046347, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9704 + }, + { + "epoch": 3.877347183379944, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9705 + }, + { + "epoch": 3.877746703955254, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9706 + }, + { + "epoch": 3.878146224530563, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9707 + }, + { + "epoch": 3.878545745105873, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 9708 + }, + { + "epoch": 3.8789452656811827, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9709 + }, + { + "epoch": 3.8793447862564925, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 9710 + }, + { + "epoch": 3.879744306831802, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 9711 + }, + { + "epoch": 3.8801438274071116, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 9712 + }, + { + "epoch": 3.880543347982421, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9713 + }, + { + "epoch": 3.8809428685577307, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 9714 + }, + { + "epoch": 3.8813423891330405, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9715 + }, + { + "epoch": 3.88174190970835, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9716 + }, + { + "epoch": 3.8821414302836597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9717 + }, + { + "epoch": 3.882540950858969, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9718 + }, + { + "epoch": 3.882940471434279, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9719 + }, + { + "epoch": 3.8833399920095886, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9720 + }, + { + "epoch": 3.8837395125848984, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9721 + }, + { + "epoch": 3.8841390331602077, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9722 + }, + { + "epoch": 3.8845385537355175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9723 + }, + { + "epoch": 3.884938074310827, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 9724 + }, + { + "epoch": 3.8853375948861366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9725 + }, + { + "epoch": 3.8857371154614464, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 9726 + }, + { + "epoch": 3.8861366360367557, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9727 + }, + { + "epoch": 3.8865361566120655, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 9728 + }, + { + "epoch": 3.8869356771873753, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9729 + }, + { + "epoch": 3.8873351977626847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9730 + }, + { + "epoch": 3.8877347183379944, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9731 + }, + { + "epoch": 3.8881342389133042, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9732 + }, + { + "epoch": 3.8885337594886136, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9733 + }, + { + "epoch": 3.8889332800639234, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9734 + }, + { + "epoch": 3.8893328006392327, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 9735 + }, + { + "epoch": 3.8897323212145425, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 9736 + }, + { + "epoch": 3.8901318417898523, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9737 + }, + { + "epoch": 3.890531362365162, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9738 + }, + { + "epoch": 3.8909308829404714, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 9739 + }, + { + "epoch": 3.891330403515781, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9740 + }, + { + "epoch": 3.8917299240910905, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9741 + }, + { + "epoch": 3.8921294446664003, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9742 + }, + { + "epoch": 3.89252896524171, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9743 + }, + { + "epoch": 3.8929284858170194, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9744 + }, + { + "epoch": 3.8933280063923292, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 9745 + }, + { + "epoch": 3.8937275269676386, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9746 + }, + { + "epoch": 3.8941270475429484, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 9747 + }, + { + "epoch": 3.894526568118258, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9748 + }, + { + "epoch": 3.894926088693568, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 9749 + }, + { + "epoch": 3.8953256092688773, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 9750 + }, + { + "epoch": 3.895725129844187, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 9751 + }, + { + "epoch": 3.8961246504194964, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9752 + }, + { + "epoch": 3.896524170994806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9753 + }, + { + "epoch": 3.896923691570116, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9754 + }, + { + "epoch": 3.8973232121454258, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9755 + }, + { + "epoch": 3.897722732720735, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9756 + }, + { + "epoch": 3.898122253296045, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 9757 + }, + { + "epoch": 3.8985217738713542, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9758 + }, + { + "epoch": 3.898921294446664, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 9759 + }, + { + "epoch": 3.899320815021974, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9760 + }, + { + "epoch": 3.899720335597283, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 9761 + }, + { + "epoch": 3.900119856172593, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9762 + }, + { + "epoch": 3.9005193767479023, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 9763 + }, + { + "epoch": 3.900918897323212, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 9764 + }, + { + "epoch": 3.901318417898522, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9765 + }, + { + "epoch": 3.9017179384738316, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9766 + }, + { + "epoch": 3.902117459049141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 9767 + }, + { + "epoch": 3.9025169796244508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9768 + }, + { + "epoch": 3.90291650019976, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 9769 + }, + { + "epoch": 3.90331602077507, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 9770 + }, + { + "epoch": 3.9037155413503797, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 9771 + }, + { + "epoch": 3.904115061925689, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9772 + }, + { + "epoch": 3.904514582500999, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 9773 + }, + { + "epoch": 3.9049141030763086, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 9774 + }, + { + "epoch": 3.905313623651618, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9775 + }, + { + "epoch": 3.9057131442269277, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9776 + }, + { + "epoch": 3.9061126648022375, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9777 + }, + { + "epoch": 3.906512185377547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 9778 + }, + { + "epoch": 3.9069117059528566, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9779 + }, + { + "epoch": 3.907311226528166, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9780 + }, + { + "epoch": 3.9077107471034758, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9781 + }, + { + "epoch": 3.9081102676787856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 9782 + }, + { + "epoch": 3.9085097882540953, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9783 + }, + { + "epoch": 3.9089093088294047, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 9784 + }, + { + "epoch": 3.9093088294047145, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 9785 + }, + { + "epoch": 3.909708349980024, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9786 + }, + { + "epoch": 3.9101078705553336, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9787 + }, + { + "epoch": 3.9105073911306434, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9788 + }, + { + "epoch": 3.9109069117059527, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 9789 + }, + { + "epoch": 3.9113064322812625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9790 + }, + { + "epoch": 3.911705952856572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9791 + }, + { + "epoch": 3.9121054734318816, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9792 + }, + { + "epoch": 3.9125049940071914, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 9793 + }, + { + "epoch": 3.912904514582501, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9794 + }, + { + "epoch": 3.9133040351578106, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 9795 + }, + { + "epoch": 3.9137035557331203, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 9796 + }, + { + "epoch": 3.9141030763084297, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 9797 + }, + { + "epoch": 3.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9798 + }, + { + "epoch": 3.9149021174590493, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 9799 + }, + { + "epoch": 3.915301638034359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9800 + }, + { + "epoch": 3.9157011586096684, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 9801 + }, + { + "epoch": 3.916100679184978, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9802 + }, + { + "epoch": 3.9165001997602875, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9803 + }, + { + "epoch": 3.9168997203355973, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 9804 + }, + { + "epoch": 3.917299240910907, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9805 + }, + { + "epoch": 3.9176987614862164, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9806 + }, + { + "epoch": 3.918098282061526, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 9807 + }, + { + "epoch": 3.9184978026368356, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 9808 + }, + { + "epoch": 3.9188973232121453, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3394, + "step": 9809 + }, + { + "epoch": 3.919296843787455, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9810 + }, + { + "epoch": 3.919696364362765, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 9811 + }, + { + "epoch": 3.9200958849380743, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9812 + }, + { + "epoch": 3.920495405513384, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 9813 + }, + { + "epoch": 3.9208949260886934, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9814 + }, + { + "epoch": 3.921294446664003, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 9815 + }, + { + "epoch": 3.921693967239313, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9816 + }, + { + "epoch": 3.9220934878146223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 9817 + }, + { + "epoch": 3.922493008389932, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9818 + }, + { + "epoch": 3.922892528965242, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 9819 + }, + { + "epoch": 3.923292049540551, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9820 + }, + { + "epoch": 3.923691570115861, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9821 + }, + { + "epoch": 3.924091090691171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9822 + }, + { + "epoch": 3.92449061126648, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9823 + }, + { + "epoch": 3.92489013184179, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 9824 + }, + { + "epoch": 3.9252896524170993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 9825 + }, + { + "epoch": 3.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9826 + }, + { + "epoch": 3.926088693567719, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 9827 + }, + { + "epoch": 3.9264882141430286, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9828 + }, + { + "epoch": 3.926887734718338, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9829 + }, + { + "epoch": 3.9272872552936477, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 9830 + }, + { + "epoch": 3.927686775868957, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 9831 + }, + { + "epoch": 3.928086296444267, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 9832 + }, + { + "epoch": 3.9284858170195767, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 9833 + }, + { + "epoch": 3.928885337594886, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 9834 + }, + { + "epoch": 3.929284858170196, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9835 + }, + { + "epoch": 3.929684378745505, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 9836 + }, + { + "epoch": 3.930083899320815, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 9837 + }, + { + "epoch": 3.9304834198961247, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9838 + }, + { + "epoch": 3.9308829404714345, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9839 + }, + { + "epoch": 3.931282461046744, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 9840 + }, + { + "epoch": 3.9316819816220536, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 9841 + }, + { + "epoch": 3.932081502197363, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 9842 + }, + { + "epoch": 3.9324810227726728, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9843 + }, + { + "epoch": 3.9328805433479825, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9844 + }, + { + "epoch": 3.9332800639232923, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 9845 + }, + { + "epoch": 3.9336795844986017, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 9846 + }, + { + "epoch": 3.9340791050739115, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9847 + }, + { + "epoch": 3.934478625649221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9848 + }, + { + "epoch": 3.9348781462245306, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 9849 + }, + { + "epoch": 3.9352776667998404, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9850 + }, + { + "epoch": 3.9356771873751497, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 9851 + }, + { + "epoch": 3.9360767079504595, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9852 + }, + { + "epoch": 3.936476228525769, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 9853 + }, + { + "epoch": 3.9368757491010786, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9854 + }, + { + "epoch": 3.9372752696763884, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 9855 + }, + { + "epoch": 3.937674790251698, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 9856 + }, + { + "epoch": 3.9380743108270075, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 9857 + }, + { + "epoch": 3.9384738314023173, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 9858 + }, + { + "epoch": 3.9388733519776267, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 9859 + }, + { + "epoch": 3.9392728725529365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9860 + }, + { + "epoch": 3.9396723931282462, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9861 + }, + { + "epoch": 3.9400719137035556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 9862 + }, + { + "epoch": 3.9404714342788654, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 9863 + }, + { + "epoch": 3.9408709548541747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9864 + }, + { + "epoch": 3.9412704754294845, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9865 + }, + { + "epoch": 3.9416699960047943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 9866 + }, + { + "epoch": 3.942069516580104, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 9867 + }, + { + "epoch": 3.9424690371554134, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9868 + }, + { + "epoch": 3.942868557730723, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 9869 + }, + { + "epoch": 3.9432680783060325, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9870 + }, + { + "epoch": 3.9436675988813423, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9871 + }, + { + "epoch": 3.944067119456652, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 9872 + }, + { + "epoch": 3.944466640031962, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 9873 + }, + { + "epoch": 3.9448661606072712, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9874 + }, + { + "epoch": 3.945265681182581, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9875 + }, + { + "epoch": 3.9456652017578904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 9876 + }, + { + "epoch": 3.9460647223332, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 9877 + }, + { + "epoch": 3.94646424290851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 9878 + }, + { + "epoch": 3.9468637634838193, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9879 + }, + { + "epoch": 3.947263284059129, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9880 + }, + { + "epoch": 3.9476628046344384, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 9881 + }, + { + "epoch": 3.948062325209748, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9882 + }, + { + "epoch": 3.948461845785058, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 9883 + }, + { + "epoch": 3.9488613663603678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9884 + }, + { + "epoch": 3.949260886935677, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9885 + }, + { + "epoch": 3.949660407510987, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9886 + }, + { + "epoch": 3.9500599280862962, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 9887 + }, + { + "epoch": 3.950459448661606, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 9888 + }, + { + "epoch": 3.950858969236916, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 9889 + }, + { + "epoch": 3.9512584898122256, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9890 + }, + { + "epoch": 3.951658010387535, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 9891 + }, + { + "epoch": 3.9520575309628447, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9892 + }, + { + "epoch": 3.952457051538154, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 9893 + }, + { + "epoch": 3.952856572113464, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 9894 + }, + { + "epoch": 3.9532560926887736, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 9895 + }, + { + "epoch": 3.953655613264083, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 9896 + }, + { + "epoch": 3.9540551338393928, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9897 + }, + { + "epoch": 3.954454654414702, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9898 + }, + { + "epoch": 3.954854174990012, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 9899 + }, + { + "epoch": 3.9552536955653217, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 9900 + }, + { + "epoch": 3.9556532161406315, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9901 + }, + { + "epoch": 3.956052736715941, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 9902 + }, + { + "epoch": 3.9564522572912506, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9903 + }, + { + "epoch": 3.95685177786656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 9904 + }, + { + "epoch": 3.9572512984418697, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 9905 + }, + { + "epoch": 3.9576508190171795, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 9906 + }, + { + "epoch": 3.958050339592489, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9907 + }, + { + "epoch": 3.9584498601677987, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9908 + }, + { + "epoch": 3.958849380743108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9909 + }, + { + "epoch": 3.959248901318418, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9910 + }, + { + "epoch": 3.9596484218937276, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 9911 + }, + { + "epoch": 3.9600479424690374, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 9912 + }, + { + "epoch": 3.9604474630443467, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9913 + }, + { + "epoch": 3.9608469836196565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9914 + }, + { + "epoch": 3.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 9915 + }, + { + "epoch": 3.9616460247702756, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9916 + }, + { + "epoch": 3.9620455453455854, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 9917 + }, + { + "epoch": 3.962445065920895, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 9918 + }, + { + "epoch": 3.9628445864962045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 9919 + }, + { + "epoch": 3.9632441070715143, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9920 + }, + { + "epoch": 3.9636436276468237, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 9921 + }, + { + "epoch": 3.9640431482221334, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 9922 + }, + { + "epoch": 3.9644426687974432, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 9923 + }, + { + "epoch": 3.9648421893727526, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 9924 + }, + { + "epoch": 3.9652417099480624, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 9925 + }, + { + "epoch": 3.9656412305233717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 9926 + }, + { + "epoch": 3.9660407510986815, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 9927 + }, + { + "epoch": 3.9664402716739913, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 9928 + }, + { + "epoch": 3.966839792249301, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 9929 + }, + { + "epoch": 3.9672393128246104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 9930 + }, + { + "epoch": 3.96763883339992, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 9931 + }, + { + "epoch": 3.9680383539752295, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 9932 + }, + { + "epoch": 3.9684378745505393, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9933 + }, + { + "epoch": 3.968837395125849, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 9934 + }, + { + "epoch": 3.969236915701159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 9935 + }, + { + "epoch": 3.9696364362764682, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 9936 + }, + { + "epoch": 3.970035956851778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 9937 + }, + { + "epoch": 3.9704354774270874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9938 + }, + { + "epoch": 3.970834998002397, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9939 + }, + { + "epoch": 3.971234518577707, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 9940 + }, + { + "epoch": 3.9716340391530163, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 9941 + }, + { + "epoch": 3.972033559728326, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 9942 + }, + { + "epoch": 3.9724330803036354, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 9943 + }, + { + "epoch": 3.972832600878945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 9944 + }, + { + "epoch": 3.973232121454255, + "grad_norm": 0.04736328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 9945 + }, + { + "epoch": 3.9736316420295648, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9946 + }, + { + "epoch": 3.974031162604874, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 9947 + }, + { + "epoch": 3.974430683180184, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 9948 + }, + { + "epoch": 3.9748302037554932, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 9949 + }, + { + "epoch": 3.975229724330803, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 9950 + }, + { + "epoch": 3.975629244906113, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 9951 + }, + { + "epoch": 3.976028765481422, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9952 + }, + { + "epoch": 3.976428286056732, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 9953 + }, + { + "epoch": 3.9768278066320413, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 9954 + }, + { + "epoch": 3.977227327207351, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 9955 + }, + { + "epoch": 3.977626847782661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 9956 + }, + { + "epoch": 3.9780263683579706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 9957 + }, + { + "epoch": 3.97842588893328, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 9958 + }, + { + "epoch": 3.9788254095085898, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 9959 + }, + { + "epoch": 3.979224930083899, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 9960 + }, + { + "epoch": 3.979624450659209, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 9961 + }, + { + "epoch": 3.9800239712345187, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 9962 + }, + { + "epoch": 3.9804234918098285, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9963 + }, + { + "epoch": 3.980823012385138, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 9964 + }, + { + "epoch": 3.9812225329604476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 9965 + }, + { + "epoch": 3.981622053535757, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 9966 + }, + { + "epoch": 3.9820215741110667, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 9967 + }, + { + "epoch": 3.9824210946863765, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 9968 + }, + { + "epoch": 3.982820615261686, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 9969 + }, + { + "epoch": 3.9832201358369956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 9970 + }, + { + "epoch": 3.983619656412305, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 9971 + }, + { + "epoch": 3.9840191769876148, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 9972 + }, + { + "epoch": 3.9844186975629245, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 9973 + }, + { + "epoch": 3.9848182181382343, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 9974 + }, + { + "epoch": 3.9852177387135437, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9975 + }, + { + "epoch": 3.9856172592888535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9976 + }, + { + "epoch": 3.986016779864163, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 9977 + }, + { + "epoch": 3.9864163004394726, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9978 + }, + { + "epoch": 3.9868158210147824, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 9979 + }, + { + "epoch": 3.987215341590092, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 9980 + }, + { + "epoch": 3.9876148621654015, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9981 + }, + { + "epoch": 3.9880143827407113, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 9982 + }, + { + "epoch": 3.9884139033160206, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 9983 + }, + { + "epoch": 3.9888134238913304, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 9984 + }, + { + "epoch": 3.98921294446664, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 9985 + }, + { + "epoch": 3.9896124650419496, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 9986 + }, + { + "epoch": 3.9900119856172593, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 9987 + }, + { + "epoch": 3.9904115061925687, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 9988 + }, + { + "epoch": 3.9908110267678785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 9989 + }, + { + "epoch": 3.9912105473431883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 9990 + }, + { + "epoch": 3.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 9991 + }, + { + "epoch": 3.9920095884938074, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 9992 + }, + { + "epoch": 3.992409109069117, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 9993 + }, + { + "epoch": 3.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 9994 + }, + { + "epoch": 3.9932081502197363, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 9995 + }, + { + "epoch": 3.993607670795046, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 9996 + }, + { + "epoch": 3.9940071913703554, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 9997 + }, + { + "epoch": 3.994406711945665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 9998 + }, + { + "epoch": 3.9948062325209746, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 9999 + }, + { + "epoch": 3.9952057530962843, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10000 + }, + { + "epoch": 3.995605273671594, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10001 + }, + { + "epoch": 3.996004794246904, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10002 + }, + { + "epoch": 3.9964043148222133, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10003 + }, + { + "epoch": 3.996803835397523, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10004 + }, + { + "epoch": 3.9972033559728324, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10005 + }, + { + "epoch": 3.997602876548142, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10006 + }, + { + "epoch": 3.998002397123452, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10007 + }, + { + "epoch": 3.9984019176987617, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10008 + }, + { + "epoch": 3.998801438274071, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10009 + }, + { + "epoch": 3.999200958849381, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 10010 + }, + { + "epoch": 3.99960047942469, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 10011 + }, + { + "epoch": 4.0, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10012 + }, + { + "epoch": 4.00039952057531, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10013 + }, + { + "epoch": 4.00079904115062, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10014 + }, + { + "epoch": 4.0011985617259285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10015 + }, + { + "epoch": 4.001598082301238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10016 + }, + { + "epoch": 4.001997602876548, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10017 + }, + { + "epoch": 4.002397123451858, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10018 + }, + { + "epoch": 4.002796644027168, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10019 + }, + { + "epoch": 4.003196164602477, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10020 + }, + { + "epoch": 4.003595685177786, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10021 + }, + { + "epoch": 4.003995205753096, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10022 + }, + { + "epoch": 4.004394726328406, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10023 + }, + { + "epoch": 4.004794246903716, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10024 + }, + { + "epoch": 4.0051937674790254, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10025 + }, + { + "epoch": 4.005593288054335, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10026 + }, + { + "epoch": 4.005992808629644, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10027 + }, + { + "epoch": 4.006392329204954, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10028 + }, + { + "epoch": 4.006791849780264, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10029 + }, + { + "epoch": 4.0071913703555735, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10030 + }, + { + "epoch": 4.007590890930883, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10031 + }, + { + "epoch": 4.007990411506192, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10032 + }, + { + "epoch": 4.008389932081502, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10033 + }, + { + "epoch": 4.008789452656812, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10034 + }, + { + "epoch": 4.0091889732321215, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10035 + }, + { + "epoch": 4.009588493807431, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 10036 + }, + { + "epoch": 4.009988014382741, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10037 + }, + { + "epoch": 4.01038753495805, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 10038 + }, + { + "epoch": 4.01078705553336, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10039 + }, + { + "epoch": 4.01118657610867, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10040 + }, + { + "epoch": 4.011586096683979, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10041 + }, + { + "epoch": 4.011985617259289, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10042 + }, + { + "epoch": 4.012385137834598, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10043 + }, + { + "epoch": 4.012784658409908, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10044 + }, + { + "epoch": 4.013184178985218, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10045 + }, + { + "epoch": 4.013583699560527, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10046 + }, + { + "epoch": 4.013983220135837, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10047 + }, + { + "epoch": 4.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10048 + }, + { + "epoch": 4.014782261286456, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 10049 + }, + { + "epoch": 4.015181781861766, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10050 + }, + { + "epoch": 4.0155813024370755, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10051 + }, + { + "epoch": 4.015980823012385, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10052 + }, + { + "epoch": 4.016380343587695, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10053 + }, + { + "epoch": 4.016779864163005, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10054 + }, + { + "epoch": 4.017179384738314, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10055 + }, + { + "epoch": 4.0175789053136235, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10056 + }, + { + "epoch": 4.017978425888933, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10057 + }, + { + "epoch": 4.018377946464243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10058 + }, + { + "epoch": 4.018777467039553, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10059 + }, + { + "epoch": 4.019176987614862, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 10060 + }, + { + "epoch": 4.0195765081901715, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10061 + }, + { + "epoch": 4.019976028765481, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10062 + }, + { + "epoch": 4.020375549340791, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10063 + }, + { + "epoch": 4.020775069916101, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10064 + }, + { + "epoch": 4.021174590491411, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10065 + }, + { + "epoch": 4.02157411106672, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10066 + }, + { + "epoch": 4.021973631642029, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10067 + }, + { + "epoch": 4.022373152217339, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10068 + }, + { + "epoch": 4.022772672792649, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10069 + }, + { + "epoch": 4.023172193367959, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10070 + }, + { + "epoch": 4.0235717139432685, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10071 + }, + { + "epoch": 4.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10072 + }, + { + "epoch": 4.024370755093887, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10073 + }, + { + "epoch": 4.024770275669197, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10074 + }, + { + "epoch": 4.025169796244507, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10075 + }, + { + "epoch": 4.025569316819817, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10076 + }, + { + "epoch": 4.0259688373951255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10077 + }, + { + "epoch": 4.026368357970435, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10078 + }, + { + "epoch": 4.026767878545745, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 10079 + }, + { + "epoch": 4.027167399121055, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10080 + }, + { + "epoch": 4.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10081 + }, + { + "epoch": 4.027966440271674, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10082 + }, + { + "epoch": 4.028365960846983, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10083 + }, + { + "epoch": 4.028765481422293, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10084 + }, + { + "epoch": 4.029165001997603, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10085 + }, + { + "epoch": 4.029564522572913, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10086 + }, + { + "epoch": 4.029964043148222, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10087 + }, + { + "epoch": 4.030363563723531, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10088 + }, + { + "epoch": 4.030763084298841, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10089 + }, + { + "epoch": 4.031162604874151, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10090 + }, + { + "epoch": 4.031562125449461, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 10091 + }, + { + "epoch": 4.0319616460247705, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10092 + }, + { + "epoch": 4.03236116660008, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10093 + }, + { + "epoch": 4.032760687175389, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10094 + }, + { + "epoch": 4.033160207750699, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10095 + }, + { + "epoch": 4.033559728326009, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10096 + }, + { + "epoch": 4.0339592489013185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10097 + }, + { + "epoch": 4.034358769476628, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10098 + }, + { + "epoch": 4.034758290051938, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10099 + }, + { + "epoch": 4.035157810627247, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10100 + }, + { + "epoch": 4.035557331202557, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10101 + }, + { + "epoch": 4.035956851777867, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10102 + }, + { + "epoch": 4.036356372353176, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10103 + }, + { + "epoch": 4.036755892928486, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10104 + }, + { + "epoch": 4.037155413503795, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10105 + }, + { + "epoch": 4.037554934079105, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10106 + }, + { + "epoch": 4.037954454654415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 10107 + }, + { + "epoch": 4.038353975229724, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10108 + }, + { + "epoch": 4.038753495805034, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10109 + }, + { + "epoch": 4.039153016380344, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10110 + }, + { + "epoch": 4.039552536955653, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10111 + }, + { + "epoch": 4.039952057530963, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10112 + }, + { + "epoch": 4.040351578106272, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10113 + }, + { + "epoch": 4.040751098681582, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10114 + }, + { + "epoch": 4.041150619256892, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10115 + }, + { + "epoch": 4.041550139832202, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10116 + }, + { + "epoch": 4.041949660407511, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10117 + }, + { + "epoch": 4.0423491809828205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10118 + }, + { + "epoch": 4.04274870155813, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10119 + }, + { + "epoch": 4.04314822213344, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10120 + }, + { + "epoch": 4.04354774270875, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10121 + }, + { + "epoch": 4.043947263284059, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10122 + }, + { + "epoch": 4.0443467838593685, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10123 + }, + { + "epoch": 4.044746304434678, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10124 + }, + { + "epoch": 4.045145825009988, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10125 + }, + { + "epoch": 4.045545345585298, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 10126 + }, + { + "epoch": 4.045944866160608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10127 + }, + { + "epoch": 4.046344386735917, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10128 + }, + { + "epoch": 4.046743907311226, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10129 + }, + { + "epoch": 4.047143427886536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10130 + }, + { + "epoch": 4.047542948461846, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10131 + }, + { + "epoch": 4.047942469037156, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10132 + }, + { + "epoch": 4.048341989612465, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10133 + }, + { + "epoch": 4.048741510187774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10134 + }, + { + "epoch": 4.049141030763084, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10135 + }, + { + "epoch": 4.049540551338394, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10136 + }, + { + "epoch": 4.049940071913704, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10137 + }, + { + "epoch": 4.0503395924890135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10138 + }, + { + "epoch": 4.050739113064322, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10139 + }, + { + "epoch": 4.051138633639632, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3388, + "step": 10140 + }, + { + "epoch": 4.051538154214942, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10141 + }, + { + "epoch": 4.051937674790252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10142 + }, + { + "epoch": 4.052337195365562, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10143 + }, + { + "epoch": 4.052736715940871, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 10144 + }, + { + "epoch": 4.05313623651618, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10145 + }, + { + "epoch": 4.05353575709149, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10146 + }, + { + "epoch": 4.0539352776668, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10147 + }, + { + "epoch": 4.05433479824211, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10148 + }, + { + "epoch": 4.054734318817419, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10149 + }, + { + "epoch": 4.055133839392728, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10150 + }, + { + "epoch": 4.055533359968038, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10151 + }, + { + "epoch": 4.055932880543348, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10152 + }, + { + "epoch": 4.056332401118658, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10153 + }, + { + "epoch": 4.0567319216939675, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10154 + }, + { + "epoch": 4.057131442269277, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10155 + }, + { + "epoch": 4.057530962844586, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10156 + }, + { + "epoch": 4.057930483419896, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10157 + }, + { + "epoch": 4.058330003995206, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10158 + }, + { + "epoch": 4.0587295245705155, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10159 + }, + { + "epoch": 4.059129045145825, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10160 + }, + { + "epoch": 4.059528565721135, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10161 + }, + { + "epoch": 4.059928086296444, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10162 + }, + { + "epoch": 4.060327606871754, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10163 + }, + { + "epoch": 4.0607271274470635, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10164 + }, + { + "epoch": 4.061126648022373, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10165 + }, + { + "epoch": 4.061526168597683, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10166 + }, + { + "epoch": 4.061925689172992, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10167 + }, + { + "epoch": 4.062325209748302, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10168 + }, + { + "epoch": 4.062724730323612, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10169 + }, + { + "epoch": 4.063124250898921, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10170 + }, + { + "epoch": 4.063523771474231, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10171 + }, + { + "epoch": 4.063923292049541, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10172 + }, + { + "epoch": 4.06432281262485, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10173 + }, + { + "epoch": 4.06472233320016, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10174 + }, + { + "epoch": 4.065121853775469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10175 + }, + { + "epoch": 4.065521374350779, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10176 + }, + { + "epoch": 4.065920894926089, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10177 + }, + { + "epoch": 4.066320415501398, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10178 + }, + { + "epoch": 4.066719936076708, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10179 + }, + { + "epoch": 4.0671194566520175, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10180 + }, + { + "epoch": 4.067518977227327, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10181 + }, + { + "epoch": 4.067918497802637, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10182 + }, + { + "epoch": 4.068318018377947, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10183 + }, + { + "epoch": 4.068717538953256, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10184 + }, + { + "epoch": 4.0691170595285655, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10185 + }, + { + "epoch": 4.069516580103875, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10186 + }, + { + "epoch": 4.069916100679185, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10187 + }, + { + "epoch": 4.070315621254495, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10188 + }, + { + "epoch": 4.070715141829805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10189 + }, + { + "epoch": 4.0711146624051135, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10190 + }, + { + "epoch": 4.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10191 + }, + { + "epoch": 4.071913703555733, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10192 + }, + { + "epoch": 4.072313224131043, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10193 + }, + { + "epoch": 4.072712744706353, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10194 + }, + { + "epoch": 4.073112265281662, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 10195 + }, + { + "epoch": 4.073511785856971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10196 + }, + { + "epoch": 4.073911306432281, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10197 + }, + { + "epoch": 4.074310827007591, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10198 + }, + { + "epoch": 4.074710347582901, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10199 + }, + { + "epoch": 4.0751098681582105, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10200 + }, + { + "epoch": 4.075509388733519, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10201 + }, + { + "epoch": 4.075908909308829, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 10202 + }, + { + "epoch": 4.076308429884139, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10203 + }, + { + "epoch": 4.076707950459449, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3406, + "step": 10204 + }, + { + "epoch": 4.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10205 + }, + { + "epoch": 4.077506991610068, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10206 + }, + { + "epoch": 4.077906512185377, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10207 + }, + { + "epoch": 4.078306032760687, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10208 + }, + { + "epoch": 4.078705553335997, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10209 + }, + { + "epoch": 4.079105073911307, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10210 + }, + { + "epoch": 4.079504594486616, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10211 + }, + { + "epoch": 4.079904115061925, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10212 + }, + { + "epoch": 4.080303635637235, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10213 + }, + { + "epoch": 4.080703156212545, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10214 + }, + { + "epoch": 4.081102676787855, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10215 + }, + { + "epoch": 4.081502197363164, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10216 + }, + { + "epoch": 4.081901717938474, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10217 + }, + { + "epoch": 4.082301238513783, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10218 + }, + { + "epoch": 4.082700759089093, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10219 + }, + { + "epoch": 4.083100279664403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10220 + }, + { + "epoch": 4.0834998002397125, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 10221 + }, + { + "epoch": 4.083899320815022, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10222 + }, + { + "epoch": 4.084298841390331, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10223 + }, + { + "epoch": 4.084698361965641, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10224 + }, + { + "epoch": 4.085097882540951, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 10225 + }, + { + "epoch": 4.0854974031162605, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10226 + }, + { + "epoch": 4.08589692369157, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10227 + }, + { + "epoch": 4.08629644426688, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10228 + }, + { + "epoch": 4.086695964842189, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10229 + }, + { + "epoch": 4.087095485417499, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10230 + }, + { + "epoch": 4.087495005992809, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10231 + }, + { + "epoch": 4.087894526568118, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10232 + }, + { + "epoch": 4.088294047143428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10233 + }, + { + "epoch": 4.088693567718738, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10234 + }, + { + "epoch": 4.089093088294047, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10235 + }, + { + "epoch": 4.089492608869357, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10236 + }, + { + "epoch": 4.089892129444666, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10237 + }, + { + "epoch": 4.090291650019976, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10238 + }, + { + "epoch": 4.090691170595286, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 10239 + }, + { + "epoch": 4.091090691170595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10240 + }, + { + "epoch": 4.091490211745905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10241 + }, + { + "epoch": 4.0918897323212144, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10242 + }, + { + "epoch": 4.092289252896524, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10243 + }, + { + "epoch": 4.092688773471834, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10244 + }, + { + "epoch": 4.093088294047144, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10245 + }, + { + "epoch": 4.093487814622453, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10246 + }, + { + "epoch": 4.0938873351977625, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10247 + }, + { + "epoch": 4.094286855773072, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10248 + }, + { + "epoch": 4.094686376348382, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 10249 + }, + { + "epoch": 4.095085896923692, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10250 + }, + { + "epoch": 4.095485417499002, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10251 + }, + { + "epoch": 4.0958849380743105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10252 + }, + { + "epoch": 4.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10253 + }, + { + "epoch": 4.09668397922493, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10254 + }, + { + "epoch": 4.09708349980024, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10255 + }, + { + "epoch": 4.09748302037555, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10256 + }, + { + "epoch": 4.097882540950859, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10257 + }, + { + "epoch": 4.098282061526168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10258 + }, + { + "epoch": 4.098681582101478, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10259 + }, + { + "epoch": 4.099081102676788, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10260 + }, + { + "epoch": 4.099480623252098, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10261 + }, + { + "epoch": 4.0998801438274075, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10262 + }, + { + "epoch": 4.100279664402716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10263 + }, + { + "epoch": 4.100679184978026, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10264 + }, + { + "epoch": 4.101078705553336, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10265 + }, + { + "epoch": 4.101478226128646, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10266 + }, + { + "epoch": 4.1018777467039556, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 10267 + }, + { + "epoch": 4.1022772672792644, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10268 + }, + { + "epoch": 4.102676787854574, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 10269 + }, + { + "epoch": 4.103076308429884, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10270 + }, + { + "epoch": 4.103475829005194, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10271 + }, + { + "epoch": 4.103875349580504, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10272 + }, + { + "epoch": 4.104274870155813, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3423, + "step": 10273 + }, + { + "epoch": 4.104674390731122, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10274 + }, + { + "epoch": 4.105073911306432, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10275 + }, + { + "epoch": 4.105473431881742, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10276 + }, + { + "epoch": 4.105872952457052, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10277 + }, + { + "epoch": 4.106272473032361, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 10278 + }, + { + "epoch": 4.106671993607671, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10279 + }, + { + "epoch": 4.10707151418298, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10280 + }, + { + "epoch": 4.10747103475829, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10281 + }, + { + "epoch": 4.1078705553336, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10282 + }, + { + "epoch": 4.1082700759089095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10283 + }, + { + "epoch": 4.108669596484219, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10284 + }, + { + "epoch": 4.109069117059528, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10285 + }, + { + "epoch": 4.109468637634838, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10286 + }, + { + "epoch": 4.109868158210148, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10287 + }, + { + "epoch": 4.1102676787854575, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10288 + }, + { + "epoch": 4.110667199360767, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10289 + }, + { + "epoch": 4.111066719936077, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10290 + }, + { + "epoch": 4.111466240511386, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10291 + }, + { + "epoch": 4.111865761086696, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10292 + }, + { + "epoch": 4.112265281662006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10293 + }, + { + "epoch": 4.112664802237315, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10294 + }, + { + "epoch": 4.113064322812625, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 10295 + }, + { + "epoch": 4.113463843387935, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10296 + }, + { + "epoch": 4.113863363963244, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10297 + }, + { + "epoch": 4.114262884538554, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10298 + }, + { + "epoch": 4.114662405113863, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10299 + }, + { + "epoch": 4.115061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10300 + }, + { + "epoch": 4.115461446264483, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10301 + }, + { + "epoch": 4.115860966839792, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10302 + }, + { + "epoch": 4.116260487415102, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10303 + }, + { + "epoch": 4.116660007990411, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10304 + }, + { + "epoch": 4.117059528565721, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10305 + }, + { + "epoch": 4.117459049141031, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10306 + }, + { + "epoch": 4.117858569716341, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10307 + }, + { + "epoch": 4.11825809029165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10308 + }, + { + "epoch": 4.1186576108669595, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10309 + }, + { + "epoch": 4.119057131442269, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10310 + }, + { + "epoch": 4.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 10311 + }, + { + "epoch": 4.119856172592889, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 10312 + }, + { + "epoch": 4.120255693168198, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10313 + }, + { + "epoch": 4.1206552137435075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 10314 + }, + { + "epoch": 4.121054734318817, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10315 + }, + { + "epoch": 4.121454254894127, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10316 + }, + { + "epoch": 4.121853775469437, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10317 + }, + { + "epoch": 4.122253296044747, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 10318 + }, + { + "epoch": 4.122652816620056, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10319 + }, + { + "epoch": 4.123052337195365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10320 + }, + { + "epoch": 4.123451857770675, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10321 + }, + { + "epoch": 4.123851378345985, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10322 + }, + { + "epoch": 4.124250898921295, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10323 + }, + { + "epoch": 4.1246504194966045, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10324 + }, + { + "epoch": 4.125049940071913, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10325 + }, + { + "epoch": 4.125449460647223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10326 + }, + { + "epoch": 4.125848981222533, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 10327 + }, + { + "epoch": 4.126248501797843, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 10328 + }, + { + "epoch": 4.1266480223731525, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10329 + }, + { + "epoch": 4.127047542948461, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10330 + }, + { + "epoch": 4.127447063523771, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10331 + }, + { + "epoch": 4.127846584099081, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10332 + }, + { + "epoch": 4.128246104674391, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 10333 + }, + { + "epoch": 4.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10334 + }, + { + "epoch": 4.12904514582501, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10335 + }, + { + "epoch": 4.129444666400319, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10336 + }, + { + "epoch": 4.129844186975629, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 10337 + }, + { + "epoch": 4.130243707550939, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10338 + }, + { + "epoch": 4.130643228126249, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10339 + }, + { + "epoch": 4.131042748701558, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10340 + }, + { + "epoch": 4.131442269276867, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10341 + }, + { + "epoch": 4.131841789852177, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10342 + }, + { + "epoch": 4.132241310427487, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10343 + }, + { + "epoch": 4.132640831002797, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10344 + }, + { + "epoch": 4.1330403515781065, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10345 + }, + { + "epoch": 4.133439872153416, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10346 + }, + { + "epoch": 4.133839392728725, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10347 + }, + { + "epoch": 4.134238913304035, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10348 + }, + { + "epoch": 4.134638433879345, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10349 + }, + { + "epoch": 4.1350379544546545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10350 + }, + { + "epoch": 4.135437475029964, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 10351 + }, + { + "epoch": 4.135836995605274, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 10352 + }, + { + "epoch": 4.136236516180583, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10353 + }, + { + "epoch": 4.136636036755893, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10354 + }, + { + "epoch": 4.1370355573312025, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10355 + }, + { + "epoch": 4.137435077906512, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10356 + }, + { + "epoch": 4.137834598481822, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10357 + }, + { + "epoch": 4.138234119057131, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10358 + }, + { + "epoch": 4.138633639632441, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10359 + }, + { + "epoch": 4.139033160207751, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10360 + }, + { + "epoch": 4.13943268078306, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10361 + }, + { + "epoch": 4.13983220135837, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10362 + }, + { + "epoch": 4.14023172193368, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10363 + }, + { + "epoch": 4.140631242508989, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10364 + }, + { + "epoch": 4.141030763084299, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10365 + }, + { + "epoch": 4.141430283659608, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10366 + }, + { + "epoch": 4.141829804234918, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10367 + }, + { + "epoch": 4.142229324810228, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10368 + }, + { + "epoch": 4.142628845385538, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10369 + }, + { + "epoch": 4.143028365960847, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10370 + }, + { + "epoch": 4.1434278865361565, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10371 + }, + { + "epoch": 4.143827407111466, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10372 + }, + { + "epoch": 4.144226927686776, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10373 + }, + { + "epoch": 4.144626448262086, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10374 + }, + { + "epoch": 4.145025968837395, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10375 + }, + { + "epoch": 4.1454254894127045, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10376 + }, + { + "epoch": 4.145825009988014, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10377 + }, + { + "epoch": 4.146224530563324, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10378 + }, + { + "epoch": 4.146624051138634, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10379 + }, + { + "epoch": 4.147023571713944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10380 + }, + { + "epoch": 4.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10381 + }, + { + "epoch": 4.147822612864562, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10382 + }, + { + "epoch": 4.148222133439872, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10383 + }, + { + "epoch": 4.148621654015182, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10384 + }, + { + "epoch": 4.149021174590492, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10385 + }, + { + "epoch": 4.1494206951658015, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 10386 + }, + { + "epoch": 4.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10387 + }, + { + "epoch": 4.15021973631642, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10388 + }, + { + "epoch": 4.15061925689173, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10389 + }, + { + "epoch": 4.15101877746704, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10390 + }, + { + "epoch": 4.1514182980423495, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10391 + }, + { + "epoch": 4.151817818617658, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10392 + }, + { + "epoch": 4.152217339192968, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10393 + }, + { + "epoch": 4.152616859768278, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 10394 + }, + { + "epoch": 4.153016380343588, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10395 + }, + { + "epoch": 4.153415900918898, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 10396 + }, + { + "epoch": 4.153815421494207, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10397 + }, + { + "epoch": 4.154214942069516, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10398 + }, + { + "epoch": 4.154614462644826, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10399 + }, + { + "epoch": 4.155013983220136, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10400 + }, + { + "epoch": 4.155413503795446, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10401 + }, + { + "epoch": 4.155813024370755, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10402 + }, + { + "epoch": 4.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10403 + }, + { + "epoch": 4.156612065521374, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10404 + }, + { + "epoch": 4.157011586096684, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10405 + }, + { + "epoch": 4.157411106671994, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10406 + }, + { + "epoch": 4.157810627247303, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 10407 + }, + { + "epoch": 4.158210147822613, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10408 + }, + { + "epoch": 4.158609668397922, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10409 + }, + { + "epoch": 4.159009188973232, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10410 + }, + { + "epoch": 4.159408709548542, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10411 + }, + { + "epoch": 4.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10412 + }, + { + "epoch": 4.160207750699161, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 10413 + }, + { + "epoch": 4.160607271274471, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10414 + }, + { + "epoch": 4.16100679184978, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10415 + }, + { + "epoch": 4.16140631242509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10416 + }, + { + "epoch": 4.1618058330003995, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10417 + }, + { + "epoch": 4.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10418 + }, + { + "epoch": 4.162604874151019, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10419 + }, + { + "epoch": 4.163004394726328, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10420 + }, + { + "epoch": 4.163403915301638, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10421 + }, + { + "epoch": 4.163803435876948, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10422 + }, + { + "epoch": 4.164202956452257, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10423 + }, + { + "epoch": 4.164602477027567, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10424 + }, + { + "epoch": 4.165001997602877, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10425 + }, + { + "epoch": 4.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10426 + }, + { + "epoch": 4.165801038753496, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10427 + }, + { + "epoch": 4.166200559328805, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10428 + }, + { + "epoch": 4.166600079904115, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 10429 + }, + { + "epoch": 4.166999600479425, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10430 + }, + { + "epoch": 4.167399121054734, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10431 + }, + { + "epoch": 4.167798641630044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10432 + }, + { + "epoch": 4.168198162205353, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 10433 + }, + { + "epoch": 4.168597682780663, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 10434 + }, + { + "epoch": 4.168997203355973, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10435 + }, + { + "epoch": 4.169396723931283, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10436 + }, + { + "epoch": 4.169796244506592, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10437 + }, + { + "epoch": 4.1701957650819015, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10438 + }, + { + "epoch": 4.170595285657211, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10439 + }, + { + "epoch": 4.170994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10440 + }, + { + "epoch": 4.171394326807831, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10441 + }, + { + "epoch": 4.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10442 + }, + { + "epoch": 4.1721933679584495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10443 + }, + { + "epoch": 4.172592888533759, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 10444 + }, + { + "epoch": 4.172992409109069, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10445 + }, + { + "epoch": 4.173391929684379, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3401, + "step": 10446 + }, + { + "epoch": 4.173791450259689, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10447 + }, + { + "epoch": 4.174190970834998, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10448 + }, + { + "epoch": 4.174590491410307, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10449 + }, + { + "epoch": 4.174990011985617, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10450 + }, + { + "epoch": 4.175389532560927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10451 + }, + { + "epoch": 4.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10452 + }, + { + "epoch": 4.1761885737115465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 10453 + }, + { + "epoch": 4.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 10454 + }, + { + "epoch": 4.176987614862165, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10455 + }, + { + "epoch": 4.177387135437475, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10456 + }, + { + "epoch": 4.177786656012785, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10457 + }, + { + "epoch": 4.1781861765880945, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10458 + }, + { + "epoch": 4.178585697163404, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10459 + }, + { + "epoch": 4.178985217738713, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10460 + }, + { + "epoch": 4.179384738314023, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10461 + }, + { + "epoch": 4.179784258889333, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 10462 + }, + { + "epoch": 4.180183779464643, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10463 + }, + { + "epoch": 4.180583300039952, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10464 + }, + { + "epoch": 4.180982820615261, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10465 + }, + { + "epoch": 4.181382341190571, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10466 + }, + { + "epoch": 4.181781861765881, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10467 + }, + { + "epoch": 4.182181382341191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10468 + }, + { + "epoch": 4.1825809029165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 10469 + }, + { + "epoch": 4.18298042349181, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10470 + }, + { + "epoch": 4.183379944067119, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10471 + }, + { + "epoch": 4.183779464642429, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10472 + }, + { + "epoch": 4.184178985217739, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10473 + }, + { + "epoch": 4.1845785057930485, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10474 + }, + { + "epoch": 4.184978026368358, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10475 + }, + { + "epoch": 4.185377546943668, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10476 + }, + { + "epoch": 4.185777067518977, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10477 + }, + { + "epoch": 4.186176588094287, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 10478 + }, + { + "epoch": 4.1865761086695965, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10479 + }, + { + "epoch": 4.186975629244906, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10480 + }, + { + "epoch": 4.187375149820216, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10481 + }, + { + "epoch": 4.187774670395525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10482 + }, + { + "epoch": 4.188174190970835, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10483 + }, + { + "epoch": 4.1885737115461446, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10484 + }, + { + "epoch": 4.188973232121454, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10485 + }, + { + "epoch": 4.189372752696764, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10486 + }, + { + "epoch": 4.189772273272074, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10487 + }, + { + "epoch": 4.190171793847383, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10488 + }, + { + "epoch": 4.190571314422693, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10489 + }, + { + "epoch": 4.190970834998002, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10490 + }, + { + "epoch": 4.191370355573312, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10491 + }, + { + "epoch": 4.191769876148622, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 10492 + }, + { + "epoch": 4.192169396723932, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10493 + }, + { + "epoch": 4.192568917299241, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10494 + }, + { + "epoch": 4.19296843787455, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10495 + }, + { + "epoch": 4.19336795844986, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 10496 + }, + { + "epoch": 4.19376747902517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3366, + "step": 10497 + }, + { + "epoch": 4.19416699960048, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 10498 + }, + { + "epoch": 4.194566520175789, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10499 + }, + { + "epoch": 4.1949660407510985, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 10500 + }, + { + "epoch": 4.195365561326408, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 10501 + }, + { + "epoch": 4.195765081901718, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10502 + }, + { + "epoch": 4.196164602477028, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10503 + }, + { + "epoch": 4.196564123052338, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10504 + }, + { + "epoch": 4.1969636436276465, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10505 + }, + { + "epoch": 4.197363164202956, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10506 + }, + { + "epoch": 4.197762684778266, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10507 + }, + { + "epoch": 4.198162205353576, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10508 + }, + { + "epoch": 4.198561725928886, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10509 + }, + { + "epoch": 4.1989612465041946, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10510 + }, + { + "epoch": 4.199360767079504, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10511 + }, + { + "epoch": 4.199760287654814, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10512 + }, + { + "epoch": 4.200159808230124, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10513 + }, + { + "epoch": 4.200559328805434, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10514 + }, + { + "epoch": 4.2009588493807435, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10515 + }, + { + "epoch": 4.201358369956052, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10516 + }, + { + "epoch": 4.201757890531362, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10517 + }, + { + "epoch": 4.202157411106672, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3434, + "step": 10518 + }, + { + "epoch": 4.202556931681982, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10519 + }, + { + "epoch": 4.2029564522572915, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10520 + }, + { + "epoch": 4.2033559728326, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10521 + }, + { + "epoch": 4.20375549340791, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10522 + }, + { + "epoch": 4.20415501398322, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10523 + }, + { + "epoch": 4.20455453455853, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10524 + }, + { + "epoch": 4.20495405513384, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10525 + }, + { + "epoch": 4.205353575709149, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10526 + }, + { + "epoch": 4.205753096284458, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10527 + }, + { + "epoch": 4.206152616859768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10528 + }, + { + "epoch": 4.206552137435078, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10529 + }, + { + "epoch": 4.206951658010388, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10530 + }, + { + "epoch": 4.207351178585697, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10531 + }, + { + "epoch": 4.207750699161007, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 10532 + }, + { + "epoch": 4.208150219736316, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 10533 + }, + { + "epoch": 4.208549740311626, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10534 + }, + { + "epoch": 4.208949260886936, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10535 + }, + { + "epoch": 4.2093487814622454, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10536 + }, + { + "epoch": 4.209748302037555, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 10537 + }, + { + "epoch": 4.210147822612864, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10538 + }, + { + "epoch": 4.210547343188174, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10539 + }, + { + "epoch": 4.210946863763484, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10540 + }, + { + "epoch": 4.2113463843387935, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10541 + }, + { + "epoch": 4.211745904914103, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 10542 + }, + { + "epoch": 4.212145425489413, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10543 + }, + { + "epoch": 4.212544946064722, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10544 + }, + { + "epoch": 4.212944466640032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10545 + }, + { + "epoch": 4.2133439872153415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 10546 + }, + { + "epoch": 4.213743507790651, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10547 + }, + { + "epoch": 4.214143028365961, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10548 + }, + { + "epoch": 4.214542548941271, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10549 + }, + { + "epoch": 4.21494206951658, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10550 + }, + { + "epoch": 4.21534159009189, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10551 + }, + { + "epoch": 4.215741110667199, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10552 + }, + { + "epoch": 4.216140631242509, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10553 + }, + { + "epoch": 4.216540151817819, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10554 + }, + { + "epoch": 4.216939672393128, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10555 + }, + { + "epoch": 4.217339192968438, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10556 + }, + { + "epoch": 4.217738713543747, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 10557 + }, + { + "epoch": 4.218138234119057, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10558 + }, + { + "epoch": 4.218537754694367, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10559 + }, + { + "epoch": 4.218937275269677, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10560 + }, + { + "epoch": 4.219336795844986, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10561 + }, + { + "epoch": 4.2197363164202955, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10562 + }, + { + "epoch": 4.220135836995605, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10563 + }, + { + "epoch": 4.220535357570915, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10564 + }, + { + "epoch": 4.220934878146225, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10565 + }, + { + "epoch": 4.221334398721535, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10566 + }, + { + "epoch": 4.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10567 + }, + { + "epoch": 4.222133439872153, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 10568 + }, + { + "epoch": 4.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10569 + }, + { + "epoch": 4.222932481022773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 10570 + }, + { + "epoch": 4.223332001598083, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10571 + }, + { + "epoch": 4.2237315221733915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10572 + }, + { + "epoch": 4.224131042748701, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10573 + }, + { + "epoch": 4.224530563324011, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 10574 + }, + { + "epoch": 4.224930083899321, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10575 + }, + { + "epoch": 4.225329604474631, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10576 + }, + { + "epoch": 4.2257291250499405, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 10577 + }, + { + "epoch": 4.226128645625249, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3502, + "step": 10578 + }, + { + "epoch": 4.226528166200559, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10579 + }, + { + "epoch": 4.226927686775869, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10580 + }, + { + "epoch": 4.227327207351179, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10581 + }, + { + "epoch": 4.2277267279264885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10582 + }, + { + "epoch": 4.228126248501797, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10583 + }, + { + "epoch": 4.228525769077107, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10584 + }, + { + "epoch": 4.228925289652417, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10585 + }, + { + "epoch": 4.229324810227727, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 10586 + }, + { + "epoch": 4.229724330803037, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 10587 + }, + { + "epoch": 4.230123851378346, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10588 + }, + { + "epoch": 4.230523371953655, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10589 + }, + { + "epoch": 4.230922892528965, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10590 + }, + { + "epoch": 4.231322413104275, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10591 + }, + { + "epoch": 4.231721933679585, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 10592 + }, + { + "epoch": 4.232121454254894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10593 + }, + { + "epoch": 4.232520974830204, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10594 + }, + { + "epoch": 4.232920495405513, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10595 + }, + { + "epoch": 4.233320015980823, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10596 + }, + { + "epoch": 4.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10597 + }, + { + "epoch": 4.234119057131442, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10598 + }, + { + "epoch": 4.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10599 + }, + { + "epoch": 4.234918098282061, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10600 + }, + { + "epoch": 4.235317618857371, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 10601 + }, + { + "epoch": 4.235717139432681, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10602 + }, + { + "epoch": 4.2361166600079905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10603 + }, + { + "epoch": 4.2365161805833, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10604 + }, + { + "epoch": 4.23691570115861, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 10605 + }, + { + "epoch": 4.237315221733919, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10606 + }, + { + "epoch": 4.237714742309229, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10607 + }, + { + "epoch": 4.2381142628845385, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10608 + }, + { + "epoch": 4.238513783459848, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10609 + }, + { + "epoch": 4.238913304035158, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10610 + }, + { + "epoch": 4.239312824610467, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10611 + }, + { + "epoch": 4.239712345185777, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 10612 + }, + { + "epoch": 4.240111865761087, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10613 + }, + { + "epoch": 4.240511386336396, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10614 + }, + { + "epoch": 4.240910906911706, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3343, + "step": 10615 + }, + { + "epoch": 4.241310427487016, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10616 + }, + { + "epoch": 4.241709948062325, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 10617 + }, + { + "epoch": 4.242109468637635, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10618 + }, + { + "epoch": 4.242508989212944, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10619 + }, + { + "epoch": 4.242908509788254, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 10620 + }, + { + "epoch": 4.243308030363564, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 10621 + }, + { + "epoch": 4.243707550938874, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 10622 + }, + { + "epoch": 4.244107071514183, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 10623 + }, + { + "epoch": 4.244506592089492, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10624 + }, + { + "epoch": 4.244906112664802, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10625 + }, + { + "epoch": 4.245305633240112, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10626 + }, + { + "epoch": 4.245705153815422, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10627 + }, + { + "epoch": 4.246104674390731, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 10628 + }, + { + "epoch": 4.2465041949660405, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 10629 + }, + { + "epoch": 4.24690371554135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10630 + }, + { + "epoch": 4.24730323611666, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10631 + }, + { + "epoch": 4.24770275669197, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10632 + }, + { + "epoch": 4.24810227726728, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 10633 + }, + { + "epoch": 4.2485017978425885, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10634 + }, + { + "epoch": 4.248901318417898, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10635 + }, + { + "epoch": 4.249300838993208, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10636 + }, + { + "epoch": 4.249700359568518, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10637 + }, + { + "epoch": 4.250099880143828, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10638 + }, + { + "epoch": 4.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10639 + }, + { + "epoch": 4.250898921294446, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10640 + }, + { + "epoch": 4.251298441869756, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10641 + }, + { + "epoch": 4.251697962445066, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10642 + }, + { + "epoch": 4.252097483020376, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10643 + }, + { + "epoch": 4.2524970035956855, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10644 + }, + { + "epoch": 4.252896524170994, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10645 + }, + { + "epoch": 4.253296044746304, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10646 + }, + { + "epoch": 4.253695565321614, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10647 + }, + { + "epoch": 4.254095085896924, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10648 + }, + { + "epoch": 4.2544946064722335, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 10649 + }, + { + "epoch": 4.254894127047543, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 10650 + }, + { + "epoch": 4.255293647622852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10651 + }, + { + "epoch": 4.255693168198162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10652 + }, + { + "epoch": 4.256092688773472, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 10653 + }, + { + "epoch": 4.256492209348782, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10654 + }, + { + "epoch": 4.256891729924091, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10655 + }, + { + "epoch": 4.257291250499401, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10656 + }, + { + "epoch": 4.25769077107471, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 10657 + }, + { + "epoch": 4.25809029165002, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 10658 + }, + { + "epoch": 4.25848981222533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10659 + }, + { + "epoch": 4.258889332800639, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10660 + }, + { + "epoch": 4.259288853375949, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10661 + }, + { + "epoch": 4.259688373951258, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 10662 + }, + { + "epoch": 4.260087894526568, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10663 + }, + { + "epoch": 4.260487415101878, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10664 + }, + { + "epoch": 4.2608869356771875, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10665 + }, + { + "epoch": 4.261286456252497, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10666 + }, + { + "epoch": 4.261685976827807, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 10667 + }, + { + "epoch": 4.262085497403116, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10668 + }, + { + "epoch": 4.262485017978426, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10669 + }, + { + "epoch": 4.2628845385537355, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 10670 + }, + { + "epoch": 4.263284059129045, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 10671 + }, + { + "epoch": 4.263683579704355, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10672 + }, + { + "epoch": 4.264083100279665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 10673 + }, + { + "epoch": 4.264482620854974, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10674 + }, + { + "epoch": 4.2648821414302835, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10675 + }, + { + "epoch": 4.265281662005593, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10676 + }, + { + "epoch": 4.265681182580903, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10677 + }, + { + "epoch": 4.266080703156213, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10678 + }, + { + "epoch": 4.266480223731522, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10679 + }, + { + "epoch": 4.266879744306832, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 10680 + }, + { + "epoch": 4.267279264882141, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 10681 + }, + { + "epoch": 4.267678785457451, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10682 + }, + { + "epoch": 4.268078306032761, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3352, + "step": 10683 + }, + { + "epoch": 4.26847782660807, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 10684 + }, + { + "epoch": 4.26887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 10685 + }, + { + "epoch": 4.269276867758689, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 10686 + }, + { + "epoch": 4.269676388333999, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 10687 + }, + { + "epoch": 4.270075908909309, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10688 + }, + { + "epoch": 4.270475429484619, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10689 + }, + { + "epoch": 4.270874950059928, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10690 + }, + { + "epoch": 4.2712744706352375, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10691 + }, + { + "epoch": 4.271673991210547, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 10692 + }, + { + "epoch": 4.272073511785857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10693 + }, + { + "epoch": 4.272473032361167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 10694 + }, + { + "epoch": 4.272872552936477, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10695 + }, + { + "epoch": 4.2732720735117855, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10696 + }, + { + "epoch": 4.273671594087095, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10697 + }, + { + "epoch": 4.274071114662405, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10698 + }, + { + "epoch": 4.274470635237715, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 10699 + }, + { + "epoch": 4.274870155813025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10700 + }, + { + "epoch": 4.2752696763883336, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10701 + }, + { + "epoch": 4.275669196963643, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 10702 + }, + { + "epoch": 4.276068717538953, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10703 + }, + { + "epoch": 4.276468238114263, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 10704 + }, + { + "epoch": 4.276867758689573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 10705 + }, + { + "epoch": 4.2772672792648825, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10706 + }, + { + "epoch": 4.277666799840191, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10707 + }, + { + "epoch": 4.278066320415501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10708 + }, + { + "epoch": 4.278465840990811, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10709 + }, + { + "epoch": 4.278865361566121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10710 + }, + { + "epoch": 4.2792648821414305, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10711 + }, + { + "epoch": 4.27966440271674, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 10712 + }, + { + "epoch": 4.280063923292049, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10713 + }, + { + "epoch": 4.280463443867359, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 10714 + }, + { + "epoch": 4.280862964442669, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10715 + }, + { + "epoch": 4.281262485017979, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10716 + }, + { + "epoch": 4.281662005593288, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10717 + }, + { + "epoch": 4.282061526168597, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10718 + }, + { + "epoch": 4.282461046743907, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10719 + }, + { + "epoch": 4.282860567319217, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 10720 + }, + { + "epoch": 4.283260087894527, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 10721 + }, + { + "epoch": 4.283659608469836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10722 + }, + { + "epoch": 4.284059129045146, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10723 + }, + { + "epoch": 4.284458649620455, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10724 + }, + { + "epoch": 4.284858170195765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 10725 + }, + { + "epoch": 4.285257690771075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10726 + }, + { + "epoch": 4.285657211346384, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 10727 + }, + { + "epoch": 4.286056731921694, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10728 + }, + { + "epoch": 4.286456252497004, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10729 + }, + { + "epoch": 4.286855773072313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10730 + }, + { + "epoch": 4.287255293647623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10731 + }, + { + "epoch": 4.2876548142229325, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10732 + }, + { + "epoch": 4.288054334798242, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10733 + }, + { + "epoch": 4.288453855373552, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 10734 + }, + { + "epoch": 4.288853375948861, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10735 + }, + { + "epoch": 4.289252896524171, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 10736 + }, + { + "epoch": 4.2896524170994805, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10737 + }, + { + "epoch": 4.29005193767479, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 10738 + }, + { + "epoch": 4.2904514582501, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 10739 + }, + { + "epoch": 4.29085097882541, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 10740 + }, + { + "epoch": 4.291250499400719, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 10741 + }, + { + "epoch": 4.291650019976029, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 10742 + }, + { + "epoch": 4.292049540551338, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10743 + }, + { + "epoch": 4.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10744 + }, + { + "epoch": 4.292848581701958, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 10745 + }, + { + "epoch": 4.293248102277268, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10746 + }, + { + "epoch": 4.293647622852577, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 10747 + }, + { + "epoch": 4.294047143427886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10748 + }, + { + "epoch": 4.294446664003196, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10749 + }, + { + "epoch": 4.294846184578506, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10750 + }, + { + "epoch": 4.295245705153816, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 10751 + }, + { + "epoch": 4.295645225729125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3386, + "step": 10752 + }, + { + "epoch": 4.2960447463044344, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 10753 + }, + { + "epoch": 4.296444266879744, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 10754 + }, + { + "epoch": 4.296843787455054, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10755 + }, + { + "epoch": 4.297243308030364, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 10756 + }, + { + "epoch": 4.297642828605674, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 10757 + }, + { + "epoch": 4.2980423491809825, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10758 + }, + { + "epoch": 4.298441869756292, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 10759 + }, + { + "epoch": 4.298841390331602, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 10760 + }, + { + "epoch": 4.299240910906912, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10761 + }, + { + "epoch": 4.299640431482222, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 10762 + }, + { + "epoch": 4.300039952057531, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10763 + }, + { + "epoch": 4.30043947263284, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 10764 + }, + { + "epoch": 4.30083899320815, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10765 + }, + { + "epoch": 4.30123851378346, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 10766 + }, + { + "epoch": 4.30163803435877, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 10767 + }, + { + "epoch": 4.3020375549340795, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 10768 + }, + { + "epoch": 4.302437075509388, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10769 + }, + { + "epoch": 4.302836596084698, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10770 + }, + { + "epoch": 4.303236116660008, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10771 + }, + { + "epoch": 4.303635637235318, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10772 + }, + { + "epoch": 4.3040351578106275, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10773 + }, + { + "epoch": 4.304434678385936, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10774 + }, + { + "epoch": 4.304834198961246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10775 + }, + { + "epoch": 4.305233719536556, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10776 + }, + { + "epoch": 4.305633240111866, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10777 + }, + { + "epoch": 4.3060327606871756, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10778 + }, + { + "epoch": 4.306432281262485, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10779 + }, + { + "epoch": 4.306831801837794, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10780 + }, + { + "epoch": 4.307231322413104, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 10781 + }, + { + "epoch": 4.307630842988414, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 10782 + }, + { + "epoch": 4.308030363563724, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10783 + }, + { + "epoch": 4.308429884139033, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10784 + }, + { + "epoch": 4.308829404714343, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 10785 + }, + { + "epoch": 4.309228925289652, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10786 + }, + { + "epoch": 4.309628445864962, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 10787 + }, + { + "epoch": 4.310027966440272, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10788 + }, + { + "epoch": 4.310427487015581, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10789 + }, + { + "epoch": 4.310827007590891, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 10790 + }, + { + "epoch": 4.3112265281662, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10791 + }, + { + "epoch": 4.31162604874151, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 10792 + }, + { + "epoch": 4.31202556931682, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10793 + }, + { + "epoch": 4.3124250898921295, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 10794 + }, + { + "epoch": 4.312824610467439, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10795 + }, + { + "epoch": 4.313224131042749, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10796 + }, + { + "epoch": 4.313623651618058, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 10797 + }, + { + "epoch": 4.314023172193368, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10798 + }, + { + "epoch": 4.3144226927686775, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 10799 + }, + { + "epoch": 4.314822213343987, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 10800 + }, + { + "epoch": 4.315221733919297, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10801 + }, + { + "epoch": 4.315621254494607, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 10802 + }, + { + "epoch": 4.316020775069916, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10803 + }, + { + "epoch": 4.316420295645226, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 10804 + }, + { + "epoch": 4.316819816220535, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 10805 + }, + { + "epoch": 4.317219336795845, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10806 + }, + { + "epoch": 4.317618857371155, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10807 + }, + { + "epoch": 4.318018377946464, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10808 + }, + { + "epoch": 4.318417898521774, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 10809 + }, + { + "epoch": 4.318817419097083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 10810 + }, + { + "epoch": 4.319216939672393, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10811 + }, + { + "epoch": 4.319616460247703, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10812 + }, + { + "epoch": 4.320015980823013, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 10813 + }, + { + "epoch": 4.320415501398322, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10814 + }, + { + "epoch": 4.320815021973631, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10815 + }, + { + "epoch": 4.321214542548941, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 10816 + }, + { + "epoch": 4.321614063124251, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10817 + }, + { + "epoch": 4.322013583699561, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 10818 + }, + { + "epoch": 4.322413104274871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10819 + }, + { + "epoch": 4.3228126248501795, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 10820 + }, + { + "epoch": 4.323212145425489, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 10821 + }, + { + "epoch": 4.323611666000799, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 10822 + }, + { + "epoch": 4.324011186576109, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 10823 + }, + { + "epoch": 4.324410707151419, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 10824 + }, + { + "epoch": 4.3248102277267275, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10825 + }, + { + "epoch": 4.325209748302037, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 10826 + }, + { + "epoch": 4.325609268877347, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10827 + }, + { + "epoch": 4.326008789452657, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10828 + }, + { + "epoch": 4.326408310027967, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10829 + }, + { + "epoch": 4.3268078306032765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10830 + }, + { + "epoch": 4.327207351178585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 10831 + }, + { + "epoch": 4.327606871753895, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 10832 + }, + { + "epoch": 4.328006392329205, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 10833 + }, + { + "epoch": 4.328405912904515, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 10834 + }, + { + "epoch": 4.3288054334798245, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 10835 + }, + { + "epoch": 4.329204954055134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 10836 + }, + { + "epoch": 4.329604474630443, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10837 + }, + { + "epoch": 4.330003995205753, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 10838 + }, + { + "epoch": 4.330403515781063, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10839 + }, + { + "epoch": 4.3308030363563725, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10840 + }, + { + "epoch": 4.331202556931682, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 10841 + }, + { + "epoch": 4.331602077506991, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10842 + }, + { + "epoch": 4.332001598082301, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 10843 + }, + { + "epoch": 4.332401118657611, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 10844 + }, + { + "epoch": 4.332800639232921, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10845 + }, + { + "epoch": 4.33320015980823, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 10846 + }, + { + "epoch": 4.33359968038354, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 10847 + }, + { + "epoch": 4.333999200958849, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10848 + }, + { + "epoch": 4.334398721534159, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 10849 + }, + { + "epoch": 4.334798242109469, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 10850 + }, + { + "epoch": 4.335197762684778, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10851 + }, + { + "epoch": 4.335597283260088, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 10852 + }, + { + "epoch": 4.335996803835398, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 10853 + }, + { + "epoch": 4.336396324410707, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 10854 + }, + { + "epoch": 4.336795844986017, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10855 + }, + { + "epoch": 4.3371953655613265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10856 + }, + { + "epoch": 4.337594886136636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10857 + }, + { + "epoch": 4.337994406711946, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 10858 + }, + { + "epoch": 4.338393927287255, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 10859 + }, + { + "epoch": 4.338793447862565, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 10860 + }, + { + "epoch": 4.3391929684378745, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 10861 + }, + { + "epoch": 4.339592489013184, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 10862 + }, + { + "epoch": 4.339992009588494, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 10863 + }, + { + "epoch": 4.340391530163803, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 10864 + }, + { + "epoch": 4.340791050739113, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 10865 + }, + { + "epoch": 4.3411905713144225, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 10866 + }, + { + "epoch": 4.341590091889732, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10867 + }, + { + "epoch": 4.341989612465042, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10868 + }, + { + "epoch": 4.342389133040352, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10869 + }, + { + "epoch": 4.342788653615661, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 10870 + }, + { + "epoch": 4.343188174190971, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 10871 + }, + { + "epoch": 4.34358769476628, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3359, + "step": 10872 + }, + { + "epoch": 4.34398721534159, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 10873 + }, + { + "epoch": 4.3443867359169, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 10874 + }, + { + "epoch": 4.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10875 + }, + { + "epoch": 4.345185777067519, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 10876 + }, + { + "epoch": 4.345585297642828, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10877 + }, + { + "epoch": 4.345984818218138, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 10878 + }, + { + "epoch": 4.346384338793448, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 10879 + }, + { + "epoch": 4.346783859368758, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10880 + }, + { + "epoch": 4.347183379944067, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10881 + }, + { + "epoch": 4.3475829005193765, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 10882 + }, + { + "epoch": 4.347982421094686, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 10883 + }, + { + "epoch": 4.348381941669996, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 10884 + }, + { + "epoch": 4.348781462245306, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 10885 + }, + { + "epoch": 4.349180982820616, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 10886 + }, + { + "epoch": 4.3495805033959245, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 10887 + }, + { + "epoch": 4.349980023971234, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10888 + }, + { + "epoch": 4.350379544546544, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 10889 + }, + { + "epoch": 4.350779065121854, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10890 + }, + { + "epoch": 4.351178585697164, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 10891 + }, + { + "epoch": 4.351578106272473, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10892 + }, + { + "epoch": 4.351977626847782, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3377, + "step": 10893 + }, + { + "epoch": 4.352377147423092, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 10894 + }, + { + "epoch": 4.352776667998402, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 10895 + }, + { + "epoch": 4.353176188573712, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 10896 + }, + { + "epoch": 4.3535757091490215, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 10897 + }, + { + "epoch": 4.35397522972433, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 10898 + }, + { + "epoch": 4.35437475029964, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 10899 + }, + { + "epoch": 4.35477427087495, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10900 + }, + { + "epoch": 4.35517379145026, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10901 + }, + { + "epoch": 4.3555733120255695, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 10902 + }, + { + "epoch": 4.355972832600879, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10903 + }, + { + "epoch": 4.356372353176188, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 10904 + }, + { + "epoch": 4.356771873751498, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 10905 + }, + { + "epoch": 4.357171394326808, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 10906 + }, + { + "epoch": 4.357570914902118, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10907 + }, + { + "epoch": 4.357970435477427, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10908 + }, + { + "epoch": 4.358369956052737, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 10909 + }, + { + "epoch": 4.358769476628046, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 10910 + }, + { + "epoch": 4.359168997203356, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 10911 + }, + { + "epoch": 4.359568517778666, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10912 + }, + { + "epoch": 4.359968038353975, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10913 + }, + { + "epoch": 4.360367558929285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10914 + }, + { + "epoch": 4.360767079504594, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 10915 + }, + { + "epoch": 4.361166600079904, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10916 + }, + { + "epoch": 4.361566120655214, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10917 + }, + { + "epoch": 4.361965641230523, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 10918 + }, + { + "epoch": 4.362365161805833, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10919 + }, + { + "epoch": 4.362764682381143, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10920 + }, + { + "epoch": 4.363164202956452, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10921 + }, + { + "epoch": 4.363563723531762, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 10922 + }, + { + "epoch": 4.3639632441070715, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 10923 + }, + { + "epoch": 4.364362764682381, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 10924 + }, + { + "epoch": 4.364762285257691, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 10925 + }, + { + "epoch": 4.365161805833001, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 10926 + }, + { + "epoch": 4.36556132640831, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10927 + }, + { + "epoch": 4.3659608469836195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 10928 + }, + { + "epoch": 4.366360367558929, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10929 + }, + { + "epoch": 4.366759888134239, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10930 + }, + { + "epoch": 4.367159408709549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 10931 + }, + { + "epoch": 4.367558929284858, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 10932 + }, + { + "epoch": 4.367958449860168, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10933 + }, + { + "epoch": 4.368357970435477, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 10934 + }, + { + "epoch": 4.368757491010787, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 10935 + }, + { + "epoch": 4.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 10936 + }, + { + "epoch": 4.369556532161407, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 10937 + }, + { + "epoch": 4.369956052736716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 10938 + }, + { + "epoch": 4.370355573312025, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10939 + }, + { + "epoch": 4.370755093887335, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 10940 + }, + { + "epoch": 4.371154614462645, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 10941 + }, + { + "epoch": 4.371554135037955, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 10942 + }, + { + "epoch": 4.3719536556132645, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 10943 + }, + { + "epoch": 4.372353176188573, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 10944 + }, + { + "epoch": 4.372752696763883, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 10945 + }, + { + "epoch": 4.373152217339193, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 10946 + }, + { + "epoch": 4.373551737914503, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 10947 + }, + { + "epoch": 4.373951258489813, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 10948 + }, + { + "epoch": 4.3743507790651215, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 10949 + }, + { + "epoch": 4.374750299640431, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 10950 + }, + { + "epoch": 4.375149820215741, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 10951 + }, + { + "epoch": 4.375549340791051, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10952 + }, + { + "epoch": 4.375948861366361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 10953 + }, + { + "epoch": 4.3763483819416695, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.34, + "step": 10954 + }, + { + "epoch": 4.376747902516979, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 10955 + }, + { + "epoch": 4.377147423092289, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 10956 + }, + { + "epoch": 4.377546943667599, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 10957 + }, + { + "epoch": 4.377946464242909, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10958 + }, + { + "epoch": 4.3783459848182185, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 10959 + }, + { + "epoch": 4.378745505393527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 10960 + }, + { + "epoch": 4.379145025968837, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 10961 + }, + { + "epoch": 4.379544546544147, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 10962 + }, + { + "epoch": 4.379944067119457, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 10963 + }, + { + "epoch": 4.3803435876947665, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 10964 + }, + { + "epoch": 4.380743108270076, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 10965 + }, + { + "epoch": 4.381142628845385, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10966 + }, + { + "epoch": 4.381542149420695, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 10967 + }, + { + "epoch": 4.381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10968 + }, + { + "epoch": 4.3823411905713145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 10969 + }, + { + "epoch": 4.382740711146624, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 10970 + }, + { + "epoch": 4.383140231721933, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 10971 + }, + { + "epoch": 4.383539752297243, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 10972 + }, + { + "epoch": 4.383939272872553, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 10973 + }, + { + "epoch": 4.384338793447863, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 10974 + }, + { + "epoch": 4.384738314023172, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 10975 + }, + { + "epoch": 4.385137834598482, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 10976 + }, + { + "epoch": 4.385537355173791, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 10977 + }, + { + "epoch": 4.385936875749101, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 10978 + }, + { + "epoch": 4.386336396324411, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 10979 + }, + { + "epoch": 4.38673591689972, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 10980 + }, + { + "epoch": 4.38713543747503, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10981 + }, + { + "epoch": 4.38753495805034, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 10982 + }, + { + "epoch": 4.387934478625649, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 10983 + }, + { + "epoch": 4.388333999200959, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 10984 + }, + { + "epoch": 4.3887335197762685, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10985 + }, + { + "epoch": 4.389133040351578, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 10986 + }, + { + "epoch": 4.389532560926888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 10987 + }, + { + "epoch": 4.389932081502197, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 10988 + }, + { + "epoch": 4.390331602077507, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 10989 + }, + { + "epoch": 4.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 10990 + }, + { + "epoch": 4.391130643228126, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 10991 + }, + { + "epoch": 4.391530163803436, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 10992 + }, + { + "epoch": 4.391929684378746, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 10993 + }, + { + "epoch": 4.392329204954055, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 10994 + }, + { + "epoch": 4.3927287255293646, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 10995 + }, + { + "epoch": 4.393128246104674, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 10996 + }, + { + "epoch": 4.393527766679984, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 10997 + }, + { + "epoch": 4.393927287255294, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 10998 + }, + { + "epoch": 4.394326807830604, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 10999 + }, + { + "epoch": 4.394726328405913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11000 + }, + { + "epoch": 4.395125848981222, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11001 + }, + { + "epoch": 4.395525369556532, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11002 + }, + { + "epoch": 4.395924890131842, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 11003 + }, + { + "epoch": 4.396324410707152, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11004 + }, + { + "epoch": 4.396723931282461, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11005 + }, + { + "epoch": 4.39712345185777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11006 + }, + { + "epoch": 4.39752297243308, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11007 + }, + { + "epoch": 4.39792249300839, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11008 + }, + { + "epoch": 4.3983220135837, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11009 + }, + { + "epoch": 4.39872153415901, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11010 + }, + { + "epoch": 4.3991210547343185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11011 + }, + { + "epoch": 4.399520575309628, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11012 + }, + { + "epoch": 4.399920095884938, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11013 + }, + { + "epoch": 4.400319616460248, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11014 + }, + { + "epoch": 4.400719137035558, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11015 + }, + { + "epoch": 4.401118657610867, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11016 + }, + { + "epoch": 4.401518178186176, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11017 + }, + { + "epoch": 4.401917698761486, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11018 + }, + { + "epoch": 4.402317219336796, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11019 + }, + { + "epoch": 4.402716739912106, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11020 + }, + { + "epoch": 4.4031162604874154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11021 + }, + { + "epoch": 4.403515781062724, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11022 + }, + { + "epoch": 4.403915301638034, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11023 + }, + { + "epoch": 4.404314822213344, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11024 + }, + { + "epoch": 4.404714342788654, + "grad_norm": 0.11669921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11025 + }, + { + "epoch": 4.4051138633639635, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11026 + }, + { + "epoch": 4.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11027 + }, + { + "epoch": 4.405912904514582, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11028 + }, + { + "epoch": 4.406312425089892, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11029 + }, + { + "epoch": 4.406711945665202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11030 + }, + { + "epoch": 4.4071114662405115, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11031 + }, + { + "epoch": 4.407510986815821, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11032 + }, + { + "epoch": 4.407910507391131, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 11033 + }, + { + "epoch": 4.40831002796644, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11034 + }, + { + "epoch": 4.40870954854175, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11035 + }, + { + "epoch": 4.40910906911706, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11036 + }, + { + "epoch": 4.409508589692369, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11037 + }, + { + "epoch": 4.409908110267679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11038 + }, + { + "epoch": 4.410307630842988, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11039 + }, + { + "epoch": 4.410707151418298, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11040 + }, + { + "epoch": 4.411106671993608, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11041 + }, + { + "epoch": 4.411506192568917, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11042 + }, + { + "epoch": 4.411905713144227, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11043 + }, + { + "epoch": 4.412305233719536, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11044 + }, + { + "epoch": 4.412704754294846, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11045 + }, + { + "epoch": 4.413104274870156, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11046 + }, + { + "epoch": 4.4135037954454654, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11047 + }, + { + "epoch": 4.413903316020775, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11048 + }, + { + "epoch": 4.414302836596085, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11049 + }, + { + "epoch": 4.414702357171394, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11050 + }, + { + "epoch": 4.415101877746704, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11051 + }, + { + "epoch": 4.4155013983220135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11052 + }, + { + "epoch": 4.415900918897323, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11053 + }, + { + "epoch": 4.416300439472633, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11054 + }, + { + "epoch": 4.416699960047943, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11055 + }, + { + "epoch": 4.417099480623252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11056 + }, + { + "epoch": 4.4174990011985615, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11057 + }, + { + "epoch": 4.417898521773871, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11058 + }, + { + "epoch": 4.418298042349181, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11059 + }, + { + "epoch": 4.418697562924491, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11060 + }, + { + "epoch": 4.4190970834998, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11061 + }, + { + "epoch": 4.41949660407511, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11062 + }, + { + "epoch": 4.419896124650419, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11063 + }, + { + "epoch": 4.420295645225729, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11064 + }, + { + "epoch": 4.420695165801039, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11065 + }, + { + "epoch": 4.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11066 + }, + { + "epoch": 4.421494206951658, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11067 + }, + { + "epoch": 4.421893727526967, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11068 + }, + { + "epoch": 4.422293248102277, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11069 + }, + { + "epoch": 4.422692768677587, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11070 + }, + { + "epoch": 4.423092289252897, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11071 + }, + { + "epoch": 4.4234918098282066, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11072 + }, + { + "epoch": 4.4238913304035155, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11073 + }, + { + "epoch": 4.424290850978825, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11074 + }, + { + "epoch": 4.424690371554135, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11075 + }, + { + "epoch": 4.425089892129445, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11076 + }, + { + "epoch": 4.425489412704755, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11077 + }, + { + "epoch": 4.4258889332800635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11078 + }, + { + "epoch": 4.426288453855373, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11079 + }, + { + "epoch": 4.426687974430683, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11080 + }, + { + "epoch": 4.427087495005993, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11081 + }, + { + "epoch": 4.427487015581303, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11082 + }, + { + "epoch": 4.427886536156612, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11083 + }, + { + "epoch": 4.428286056731921, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11084 + }, + { + "epoch": 4.428685577307231, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11085 + }, + { + "epoch": 4.429085097882541, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11086 + }, + { + "epoch": 4.429484618457851, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11087 + }, + { + "epoch": 4.4298841390331605, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 11088 + }, + { + "epoch": 4.43028365960847, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11089 + }, + { + "epoch": 4.430683180183779, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11090 + }, + { + "epoch": 4.431082700759089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11091 + }, + { + "epoch": 4.431482221334399, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 11092 + }, + { + "epoch": 4.4318817419097085, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11093 + }, + { + "epoch": 4.432281262485018, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11094 + }, + { + "epoch": 4.432680783060327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11095 + }, + { + "epoch": 4.433080303635637, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11096 + }, + { + "epoch": 4.433479824210947, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11097 + }, + { + "epoch": 4.433879344786257, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 11098 + }, + { + "epoch": 4.434278865361566, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11099 + }, + { + "epoch": 4.434678385936876, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11100 + }, + { + "epoch": 4.435077906512185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11101 + }, + { + "epoch": 4.435477427087495, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11102 + }, + { + "epoch": 4.435876947662805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11103 + }, + { + "epoch": 4.436276468238114, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11104 + }, + { + "epoch": 4.436675988813424, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11105 + }, + { + "epoch": 4.437075509388734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11106 + }, + { + "epoch": 4.437475029964043, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 11107 + }, + { + "epoch": 4.437874550539353, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 11108 + }, + { + "epoch": 4.438274071114662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11109 + }, + { + "epoch": 4.438673591689972, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11110 + }, + { + "epoch": 4.439073112265282, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11111 + }, + { + "epoch": 4.439472632840591, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11112 + }, + { + "epoch": 4.439872153415901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11113 + }, + { + "epoch": 4.4402716739912105, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11114 + }, + { + "epoch": 4.44067119456652, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11115 + }, + { + "epoch": 4.44107071514183, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11116 + }, + { + "epoch": 4.44147023571714, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11117 + }, + { + "epoch": 4.441869756292449, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11118 + }, + { + "epoch": 4.4422692768677585, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 11119 + }, + { + "epoch": 4.442668797443068, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11120 + }, + { + "epoch": 4.443068318018378, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11121 + }, + { + "epoch": 4.443467838593688, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11122 + }, + { + "epoch": 4.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11123 + }, + { + "epoch": 4.444266879744307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11124 + }, + { + "epoch": 4.444666400319616, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11125 + }, + { + "epoch": 4.445065920894926, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11126 + }, + { + "epoch": 4.445465441470236, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11127 + }, + { + "epoch": 4.445864962045546, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11128 + }, + { + "epoch": 4.446264482620855, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11129 + }, + { + "epoch": 4.446664003196164, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11130 + }, + { + "epoch": 4.447063523771474, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11131 + }, + { + "epoch": 4.447463044346784, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11132 + }, + { + "epoch": 4.447862564922094, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11133 + }, + { + "epoch": 4.448262085497403, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11134 + }, + { + "epoch": 4.448661606072712, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11135 + }, + { + "epoch": 4.449061126648022, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11136 + }, + { + "epoch": 4.449460647223332, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 11137 + }, + { + "epoch": 4.449860167798642, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11138 + }, + { + "epoch": 4.450259688373952, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11139 + }, + { + "epoch": 4.4506592089492605, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 11140 + }, + { + "epoch": 4.45105872952457, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11141 + }, + { + "epoch": 4.45145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11142 + }, + { + "epoch": 4.45185777067519, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11143 + }, + { + "epoch": 4.4522572912505, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11144 + }, + { + "epoch": 4.452656811825809, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11145 + }, + { + "epoch": 4.453056332401118, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11146 + }, + { + "epoch": 4.453455852976428, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11147 + }, + { + "epoch": 4.453855373551738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11148 + }, + { + "epoch": 4.454254894127048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11149 + }, + { + "epoch": 4.4546544147023575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11150 + }, + { + "epoch": 4.455053935277666, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11151 + }, + { + "epoch": 4.455453455852976, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11152 + }, + { + "epoch": 4.455852976428286, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11153 + }, + { + "epoch": 4.456252497003596, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11154 + }, + { + "epoch": 4.4566520175789055, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11155 + }, + { + "epoch": 4.457051538154215, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11156 + }, + { + "epoch": 4.457451058729524, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11157 + }, + { + "epoch": 4.457850579304834, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11158 + }, + { + "epoch": 4.458250099880144, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11159 + }, + { + "epoch": 4.4586496204554535, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11160 + }, + { + "epoch": 4.459049141030763, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11161 + }, + { + "epoch": 4.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11162 + }, + { + "epoch": 4.459848182181382, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11163 + }, + { + "epoch": 4.460247702756692, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11164 + }, + { + "epoch": 4.460647223332002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11165 + }, + { + "epoch": 4.461046743907311, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11166 + }, + { + "epoch": 4.461446264482621, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11167 + }, + { + "epoch": 4.46184578505793, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11168 + }, + { + "epoch": 4.46224530563324, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11169 + }, + { + "epoch": 4.46264482620855, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11170 + }, + { + "epoch": 4.463044346783859, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11171 + }, + { + "epoch": 4.463443867359169, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 11172 + }, + { + "epoch": 4.463843387934479, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11173 + }, + { + "epoch": 4.464242908509788, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11174 + }, + { + "epoch": 4.464642429085098, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11175 + }, + { + "epoch": 4.4650419496604075, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11176 + }, + { + "epoch": 4.465441470235717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11177 + }, + { + "epoch": 4.465840990811027, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11178 + }, + { + "epoch": 4.466240511386337, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11179 + }, + { + "epoch": 4.466640031961646, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11180 + }, + { + "epoch": 4.4670395525369555, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11181 + }, + { + "epoch": 4.467439073112265, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11182 + }, + { + "epoch": 4.467838593687575, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11183 + }, + { + "epoch": 4.468238114262885, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 11184 + }, + { + "epoch": 4.468637634838194, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 11185 + }, + { + "epoch": 4.4690371554135035, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11186 + }, + { + "epoch": 4.469436675988813, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11187 + }, + { + "epoch": 4.469836196564123, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11188 + }, + { + "epoch": 4.470235717139433, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11189 + }, + { + "epoch": 4.470635237714743, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 11190 + }, + { + "epoch": 4.471034758290052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11191 + }, + { + "epoch": 4.471434278865361, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11192 + }, + { + "epoch": 4.471833799440671, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 11193 + }, + { + "epoch": 4.472233320015981, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11194 + }, + { + "epoch": 4.472632840591291, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11195 + }, + { + "epoch": 4.4730323611666005, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11196 + }, + { + "epoch": 4.473431881741909, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 11197 + }, + { + "epoch": 4.473831402317219, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11198 + }, + { + "epoch": 4.474230922892529, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11199 + }, + { + "epoch": 4.474630443467839, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 11200 + }, + { + "epoch": 4.475029964043149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11201 + }, + { + "epoch": 4.4754294846184575, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 11202 + }, + { + "epoch": 4.475829005193767, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 11203 + }, + { + "epoch": 4.476228525769077, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11204 + }, + { + "epoch": 4.476628046344387, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 11205 + }, + { + "epoch": 4.477027566919697, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11206 + }, + { + "epoch": 4.477427087495006, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11207 + }, + { + "epoch": 4.477826608070315, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11208 + }, + { + "epoch": 4.478226128645625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11209 + }, + { + "epoch": 4.478625649220935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11210 + }, + { + "epoch": 4.479025169796245, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11211 + }, + { + "epoch": 4.479424690371554, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11212 + }, + { + "epoch": 4.479824210946864, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11213 + }, + { + "epoch": 4.480223731522173, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11214 + }, + { + "epoch": 4.480623252097483, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11215 + }, + { + "epoch": 4.481022772672793, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11216 + }, + { + "epoch": 4.4814222932481025, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11217 + }, + { + "epoch": 4.481821813823412, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11218 + }, + { + "epoch": 4.482221334398721, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11219 + }, + { + "epoch": 4.482620854974031, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11220 + }, + { + "epoch": 4.483020375549341, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11221 + }, + { + "epoch": 4.4834198961246505, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11222 + }, + { + "epoch": 4.48381941669996, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11223 + }, + { + "epoch": 4.484218937275269, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11224 + }, + { + "epoch": 4.484618457850579, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11225 + }, + { + "epoch": 4.485017978425889, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11226 + }, + { + "epoch": 4.485417499001199, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11227 + }, + { + "epoch": 4.485817019576508, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11228 + }, + { + "epoch": 4.486216540151818, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11229 + }, + { + "epoch": 4.486616060727127, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11230 + }, + { + "epoch": 4.487015581302437, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11231 + }, + { + "epoch": 4.487415101877747, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11232 + }, + { + "epoch": 4.487814622453056, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11233 + }, + { + "epoch": 4.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11234 + }, + { + "epoch": 4.488613663603676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11235 + }, + { + "epoch": 4.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11236 + }, + { + "epoch": 4.489412704754295, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11237 + }, + { + "epoch": 4.4898122253296044, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11238 + }, + { + "epoch": 4.490211745904914, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11239 + }, + { + "epoch": 4.490611266480224, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11240 + }, + { + "epoch": 4.491010787055533, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11241 + }, + { + "epoch": 4.491410307630843, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 11242 + }, + { + "epoch": 4.4918098282061525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11243 + }, + { + "epoch": 4.492209348781462, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11244 + }, + { + "epoch": 4.492608869356772, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 11245 + }, + { + "epoch": 4.493008389932082, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11246 + }, + { + "epoch": 4.493407910507391, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11247 + }, + { + "epoch": 4.4938074310827005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11248 + }, + { + "epoch": 4.49420695165801, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 11249 + }, + { + "epoch": 4.49460647223332, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11250 + }, + { + "epoch": 4.49500599280863, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11251 + }, + { + "epoch": 4.49540551338394, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11252 + }, + { + "epoch": 4.495805033959249, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11253 + }, + { + "epoch": 4.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11254 + }, + { + "epoch": 4.496604075109868, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11255 + }, + { + "epoch": 4.497003595685178, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11256 + }, + { + "epoch": 4.497403116260488, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11257 + }, + { + "epoch": 4.497802636835797, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 11258 + }, + { + "epoch": 4.498202157411106, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11259 + }, + { + "epoch": 4.498601677986416, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11260 + }, + { + "epoch": 4.499001198561726, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11261 + }, + { + "epoch": 4.499400719137036, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11262 + }, + { + "epoch": 4.4998002397123456, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11263 + }, + { + "epoch": 4.5001997602876544, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11264 + }, + { + "epoch": 4.500599280862964, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11265 + }, + { + "epoch": 4.500998801438274, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 11266 + }, + { + "epoch": 4.501398322013584, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 11267 + }, + { + "epoch": 4.501797842588894, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11268 + }, + { + "epoch": 4.502197363164203, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11269 + }, + { + "epoch": 4.502596883739512, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11270 + }, + { + "epoch": 4.502996404314822, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11271 + }, + { + "epoch": 4.503395924890132, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11272 + }, + { + "epoch": 4.503795445465442, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 11273 + }, + { + "epoch": 4.504194966040751, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11274 + }, + { + "epoch": 4.50459448661606, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11275 + }, + { + "epoch": 4.50499400719137, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11276 + }, + { + "epoch": 4.50539352776668, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11277 + }, + { + "epoch": 4.50579304834199, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11278 + }, + { + "epoch": 4.5061925689172995, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11279 + }, + { + "epoch": 4.506592089492609, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11280 + }, + { + "epoch": 4.506991610067918, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11281 + }, + { + "epoch": 4.507391130643228, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11282 + }, + { + "epoch": 4.507790651218538, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11283 + }, + { + "epoch": 4.5081901717938475, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11284 + }, + { + "epoch": 4.508589692369157, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11285 + }, + { + "epoch": 4.508989212944467, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11286 + }, + { + "epoch": 4.509388733519776, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11287 + }, + { + "epoch": 4.509788254095086, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11288 + }, + { + "epoch": 4.5101877746703956, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11289 + }, + { + "epoch": 4.510587295245705, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11290 + }, + { + "epoch": 4.510986815821015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11291 + }, + { + "epoch": 4.511386336396324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 11292 + }, + { + "epoch": 4.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11293 + }, + { + "epoch": 4.512185377546944, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11294 + }, + { + "epoch": 4.512584898122253, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11295 + }, + { + "epoch": 4.512984418697563, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11296 + }, + { + "epoch": 4.513383939272872, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11297 + }, + { + "epoch": 4.513783459848182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11298 + }, + { + "epoch": 4.514182980423492, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11299 + }, + { + "epoch": 4.514582500998801, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11300 + }, + { + "epoch": 4.514982021574111, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11301 + }, + { + "epoch": 4.515381542149421, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.2975, + "step": 11302 + }, + { + "epoch": 4.515781062724731, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11303 + }, + { + "epoch": 4.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11304 + }, + { + "epoch": 4.5165801038753495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 11305 + }, + { + "epoch": 4.516979624450659, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 11306 + }, + { + "epoch": 4.517379145025969, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11307 + }, + { + "epoch": 4.517778665601279, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 11308 + }, + { + "epoch": 4.518178186176588, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11309 + }, + { + "epoch": 4.5185777067518975, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11310 + }, + { + "epoch": 4.518977227327207, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11311 + }, + { + "epoch": 4.519376747902517, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11312 + }, + { + "epoch": 4.519776268477827, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11313 + }, + { + "epoch": 4.520175789053136, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 11314 + }, + { + "epoch": 4.520575309628446, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11315 + }, + { + "epoch": 4.520974830203755, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 11316 + }, + { + "epoch": 4.521374350779065, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11317 + }, + { + "epoch": 4.521773871354375, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11318 + }, + { + "epoch": 4.522173391929685, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11319 + }, + { + "epoch": 4.5225729125049945, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 11320 + }, + { + "epoch": 4.522972433080303, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11321 + }, + { + "epoch": 4.523371953655613, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11322 + }, + { + "epoch": 4.523771474230923, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11323 + }, + { + "epoch": 4.524170994806233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11324 + }, + { + "epoch": 4.5245705153815425, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11325 + }, + { + "epoch": 4.524970035956851, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11326 + }, + { + "epoch": 4.525369556532161, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11327 + }, + { + "epoch": 4.525769077107471, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11328 + }, + { + "epoch": 4.526168597682781, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11329 + }, + { + "epoch": 4.526568118258091, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11330 + }, + { + "epoch": 4.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11331 + }, + { + "epoch": 4.527367159408709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11332 + }, + { + "epoch": 4.527766679984019, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11333 + }, + { + "epoch": 4.528166200559329, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11334 + }, + { + "epoch": 4.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 11335 + }, + { + "epoch": 4.528965241709948, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11336 + }, + { + "epoch": 4.529364762285257, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11337 + }, + { + "epoch": 4.529764282860567, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11338 + }, + { + "epoch": 4.530163803435877, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11339 + }, + { + "epoch": 4.530563324011187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11340 + }, + { + "epoch": 4.5309628445864965, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11341 + }, + { + "epoch": 4.531362365161806, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11342 + }, + { + "epoch": 4.531761885737115, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 11343 + }, + { + "epoch": 4.532161406312425, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11344 + }, + { + "epoch": 4.532560926887735, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11345 + }, + { + "epoch": 4.5329604474630445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3387, + "step": 11346 + }, + { + "epoch": 4.533359968038354, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11347 + }, + { + "epoch": 4.533759488613663, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11348 + }, + { + "epoch": 4.534159009188973, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11349 + }, + { + "epoch": 4.534558529764283, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 11350 + }, + { + "epoch": 4.5349580503395925, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11351 + }, + { + "epoch": 4.535357570914902, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11352 + }, + { + "epoch": 4.535757091490212, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 11353 + }, + { + "epoch": 4.536156612065521, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11354 + }, + { + "epoch": 4.536556132640831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 11355 + }, + { + "epoch": 4.536955653216141, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11356 + }, + { + "epoch": 4.53735517379145, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11357 + }, + { + "epoch": 4.53775469436676, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11358 + }, + { + "epoch": 4.53815421494207, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11359 + }, + { + "epoch": 4.538553735517379, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11360 + }, + { + "epoch": 4.538953256092689, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11361 + }, + { + "epoch": 4.539352776667998, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11362 + }, + { + "epoch": 4.539752297243308, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11363 + }, + { + "epoch": 4.540151817818618, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11364 + }, + { + "epoch": 4.540551338393927, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11365 + }, + { + "epoch": 4.540950858969237, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11366 + }, + { + "epoch": 4.5413503795445465, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11367 + }, + { + "epoch": 4.541749900119856, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11368 + }, + { + "epoch": 4.542149420695166, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11369 + }, + { + "epoch": 4.542548941270476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11370 + }, + { + "epoch": 4.542948461845785, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11371 + }, + { + "epoch": 4.5433479824210945, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11372 + }, + { + "epoch": 4.543747502996404, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11373 + }, + { + "epoch": 4.544147023571714, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11374 + }, + { + "epoch": 4.544546544147024, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11375 + }, + { + "epoch": 4.544946064722334, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11376 + }, + { + "epoch": 4.5453455852976425, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 11377 + }, + { + "epoch": 4.545745105872952, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11378 + }, + { + "epoch": 4.546144626448262, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11379 + }, + { + "epoch": 4.546544147023572, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11380 + }, + { + "epoch": 4.546943667598882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11381 + }, + { + "epoch": 4.547343188174191, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11382 + }, + { + "epoch": 4.5477427087495, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11383 + }, + { + "epoch": 4.54814222932481, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11384 + }, + { + "epoch": 4.54854174990012, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11385 + }, + { + "epoch": 4.54894127047543, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11386 + }, + { + "epoch": 4.549340791050739, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11387 + }, + { + "epoch": 4.549740311626048, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11388 + }, + { + "epoch": 4.550139832201358, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11389 + }, + { + "epoch": 4.550539352776668, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11390 + }, + { + "epoch": 4.550938873351978, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11391 + }, + { + "epoch": 4.551338393927288, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11392 + }, + { + "epoch": 4.551737914502597, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11393 + }, + { + "epoch": 4.552137435077906, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11394 + }, + { + "epoch": 4.552536955653216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11395 + }, + { + "epoch": 4.552936476228526, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11396 + }, + { + "epoch": 4.553335996803836, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11397 + }, + { + "epoch": 4.553735517379145, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11398 + }, + { + "epoch": 4.554135037954454, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11399 + }, + { + "epoch": 4.554534558529764, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11400 + }, + { + "epoch": 4.554934079105074, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 11401 + }, + { + "epoch": 4.555333599680384, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11402 + }, + { + "epoch": 4.555733120255693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11403 + }, + { + "epoch": 4.556132640831002, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11404 + }, + { + "epoch": 4.556532161406312, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 11405 + }, + { + "epoch": 4.556931681981622, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 11406 + }, + { + "epoch": 4.557331202556932, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11407 + }, + { + "epoch": 4.5577307231322415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11408 + }, + { + "epoch": 4.558130243707551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 11409 + }, + { + "epoch": 4.558529764282861, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11410 + }, + { + "epoch": 4.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11411 + }, + { + "epoch": 4.55932880543348, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11412 + }, + { + "epoch": 4.5597283260087895, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11413 + }, + { + "epoch": 4.560127846584099, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11414 + }, + { + "epoch": 4.560527367159409, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 11415 + }, + { + "epoch": 4.560926887734718, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11416 + }, + { + "epoch": 4.561326408310028, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11417 + }, + { + "epoch": 4.561725928885338, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11418 + }, + { + "epoch": 4.562125449460647, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11419 + }, + { + "epoch": 4.562524970035957, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11420 + }, + { + "epoch": 4.562924490611266, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11421 + }, + { + "epoch": 4.563324011186576, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11422 + }, + { + "epoch": 4.563723531761886, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11423 + }, + { + "epoch": 4.564123052337195, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11424 + }, + { + "epoch": 4.564522572912505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11425 + }, + { + "epoch": 4.564922093487815, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11426 + }, + { + "epoch": 4.565321614063124, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11427 + }, + { + "epoch": 4.565721134638434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11428 + }, + { + "epoch": 4.566120655213743, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11429 + }, + { + "epoch": 4.566520175789053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11430 + }, + { + "epoch": 4.566919696364363, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11431 + }, + { + "epoch": 4.567319216939673, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11432 + }, + { + "epoch": 4.567718737514982, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11433 + }, + { + "epoch": 4.5681182580902915, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11434 + }, + { + "epoch": 4.568517778665601, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 11435 + }, + { + "epoch": 4.568917299240911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11436 + }, + { + "epoch": 4.569316819816221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11437 + }, + { + "epoch": 4.56971634039153, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11438 + }, + { + "epoch": 4.5701158609668395, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11439 + }, + { + "epoch": 4.570515381542149, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11440 + }, + { + "epoch": 4.570914902117459, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 11441 + }, + { + "epoch": 4.571314422692769, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11442 + }, + { + "epoch": 4.571713943268079, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11443 + }, + { + "epoch": 4.572113463843388, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11444 + }, + { + "epoch": 4.572512984418697, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11445 + }, + { + "epoch": 4.572912504994007, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11446 + }, + { + "epoch": 4.573312025569317, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 11447 + }, + { + "epoch": 4.573711546144627, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11448 + }, + { + "epoch": 4.5741110667199365, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11449 + }, + { + "epoch": 4.574510587295245, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11450 + }, + { + "epoch": 4.574910107870555, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11451 + }, + { + "epoch": 4.575309628445865, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 11452 + }, + { + "epoch": 4.575709149021175, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11453 + }, + { + "epoch": 4.5761086695964845, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11454 + }, + { + "epoch": 4.5765081901717934, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11455 + }, + { + "epoch": 4.576907710747103, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11456 + }, + { + "epoch": 4.577307231322413, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 11457 + }, + { + "epoch": 4.577706751897723, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11458 + }, + { + "epoch": 4.578106272473033, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11459 + }, + { + "epoch": 4.578505793048342, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11460 + }, + { + "epoch": 4.578905313623651, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11461 + }, + { + "epoch": 4.579304834198961, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11462 + }, + { + "epoch": 4.579704354774271, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11463 + }, + { + "epoch": 4.580103875349581, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11464 + }, + { + "epoch": 4.58050339592489, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11465 + }, + { + "epoch": 4.5809029165002, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11466 + }, + { + "epoch": 4.581302437075509, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11467 + }, + { + "epoch": 4.581701957650819, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11468 + }, + { + "epoch": 4.582101478226129, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11469 + }, + { + "epoch": 4.5825009988014385, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 11470 + }, + { + "epoch": 4.582900519376748, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11471 + }, + { + "epoch": 4.583300039952057, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 11472 + }, + { + "epoch": 4.583699560527367, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 11473 + }, + { + "epoch": 4.584099081102677, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11474 + }, + { + "epoch": 4.5844986016779865, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11475 + }, + { + "epoch": 4.584898122253296, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11476 + }, + { + "epoch": 4.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11477 + }, + { + "epoch": 4.585697163403915, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11478 + }, + { + "epoch": 4.586096683979225, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11479 + }, + { + "epoch": 4.5864962045545346, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11480 + }, + { + "epoch": 4.586895725129844, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11481 + }, + { + "epoch": 4.587295245705154, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 11482 + }, + { + "epoch": 4.587694766280464, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11483 + }, + { + "epoch": 4.588094286855773, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 11484 + }, + { + "epoch": 4.588493807431083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3375, + "step": 11485 + }, + { + "epoch": 4.588893328006392, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11486 + }, + { + "epoch": 4.589292848581702, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 11487 + }, + { + "epoch": 4.589692369157012, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11488 + }, + { + "epoch": 4.590091889732321, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11489 + }, + { + "epoch": 4.590491410307631, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 11490 + }, + { + "epoch": 4.59089093088294, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3392, + "step": 11491 + }, + { + "epoch": 4.59129045145825, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11492 + }, + { + "epoch": 4.59168997203356, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11493 + }, + { + "epoch": 4.592089492608869, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11494 + }, + { + "epoch": 4.592489013184179, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11495 + }, + { + "epoch": 4.5928885337594885, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 11496 + }, + { + "epoch": 4.593288054334798, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11497 + }, + { + "epoch": 4.593687574910108, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11498 + }, + { + "epoch": 4.594087095485418, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11499 + }, + { + "epoch": 4.594486616060728, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11500 + }, + { + "epoch": 4.5948861366360365, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11501 + }, + { + "epoch": 4.595285657211346, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11502 + }, + { + "epoch": 4.595685177786656, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11503 + }, + { + "epoch": 4.596084698361966, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11504 + }, + { + "epoch": 4.596484218937276, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11505 + }, + { + "epoch": 4.5968837395125846, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11506 + }, + { + "epoch": 4.597283260087894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 11507 + }, + { + "epoch": 4.597682780663204, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 11508 + }, + { + "epoch": 4.598082301238514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11509 + }, + { + "epoch": 4.598481821813824, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11510 + }, + { + "epoch": 4.598881342389133, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11511 + }, + { + "epoch": 4.599280862964442, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11512 + }, + { + "epoch": 4.599680383539752, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11513 + }, + { + "epoch": 4.600079904115062, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11514 + }, + { + "epoch": 4.600479424690372, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11515 + }, + { + "epoch": 4.6008789452656815, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11516 + }, + { + "epoch": 4.60127846584099, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 11517 + }, + { + "epoch": 4.6016779864163, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11518 + }, + { + "epoch": 4.60207750699161, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11519 + }, + { + "epoch": 4.60247702756692, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11520 + }, + { + "epoch": 4.60287654814223, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11521 + }, + { + "epoch": 4.603276068717539, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11522 + }, + { + "epoch": 4.603675589292848, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11523 + }, + { + "epoch": 4.604075109868158, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11524 + }, + { + "epoch": 4.604474630443468, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11525 + }, + { + "epoch": 4.604874151018778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 11526 + }, + { + "epoch": 4.605273671594087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11527 + }, + { + "epoch": 4.605673192169396, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11528 + }, + { + "epoch": 4.606072712744706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11529 + }, + { + "epoch": 4.606472233320016, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11530 + }, + { + "epoch": 4.606871753895326, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 11531 + }, + { + "epoch": 4.6072712744706354, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3379, + "step": 11532 + }, + { + "epoch": 4.607670795045945, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11533 + }, + { + "epoch": 4.608070315621254, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 11534 + }, + { + "epoch": 4.608469836196564, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11535 + }, + { + "epoch": 4.608869356771874, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11536 + }, + { + "epoch": 4.6092688773471835, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 11537 + }, + { + "epoch": 4.609668397922493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11538 + }, + { + "epoch": 4.610067918497803, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11539 + }, + { + "epoch": 4.610467439073112, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11540 + }, + { + "epoch": 4.610866959648422, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11541 + }, + { + "epoch": 4.6112664802237315, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11542 + }, + { + "epoch": 4.611666000799041, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11543 + }, + { + "epoch": 4.612065521374351, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11544 + }, + { + "epoch": 4.61246504194966, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11545 + }, + { + "epoch": 4.61286456252497, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11546 + }, + { + "epoch": 4.61326408310028, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11547 + }, + { + "epoch": 4.613663603675589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 11548 + }, + { + "epoch": 4.614063124250899, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11549 + }, + { + "epoch": 4.614462644826209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11550 + }, + { + "epoch": 4.614862165401518, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11551 + }, + { + "epoch": 4.615261685976828, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 11552 + }, + { + "epoch": 4.615661206552137, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11553 + }, + { + "epoch": 4.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11554 + }, + { + "epoch": 4.616460247702757, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 11555 + }, + { + "epoch": 4.616859768278067, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11556 + }, + { + "epoch": 4.617259288853376, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11557 + }, + { + "epoch": 4.6176588094286855, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11558 + }, + { + "epoch": 4.618058330003995, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11559 + }, + { + "epoch": 4.618457850579305, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 11560 + }, + { + "epoch": 4.618857371154615, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11561 + }, + { + "epoch": 4.619256891729924, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11562 + }, + { + "epoch": 4.6196564123052335, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11563 + }, + { + "epoch": 4.620055932880543, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11564 + }, + { + "epoch": 4.620455453455853, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 11565 + }, + { + "epoch": 4.620854974031163, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11566 + }, + { + "epoch": 4.621254494606472, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11567 + }, + { + "epoch": 4.6216540151817815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11568 + }, + { + "epoch": 4.622053535757091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11569 + }, + { + "epoch": 4.622453056332401, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11570 + }, + { + "epoch": 4.622852576907711, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11571 + }, + { + "epoch": 4.623252097483021, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 11572 + }, + { + "epoch": 4.6236516180583305, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11573 + }, + { + "epoch": 4.624051138633639, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11574 + }, + { + "epoch": 4.624450659208949, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11575 + }, + { + "epoch": 4.624850179784259, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11576 + }, + { + "epoch": 4.625249700359569, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11577 + }, + { + "epoch": 4.6256492209348785, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11578 + }, + { + "epoch": 4.626048741510187, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11579 + }, + { + "epoch": 4.626448262085497, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11580 + }, + { + "epoch": 4.626847782660807, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11581 + }, + { + "epoch": 4.627247303236117, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11582 + }, + { + "epoch": 4.627646823811427, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 11583 + }, + { + "epoch": 4.6280463443867355, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11584 + }, + { + "epoch": 4.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11585 + }, + { + "epoch": 4.628845385537355, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11586 + }, + { + "epoch": 4.629244906112665, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11587 + }, + { + "epoch": 4.629644426687975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11588 + }, + { + "epoch": 4.630043947263284, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11589 + }, + { + "epoch": 4.630443467838594, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11590 + }, + { + "epoch": 4.630842988413903, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 11591 + }, + { + "epoch": 4.631242508989213, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11592 + }, + { + "epoch": 4.631642029564523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11593 + }, + { + "epoch": 4.632041550139832, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11594 + }, + { + "epoch": 4.632441070715142, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 11595 + }, + { + "epoch": 4.632840591290451, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11596 + }, + { + "epoch": 4.633240111865761, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11597 + }, + { + "epoch": 4.633639632441071, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11598 + }, + { + "epoch": 4.6340391530163805, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11599 + }, + { + "epoch": 4.63443867359169, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11600 + }, + { + "epoch": 4.634838194166999, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11601 + }, + { + "epoch": 4.635237714742309, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11602 + }, + { + "epoch": 4.635637235317619, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11603 + }, + { + "epoch": 4.6360367558929285, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 11604 + }, + { + "epoch": 4.636436276468238, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11605 + }, + { + "epoch": 4.636835797043548, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11606 + }, + { + "epoch": 4.637235317618857, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11607 + }, + { + "epoch": 4.637634838194167, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 11608 + }, + { + "epoch": 4.638034358769477, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 11609 + }, + { + "epoch": 4.638433879344786, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 11610 + }, + { + "epoch": 4.638833399920096, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11611 + }, + { + "epoch": 4.639232920495406, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11612 + }, + { + "epoch": 4.639632441070715, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11613 + }, + { + "epoch": 4.640031961646025, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11614 + }, + { + "epoch": 4.640431482221334, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11615 + }, + { + "epoch": 4.640831002796644, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11616 + }, + { + "epoch": 4.641230523371954, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11617 + }, + { + "epoch": 4.641630043947263, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11618 + }, + { + "epoch": 4.642029564522573, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11619 + }, + { + "epoch": 4.642429085097882, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 11620 + }, + { + "epoch": 4.642828605673192, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11621 + }, + { + "epoch": 4.643228126248502, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11622 + }, + { + "epoch": 4.643627646823812, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11623 + }, + { + "epoch": 4.644027167399121, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11624 + }, + { + "epoch": 4.6444266879744305, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11625 + }, + { + "epoch": 4.64482620854974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11626 + }, + { + "epoch": 4.64522572912505, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 11627 + }, + { + "epoch": 4.64562524970036, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11628 + }, + { + "epoch": 4.64602477027567, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11629 + }, + { + "epoch": 4.6464242908509785, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11630 + }, + { + "epoch": 4.646823811426288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11631 + }, + { + "epoch": 4.647223332001598, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11632 + }, + { + "epoch": 4.647622852576908, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11633 + }, + { + "epoch": 4.648022373152218, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11634 + }, + { + "epoch": 4.648421893727527, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 11635 + }, + { + "epoch": 4.648821414302836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11636 + }, + { + "epoch": 4.649220934878146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11637 + }, + { + "epoch": 4.649620455453456, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11638 + }, + { + "epoch": 4.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11639 + }, + { + "epoch": 4.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11640 + }, + { + "epoch": 4.650819017179384, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11641 + }, + { + "epoch": 4.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 11642 + }, + { + "epoch": 4.651618058330004, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11643 + }, + { + "epoch": 4.652017578905314, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11644 + }, + { + "epoch": 4.6524170994806235, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 11645 + }, + { + "epoch": 4.652816620055933, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 11646 + }, + { + "epoch": 4.653216140631242, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11647 + }, + { + "epoch": 4.653615661206552, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11648 + }, + { + "epoch": 4.654015181781862, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11649 + }, + { + "epoch": 4.654414702357172, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11650 + }, + { + "epoch": 4.654814222932481, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11651 + }, + { + "epoch": 4.65521374350779, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11652 + }, + { + "epoch": 4.6556132640831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11653 + }, + { + "epoch": 4.65601278465841, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 11654 + }, + { + "epoch": 4.65641230523372, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11655 + }, + { + "epoch": 4.656811825809029, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11656 + }, + { + "epoch": 4.657211346384338, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11657 + }, + { + "epoch": 4.657610866959648, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11658 + }, + { + "epoch": 4.658010387534958, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 11659 + }, + { + "epoch": 4.658409908110268, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11660 + }, + { + "epoch": 4.6588094286855775, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11661 + }, + { + "epoch": 4.659208949260887, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 11662 + }, + { + "epoch": 4.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11663 + }, + { + "epoch": 4.660007990411506, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11664 + }, + { + "epoch": 4.660407510986816, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 11665 + }, + { + "epoch": 4.6608070315621255, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 11666 + }, + { + "epoch": 4.661206552137435, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11667 + }, + { + "epoch": 4.661606072712745, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11668 + }, + { + "epoch": 4.662005593288054, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 11669 + }, + { + "epoch": 4.662405113863364, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11670 + }, + { + "epoch": 4.6628046344386735, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 11671 + }, + { + "epoch": 4.663204155013983, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11672 + }, + { + "epoch": 4.663603675589293, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11673 + }, + { + "epoch": 4.664003196164602, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11674 + }, + { + "epoch": 4.664402716739912, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11675 + }, + { + "epoch": 4.664802237315222, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 11676 + }, + { + "epoch": 4.665201757890531, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11677 + }, + { + "epoch": 4.665601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11678 + }, + { + "epoch": 4.666000799041151, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11679 + }, + { + "epoch": 4.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11680 + }, + { + "epoch": 4.66679984019177, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 11681 + }, + { + "epoch": 4.667199360767079, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11682 + }, + { + "epoch": 4.667598881342389, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 11683 + }, + { + "epoch": 4.667998401917699, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11684 + }, + { + "epoch": 4.668397922493009, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11685 + }, + { + "epoch": 4.668797443068318, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11686 + }, + { + "epoch": 4.6691969636436275, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11687 + }, + { + "epoch": 4.669596484218937, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11688 + }, + { + "epoch": 4.669996004794247, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11689 + }, + { + "epoch": 4.670395525369557, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 11690 + }, + { + "epoch": 4.670795045944866, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 11691 + }, + { + "epoch": 4.6711945665201755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 11692 + }, + { + "epoch": 4.671594087095485, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 11693 + }, + { + "epoch": 4.671993607670795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11694 + }, + { + "epoch": 4.672393128246105, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11695 + }, + { + "epoch": 4.672792648821415, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 11696 + }, + { + "epoch": 4.6731921693967235, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11697 + }, + { + "epoch": 4.673591689972033, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 11698 + }, + { + "epoch": 4.673991210547343, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11699 + }, + { + "epoch": 4.674390731122653, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 11700 + }, + { + "epoch": 4.674790251697963, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 11701 + }, + { + "epoch": 4.6751897722732725, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 11702 + }, + { + "epoch": 4.675589292848581, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11703 + }, + { + "epoch": 4.675988813423891, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11704 + }, + { + "epoch": 4.676388333999201, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11705 + }, + { + "epoch": 4.676787854574511, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 11706 + }, + { + "epoch": 4.6771873751498205, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 11707 + }, + { + "epoch": 4.677586895725129, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11708 + }, + { + "epoch": 4.677986416300439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11709 + }, + { + "epoch": 4.678385936875749, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11710 + }, + { + "epoch": 4.678785457451059, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 11711 + }, + { + "epoch": 4.679184978026369, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 11712 + }, + { + "epoch": 4.679584498601678, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3356, + "step": 11713 + }, + { + "epoch": 4.679984019176987, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11714 + }, + { + "epoch": 4.680383539752297, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 11715 + }, + { + "epoch": 4.680783060327607, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11716 + }, + { + "epoch": 4.681182580902917, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11717 + }, + { + "epoch": 4.681582101478226, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 11718 + }, + { + "epoch": 4.681981622053536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11719 + }, + { + "epoch": 4.682381142628845, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 11720 + }, + { + "epoch": 4.682780663204155, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 11721 + }, + { + "epoch": 4.683180183779465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11722 + }, + { + "epoch": 4.683579704354774, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 11723 + }, + { + "epoch": 4.683979224930084, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 11724 + }, + { + "epoch": 4.684378745505393, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11725 + }, + { + "epoch": 4.684778266080703, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 11726 + }, + { + "epoch": 4.685177786656013, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11727 + }, + { + "epoch": 4.6855773072313225, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11728 + }, + { + "epoch": 4.685976827806632, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11729 + }, + { + "epoch": 4.686376348381942, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11730 + }, + { + "epoch": 4.686775868957251, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11731 + }, + { + "epoch": 4.687175389532561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11732 + }, + { + "epoch": 4.6875749101078705, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 11733 + }, + { + "epoch": 4.68797443068318, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11734 + }, + { + "epoch": 4.68837395125849, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11735 + }, + { + "epoch": 4.6887734718338, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11736 + }, + { + "epoch": 4.689172992409109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11737 + }, + { + "epoch": 4.689572512984419, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 11738 + }, + { + "epoch": 4.689972033559728, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11739 + }, + { + "epoch": 4.690371554135038, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 11740 + }, + { + "epoch": 4.690771074710348, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11741 + }, + { + "epoch": 4.691170595285657, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 11742 + }, + { + "epoch": 4.691570115860967, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11743 + }, + { + "epoch": 4.691969636436276, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11744 + }, + { + "epoch": 4.692369157011586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11745 + }, + { + "epoch": 4.692768677586896, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11746 + }, + { + "epoch": 4.693168198162205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11747 + }, + { + "epoch": 4.693567718737515, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11748 + }, + { + "epoch": 4.6939672393128244, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 11749 + }, + { + "epoch": 4.694366759888134, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 11750 + }, + { + "epoch": 4.694766280463444, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11751 + }, + { + "epoch": 4.695165801038754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11752 + }, + { + "epoch": 4.695565321614064, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11753 + }, + { + "epoch": 4.6959648421893725, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 11754 + }, + { + "epoch": 4.696364362764682, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 11755 + }, + { + "epoch": 4.696763883339992, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11756 + }, + { + "epoch": 4.697163403915302, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11757 + }, + { + "epoch": 4.697562924490612, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 11758 + }, + { + "epoch": 4.6979624450659205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11759 + }, + { + "epoch": 4.69836196564123, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 11760 + }, + { + "epoch": 4.69876148621654, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 11761 + }, + { + "epoch": 4.69916100679185, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 11762 + }, + { + "epoch": 4.69956052736716, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 11763 + }, + { + "epoch": 4.699960047942469, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 11764 + }, + { + "epoch": 4.700359568517778, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11765 + }, + { + "epoch": 4.700759089093088, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11766 + }, + { + "epoch": 4.701158609668398, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 11767 + }, + { + "epoch": 4.701558130243708, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 11768 + }, + { + "epoch": 4.7019576508190175, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11769 + }, + { + "epoch": 4.702357171394327, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11770 + }, + { + "epoch": 4.702756691969636, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11771 + }, + { + "epoch": 4.703156212544946, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11772 + }, + { + "epoch": 4.703555733120256, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 11773 + }, + { + "epoch": 4.7039552536955656, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11774 + }, + { + "epoch": 4.704354774270875, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11775 + }, + { + "epoch": 4.704754294846184, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 11776 + }, + { + "epoch": 4.705153815421494, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 11777 + }, + { + "epoch": 4.705553335996804, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 11778 + }, + { + "epoch": 4.705952856572114, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11779 + }, + { + "epoch": 4.706352377147423, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 11780 + }, + { + "epoch": 4.706751897722732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11781 + }, + { + "epoch": 4.707151418298042, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3369, + "step": 11782 + }, + { + "epoch": 4.707550938873352, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11783 + }, + { + "epoch": 4.707950459448662, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 11784 + }, + { + "epoch": 4.708349980023971, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11785 + }, + { + "epoch": 4.708749500599281, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11786 + }, + { + "epoch": 4.70914902117459, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11787 + }, + { + "epoch": 4.7095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 11788 + }, + { + "epoch": 4.70994806232521, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11789 + }, + { + "epoch": 4.7103475829005195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 11790 + }, + { + "epoch": 4.710747103475829, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 11791 + }, + { + "epoch": 4.711146624051139, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11792 + }, + { + "epoch": 4.711546144626448, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 11793 + }, + { + "epoch": 4.711945665201758, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11794 + }, + { + "epoch": 4.7123451857770675, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 11795 + }, + { + "epoch": 4.712744706352377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 11796 + }, + { + "epoch": 4.713144226927687, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 11797 + }, + { + "epoch": 4.713543747502996, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11798 + }, + { + "epoch": 4.713943268078306, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11799 + }, + { + "epoch": 4.714342788653616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 11800 + }, + { + "epoch": 4.714742309228925, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 11801 + }, + { + "epoch": 4.715141829804235, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11802 + }, + { + "epoch": 4.715541350379545, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 11803 + }, + { + "epoch": 4.715940870954854, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11804 + }, + { + "epoch": 4.716340391530164, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11805 + }, + { + "epoch": 4.716739912105473, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11806 + }, + { + "epoch": 4.717139432680783, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11807 + }, + { + "epoch": 4.717538953256093, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 11808 + }, + { + "epoch": 4.717938473831403, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 11809 + }, + { + "epoch": 4.718337994406712, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11810 + }, + { + "epoch": 4.718737514982021, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11811 + }, + { + "epoch": 4.719137035557331, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11812 + }, + { + "epoch": 4.719536556132641, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11813 + }, + { + "epoch": 4.719936076707951, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 11814 + }, + { + "epoch": 4.72033559728326, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11815 + }, + { + "epoch": 4.7207351178585695, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 11816 + }, + { + "epoch": 4.721134638433879, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 11817 + }, + { + "epoch": 4.721534159009189, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 11818 + }, + { + "epoch": 4.721933679584499, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11819 + }, + { + "epoch": 4.722333200159808, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11820 + }, + { + "epoch": 4.7227327207351175, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11821 + }, + { + "epoch": 4.723132241310427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 11822 + }, + { + "epoch": 4.723531761885737, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11823 + }, + { + "epoch": 4.723931282461047, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 11824 + }, + { + "epoch": 4.724330803036357, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11825 + }, + { + "epoch": 4.7247303236116664, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11826 + }, + { + "epoch": 4.725129844186975, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11827 + }, + { + "epoch": 4.725529364762285, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11828 + }, + { + "epoch": 4.725928885337595, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 11829 + }, + { + "epoch": 4.726328405912905, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 11830 + }, + { + "epoch": 4.7267279264882145, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 11831 + }, + { + "epoch": 4.727127447063523, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11832 + }, + { + "epoch": 4.727526967638833, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11833 + }, + { + "epoch": 4.727926488214143, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 11834 + }, + { + "epoch": 4.728326008789453, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11835 + }, + { + "epoch": 4.7287255293647625, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 11836 + }, + { + "epoch": 4.729125049940071, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11837 + }, + { + "epoch": 4.729524570515381, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11838 + }, + { + "epoch": 4.729924091090691, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11839 + }, + { + "epoch": 4.730323611666001, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11840 + }, + { + "epoch": 4.730723132241311, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11841 + }, + { + "epoch": 4.73112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 11842 + }, + { + "epoch": 4.73152217339193, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11843 + }, + { + "epoch": 4.731921693967239, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 11844 + }, + { + "epoch": 4.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 11845 + }, + { + "epoch": 4.732720735117859, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11846 + }, + { + "epoch": 4.733120255693168, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11847 + }, + { + "epoch": 4.733519776268478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 11848 + }, + { + "epoch": 4.733919296843787, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 11849 + }, + { + "epoch": 4.734318817419097, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11850 + }, + { + "epoch": 4.734718337994407, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 11851 + }, + { + "epoch": 4.7351178585697165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11852 + }, + { + "epoch": 4.735517379145026, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 11853 + }, + { + "epoch": 4.735916899720335, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11854 + }, + { + "epoch": 4.736316420295645, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11855 + }, + { + "epoch": 4.736715940870955, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11856 + }, + { + "epoch": 4.7371154614462645, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11857 + }, + { + "epoch": 4.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11858 + }, + { + "epoch": 4.737914502596884, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 11859 + }, + { + "epoch": 4.738314023172194, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 11860 + }, + { + "epoch": 4.738713543747503, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 11861 + }, + { + "epoch": 4.7391130643228125, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11862 + }, + { + "epoch": 4.739512584898122, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 11863 + }, + { + "epoch": 4.739912105473432, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11864 + }, + { + "epoch": 4.740311626048742, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11865 + }, + { + "epoch": 4.740711146624051, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11866 + }, + { + "epoch": 4.741110667199361, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11867 + }, + { + "epoch": 4.74151018777467, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11868 + }, + { + "epoch": 4.74190970834998, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11869 + }, + { + "epoch": 4.74230922892529, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11870 + }, + { + "epoch": 4.742708749500599, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 11871 + }, + { + "epoch": 4.743108270075909, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11872 + }, + { + "epoch": 4.743507790651218, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 11873 + }, + { + "epoch": 4.743907311226528, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 11874 + }, + { + "epoch": 4.744306831801838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 11875 + }, + { + "epoch": 4.744706352377148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11876 + }, + { + "epoch": 4.745105872952457, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11877 + }, + { + "epoch": 4.7455053935277665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 11878 + }, + { + "epoch": 4.745904914103076, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 11879 + }, + { + "epoch": 4.746304434678386, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11880 + }, + { + "epoch": 4.746703955253696, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11881 + }, + { + "epoch": 4.747103475829006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 11882 + }, + { + "epoch": 4.7475029964043145, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 11883 + }, + { + "epoch": 4.747902516979624, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 11884 + }, + { + "epoch": 4.748302037554934, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 11885 + }, + { + "epoch": 4.748701558130244, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 11886 + }, + { + "epoch": 4.749101078705554, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 11887 + }, + { + "epoch": 4.7495005992808625, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 11888 + }, + { + "epoch": 4.749900119856172, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 11889 + }, + { + "epoch": 4.750299640431482, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11890 + }, + { + "epoch": 4.750699161006792, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11891 + }, + { + "epoch": 4.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11892 + }, + { + "epoch": 4.7514982021574115, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11893 + }, + { + "epoch": 4.75189772273272, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 11894 + }, + { + "epoch": 4.75229724330803, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11895 + }, + { + "epoch": 4.75269676388334, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 11896 + }, + { + "epoch": 4.75309628445865, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 11897 + }, + { + "epoch": 4.7534958050339595, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 11898 + }, + { + "epoch": 4.753895325609269, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 11899 + }, + { + "epoch": 4.754294846184578, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11900 + }, + { + "epoch": 4.754694366759888, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 11901 + }, + { + "epoch": 4.755093887335198, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 11902 + }, + { + "epoch": 4.755493407910508, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 11903 + }, + { + "epoch": 4.755892928485817, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 11904 + }, + { + "epoch": 4.756292449061126, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 11905 + }, + { + "epoch": 4.756691969636436, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 11906 + }, + { + "epoch": 4.757091490211746, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 11907 + }, + { + "epoch": 4.757491010787056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11908 + }, + { + "epoch": 4.757890531362365, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11909 + }, + { + "epoch": 4.758290051937674, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11910 + }, + { + "epoch": 4.758689572512984, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 11911 + }, + { + "epoch": 4.759089093088294, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11912 + }, + { + "epoch": 4.759488613663604, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 11913 + }, + { + "epoch": 4.759888134238913, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 11914 + }, + { + "epoch": 4.760287654814223, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11915 + }, + { + "epoch": 4.760687175389533, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 11916 + }, + { + "epoch": 4.761086695964842, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 11917 + }, + { + "epoch": 4.761486216540152, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 11918 + }, + { + "epoch": 4.7618857371154615, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11919 + }, + { + "epoch": 4.762285257690771, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3357, + "step": 11920 + }, + { + "epoch": 4.762684778266081, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 11921 + }, + { + "epoch": 4.76308429884139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11922 + }, + { + "epoch": 4.7634838194167, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 11923 + }, + { + "epoch": 4.7638833399920095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 11924 + }, + { + "epoch": 4.764282860567319, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 11925 + }, + { + "epoch": 4.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11926 + }, + { + "epoch": 4.765081901717938, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 11927 + }, + { + "epoch": 4.765481422293248, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11928 + }, + { + "epoch": 4.765880942868558, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 11929 + }, + { + "epoch": 4.766280463443867, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11930 + }, + { + "epoch": 4.766679984019177, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 11931 + }, + { + "epoch": 4.767079504594487, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11932 + }, + { + "epoch": 4.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 11933 + }, + { + "epoch": 4.767878545745106, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 11934 + }, + { + "epoch": 4.768278066320415, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11935 + }, + { + "epoch": 4.768677586895725, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 11936 + }, + { + "epoch": 4.769077107471035, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11937 + }, + { + "epoch": 4.769476628046345, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 11938 + }, + { + "epoch": 4.769876148621654, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 11939 + }, + { + "epoch": 4.770275669196963, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 11940 + }, + { + "epoch": 4.770675189772273, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 11941 + }, + { + "epoch": 4.771074710347583, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 11942 + }, + { + "epoch": 4.771474230922893, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11943 + }, + { + "epoch": 4.771873751498202, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11944 + }, + { + "epoch": 4.7722732720735115, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11945 + }, + { + "epoch": 4.772672792648821, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11946 + }, + { + "epoch": 4.773072313224131, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 11947 + }, + { + "epoch": 4.773471833799441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 11948 + }, + { + "epoch": 4.773871354374751, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 11949 + }, + { + "epoch": 4.77427087495006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 11950 + }, + { + "epoch": 4.774670395525369, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 11951 + }, + { + "epoch": 4.775069916100679, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 11952 + }, + { + "epoch": 4.775469436675989, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 11953 + }, + { + "epoch": 4.775868957251299, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 11954 + }, + { + "epoch": 4.7762684778266085, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 11955 + }, + { + "epoch": 4.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 11956 + }, + { + "epoch": 4.777067518977227, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 11957 + }, + { + "epoch": 4.777467039552537, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 11958 + }, + { + "epoch": 4.777866560127847, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 11959 + }, + { + "epoch": 4.7782660807031565, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11960 + }, + { + "epoch": 4.778665601278465, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 11961 + }, + { + "epoch": 4.779065121853775, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 11962 + }, + { + "epoch": 4.779464642429085, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.334, + "step": 11963 + }, + { + "epoch": 4.779864163004395, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 11964 + }, + { + "epoch": 4.7802636835797045, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 11965 + }, + { + "epoch": 4.780663204155014, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 11966 + }, + { + "epoch": 4.781062724730323, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 11967 + }, + { + "epoch": 4.781462245305633, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 11968 + }, + { + "epoch": 4.781861765880943, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 11969 + }, + { + "epoch": 4.782261286456253, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 11970 + }, + { + "epoch": 4.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 11971 + }, + { + "epoch": 4.783060327606872, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 11972 + }, + { + "epoch": 4.783459848182181, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 11973 + }, + { + "epoch": 4.783859368757491, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11974 + }, + { + "epoch": 4.784258889332801, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 11975 + }, + { + "epoch": 4.78465840990811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 11976 + }, + { + "epoch": 4.78505793048342, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 11977 + }, + { + "epoch": 4.785457451058729, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 11978 + }, + { + "epoch": 4.785856971634039, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 11979 + }, + { + "epoch": 4.786256492209349, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 11980 + }, + { + "epoch": 4.7866560127846585, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 11981 + }, + { + "epoch": 4.787055533359968, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 11982 + }, + { + "epoch": 4.787455053935278, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 11983 + }, + { + "epoch": 4.787854574510587, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 11984 + }, + { + "epoch": 4.788254095085897, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 11985 + }, + { + "epoch": 4.7886536156612065, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 11986 + }, + { + "epoch": 4.789053136236516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11987 + }, + { + "epoch": 4.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 11988 + }, + { + "epoch": 4.789852177387136, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 11989 + }, + { + "epoch": 4.790251697962445, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 11990 + }, + { + "epoch": 4.7906512185377546, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 11991 + }, + { + "epoch": 4.791050739113064, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 11992 + }, + { + "epoch": 4.791450259688374, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 11993 + }, + { + "epoch": 4.791849780263684, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 11994 + }, + { + "epoch": 4.792249300838993, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 11995 + }, + { + "epoch": 4.792648821414303, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 11996 + }, + { + "epoch": 4.793048341989612, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 11997 + }, + { + "epoch": 4.793447862564922, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 11998 + }, + { + "epoch": 4.793847383140232, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 11999 + }, + { + "epoch": 4.794246903715541, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12000 + }, + { + "epoch": 4.794646424290851, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12001 + }, + { + "epoch": 4.79504594486616, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12002 + }, + { + "epoch": 4.79544546544147, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12003 + }, + { + "epoch": 4.79584498601678, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12004 + }, + { + "epoch": 4.79624450659209, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12005 + }, + { + "epoch": 4.7966440271674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12006 + }, + { + "epoch": 4.7970435477427085, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12007 + }, + { + "epoch": 4.797443068318018, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12008 + }, + { + "epoch": 4.797842588893328, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12009 + }, + { + "epoch": 4.798242109468638, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12010 + }, + { + "epoch": 4.798641630043948, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12011 + }, + { + "epoch": 4.7990411506192565, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12012 + }, + { + "epoch": 4.799440671194566, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 12013 + }, + { + "epoch": 4.799840191769876, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12014 + }, + { + "epoch": 4.800239712345186, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12015 + }, + { + "epoch": 4.800639232920496, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12016 + }, + { + "epoch": 4.801038753495805, + "grad_norm": 0.15234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12017 + }, + { + "epoch": 4.801438274071114, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12018 + }, + { + "epoch": 4.801837794646424, + "grad_norm": 0.158203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12019 + }, + { + "epoch": 4.802237315221734, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12020 + }, + { + "epoch": 4.802636835797044, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12021 + }, + { + "epoch": 4.8030363563723535, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12022 + }, + { + "epoch": 4.803435876947663, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12023 + }, + { + "epoch": 4.803835397522972, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12024 + }, + { + "epoch": 4.804234918098282, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12025 + }, + { + "epoch": 4.804634438673592, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12026 + }, + { + "epoch": 4.8050339592489015, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12027 + }, + { + "epoch": 4.805433479824211, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12028 + }, + { + "epoch": 4.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12029 + }, + { + "epoch": 4.80623252097483, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12030 + }, + { + "epoch": 4.80663204155014, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12031 + }, + { + "epoch": 4.80703156212545, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12032 + }, + { + "epoch": 4.807431082700759, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12033 + }, + { + "epoch": 4.807830603276068, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12034 + }, + { + "epoch": 4.808230123851378, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12035 + }, + { + "epoch": 4.808629644426688, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12036 + }, + { + "epoch": 4.809029165001998, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12037 + }, + { + "epoch": 4.809428685577307, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12038 + }, + { + "epoch": 4.809828206152617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12039 + }, + { + "epoch": 4.810227726727927, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12040 + }, + { + "epoch": 4.810627247303236, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12041 + }, + { + "epoch": 4.811026767878546, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12042 + }, + { + "epoch": 4.8114262884538554, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12043 + }, + { + "epoch": 4.811825809029165, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12044 + }, + { + "epoch": 4.812225329604475, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 12045 + }, + { + "epoch": 4.812624850179784, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12046 + }, + { + "epoch": 4.813024370755094, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12047 + }, + { + "epoch": 4.8134238913304035, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12048 + }, + { + "epoch": 4.813823411905713, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12049 + }, + { + "epoch": 4.814222932481023, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12050 + }, + { + "epoch": 4.814622453056332, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12051 + }, + { + "epoch": 4.815021973631642, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12052 + }, + { + "epoch": 4.8154214942069515, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12053 + }, + { + "epoch": 4.815821014782261, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12054 + }, + { + "epoch": 4.816220535357571, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12055 + }, + { + "epoch": 4.816620055932881, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12056 + }, + { + "epoch": 4.81701957650819, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12057 + }, + { + "epoch": 4.8174190970835, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12058 + }, + { + "epoch": 4.817818617658809, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12059 + }, + { + "epoch": 4.818218138234119, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12060 + }, + { + "epoch": 4.818617658809429, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12061 + }, + { + "epoch": 4.819017179384739, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12062 + }, + { + "epoch": 4.819416699960048, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12063 + }, + { + "epoch": 4.819816220535357, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12064 + }, + { + "epoch": 4.820215741110667, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12065 + }, + { + "epoch": 4.820615261685977, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12066 + }, + { + "epoch": 4.821014782261287, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12067 + }, + { + "epoch": 4.821414302836596, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12068 + }, + { + "epoch": 4.8218138234119055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12069 + }, + { + "epoch": 4.822213343987215, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12070 + }, + { + "epoch": 4.822612864562525, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12071 + }, + { + "epoch": 4.823012385137835, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12072 + }, + { + "epoch": 4.823411905713145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12073 + }, + { + "epoch": 4.8238114262884535, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12074 + }, + { + "epoch": 4.824210946863763, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12075 + }, + { + "epoch": 4.824610467439073, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12076 + }, + { + "epoch": 4.825009988014383, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12077 + }, + { + "epoch": 4.825409508589693, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 12078 + }, + { + "epoch": 4.825809029165002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12079 + }, + { + "epoch": 4.826208549740311, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12080 + }, + { + "epoch": 4.826608070315621, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12081 + }, + { + "epoch": 4.827007590890931, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12082 + }, + { + "epoch": 4.827407111466241, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12083 + }, + { + "epoch": 4.8278066320415505, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12084 + }, + { + "epoch": 4.828206152616859, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12085 + }, + { + "epoch": 4.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12086 + }, + { + "epoch": 4.829005193767479, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12087 + }, + { + "epoch": 4.829404714342789, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12088 + }, + { + "epoch": 4.8298042349180985, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12089 + }, + { + "epoch": 4.830203755493407, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12090 + }, + { + "epoch": 4.830603276068717, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12091 + }, + { + "epoch": 4.831002796644027, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12092 + }, + { + "epoch": 4.831402317219337, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12093 + }, + { + "epoch": 4.831801837794647, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 12094 + }, + { + "epoch": 4.832201358369956, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12095 + }, + { + "epoch": 4.832600878945266, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12096 + }, + { + "epoch": 4.833000399520575, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12097 + }, + { + "epoch": 4.833399920095885, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12098 + }, + { + "epoch": 4.833799440671195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12099 + }, + { + "epoch": 4.834198961246504, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12100 + }, + { + "epoch": 4.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12101 + }, + { + "epoch": 4.834998002397123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12102 + }, + { + "epoch": 4.835397522972433, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12103 + }, + { + "epoch": 4.835797043547743, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12104 + }, + { + "epoch": 4.836196564123052, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12105 + }, + { + "epoch": 4.836596084698362, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12106 + }, + { + "epoch": 4.836995605273671, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12107 + }, + { + "epoch": 4.837395125848981, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12108 + }, + { + "epoch": 4.837794646424291, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12109 + }, + { + "epoch": 4.8381941669996005, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12110 + }, + { + "epoch": 4.83859368757491, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12111 + }, + { + "epoch": 4.83899320815022, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12112 + }, + { + "epoch": 4.83939272872553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12113 + }, + { + "epoch": 4.839792249300839, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12114 + }, + { + "epoch": 4.8401917698761485, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12115 + }, + { + "epoch": 4.840591290451458, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12116 + }, + { + "epoch": 4.840990811026768, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12117 + }, + { + "epoch": 4.841390331602078, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12118 + }, + { + "epoch": 4.841789852177387, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12119 + }, + { + "epoch": 4.842189372752697, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12120 + }, + { + "epoch": 4.842588893328006, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12121 + }, + { + "epoch": 4.842988413903316, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12122 + }, + { + "epoch": 4.843387934478626, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 12123 + }, + { + "epoch": 4.843787455053935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12124 + }, + { + "epoch": 4.844186975629245, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12125 + }, + { + "epoch": 4.844586496204554, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12126 + }, + { + "epoch": 4.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12127 + }, + { + "epoch": 4.845385537355174, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12128 + }, + { + "epoch": 4.845785057930484, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12129 + }, + { + "epoch": 4.8461845785057935, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 12130 + }, + { + "epoch": 4.846584099081102, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12131 + }, + { + "epoch": 4.846983619656412, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12132 + }, + { + "epoch": 4.847383140231722, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12133 + }, + { + "epoch": 4.847782660807032, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12134 + }, + { + "epoch": 4.848182181382342, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12135 + }, + { + "epoch": 4.8485817019576505, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12136 + }, + { + "epoch": 4.84898122253296, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 12137 + }, + { + "epoch": 4.84938074310827, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12138 + }, + { + "epoch": 4.84978026368358, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12139 + }, + { + "epoch": 4.85017978425889, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12140 + }, + { + "epoch": 4.8505793048341985, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12141 + }, + { + "epoch": 4.850978825409508, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12142 + }, + { + "epoch": 4.851378345984818, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12143 + }, + { + "epoch": 4.851777866560128, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12144 + }, + { + "epoch": 4.852177387135438, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12145 + }, + { + "epoch": 4.8525769077107475, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12146 + }, + { + "epoch": 4.852976428286056, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12147 + }, + { + "epoch": 4.853375948861366, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12148 + }, + { + "epoch": 4.853775469436676, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12149 + }, + { + "epoch": 4.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12150 + }, + { + "epoch": 4.8545745105872955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12151 + }, + { + "epoch": 4.854974031162605, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 12152 + }, + { + "epoch": 4.855373551737914, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12153 + }, + { + "epoch": 4.855773072313224, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12154 + }, + { + "epoch": 4.856172592888534, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12155 + }, + { + "epoch": 4.8565721134638435, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12156 + }, + { + "epoch": 4.856971634039153, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12157 + }, + { + "epoch": 4.857371154614462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12158 + }, + { + "epoch": 4.857770675189772, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12159 + }, + { + "epoch": 4.858170195765082, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12160 + }, + { + "epoch": 4.858569716340392, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12161 + }, + { + "epoch": 4.858969236915701, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12162 + }, + { + "epoch": 4.859368757491011, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12163 + }, + { + "epoch": 4.85976827806632, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12164 + }, + { + "epoch": 4.86016779864163, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12165 + }, + { + "epoch": 4.86056731921694, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12166 + }, + { + "epoch": 4.860966839792249, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 12167 + }, + { + "epoch": 4.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12168 + }, + { + "epoch": 4.861765880942869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12169 + }, + { + "epoch": 4.862165401518178, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12170 + }, + { + "epoch": 4.862564922093488, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12171 + }, + { + "epoch": 4.8629644426687975, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12172 + }, + { + "epoch": 4.863363963244107, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12173 + }, + { + "epoch": 4.863763483819417, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12174 + }, + { + "epoch": 4.864163004394726, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12175 + }, + { + "epoch": 4.864562524970036, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12176 + }, + { + "epoch": 4.8649620455453455, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 12177 + }, + { + "epoch": 4.865361566120655, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12178 + }, + { + "epoch": 4.865761086695965, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12179 + }, + { + "epoch": 4.866160607271274, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12180 + }, + { + "epoch": 4.866560127846584, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12181 + }, + { + "epoch": 4.8669596484218935, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12182 + }, + { + "epoch": 4.867359168997203, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12183 + }, + { + "epoch": 4.867758689572513, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12184 + }, + { + "epoch": 4.868158210147823, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 12185 + }, + { + "epoch": 4.868557730723133, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12186 + }, + { + "epoch": 4.868957251298442, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12187 + }, + { + "epoch": 4.869356771873751, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12188 + }, + { + "epoch": 4.869756292449061, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12189 + }, + { + "epoch": 4.870155813024371, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12190 + }, + { + "epoch": 4.870555333599681, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12191 + }, + { + "epoch": 4.87095485417499, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12192 + }, + { + "epoch": 4.871354374750299, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12193 + }, + { + "epoch": 4.871753895325609, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 12194 + }, + { + "epoch": 4.872153415900919, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12195 + }, + { + "epoch": 4.872552936476229, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 12196 + }, + { + "epoch": 4.872952457051538, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12197 + }, + { + "epoch": 4.8733519776268475, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12198 + }, + { + "epoch": 4.873751498202157, + "grad_norm": 0.1337890625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 12199 + }, + { + "epoch": 4.874151018777467, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12200 + }, + { + "epoch": 4.874550539352777, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 12201 + }, + { + "epoch": 4.874950059928087, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12202 + }, + { + "epoch": 4.875349580503396, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12203 + }, + { + "epoch": 4.875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12204 + }, + { + "epoch": 4.876148621654015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12205 + }, + { + "epoch": 4.876548142229325, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12206 + }, + { + "epoch": 4.876947662804635, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 12207 + }, + { + "epoch": 4.877347183379944, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12208 + }, + { + "epoch": 4.877746703955253, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12209 + }, + { + "epoch": 4.878146224530563, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12210 + }, + { + "epoch": 4.878545745105873, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12211 + }, + { + "epoch": 4.878945265681183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12212 + }, + { + "epoch": 4.8793447862564925, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12213 + }, + { + "epoch": 4.879744306831801, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12214 + }, + { + "epoch": 4.880143827407111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12215 + }, + { + "epoch": 4.880543347982421, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12216 + }, + { + "epoch": 4.880942868557731, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12217 + }, + { + "epoch": 4.8813423891330405, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.343, + "step": 12218 + }, + { + "epoch": 4.88174190970835, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12219 + }, + { + "epoch": 4.88214143028366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12220 + }, + { + "epoch": 4.882540950858969, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12221 + }, + { + "epoch": 4.882940471434279, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12222 + }, + { + "epoch": 4.883339992009589, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12223 + }, + { + "epoch": 4.883739512584898, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12224 + }, + { + "epoch": 4.884139033160208, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12225 + }, + { + "epoch": 4.884538553735517, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12226 + }, + { + "epoch": 4.884938074310827, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12227 + }, + { + "epoch": 4.885337594886137, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12228 + }, + { + "epoch": 4.885737115461446, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12229 + }, + { + "epoch": 4.886136636036756, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12230 + }, + { + "epoch": 4.886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12231 + }, + { + "epoch": 4.886935677187375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12232 + }, + { + "epoch": 4.887335197762685, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12233 + }, + { + "epoch": 4.887734718337994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12234 + }, + { + "epoch": 4.888134238913304, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12235 + }, + { + "epoch": 4.888533759488614, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12236 + }, + { + "epoch": 4.888933280063923, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12237 + }, + { + "epoch": 4.889332800639233, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12238 + }, + { + "epoch": 4.8897323212145425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12239 + }, + { + "epoch": 4.890131841789852, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12240 + }, + { + "epoch": 4.890531362365162, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12241 + }, + { + "epoch": 4.890930882940472, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12242 + }, + { + "epoch": 4.891330403515781, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12243 + }, + { + "epoch": 4.8917299240910905, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12244 + }, + { + "epoch": 4.8921294446664, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 12245 + }, + { + "epoch": 4.89252896524171, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12246 + }, + { + "epoch": 4.89292848581702, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12247 + }, + { + "epoch": 4.893328006392329, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12248 + }, + { + "epoch": 4.893727526967639, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12249 + }, + { + "epoch": 4.894127047542948, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12250 + }, + { + "epoch": 4.894526568118258, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12251 + }, + { + "epoch": 4.894926088693568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12252 + }, + { + "epoch": 4.895325609268878, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12253 + }, + { + "epoch": 4.895725129844187, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12254 + }, + { + "epoch": 4.896124650419496, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 12255 + }, + { + "epoch": 4.896524170994806, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12256 + }, + { + "epoch": 4.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12257 + }, + { + "epoch": 4.897323212145426, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12258 + }, + { + "epoch": 4.8977227327207356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12259 + }, + { + "epoch": 4.8981222532960444, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12260 + }, + { + "epoch": 4.898521773871354, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12261 + }, + { + "epoch": 4.898921294446664, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12262 + }, + { + "epoch": 4.899320815021974, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12263 + }, + { + "epoch": 4.899720335597284, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12264 + }, + { + "epoch": 4.9001198561725925, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12265 + }, + { + "epoch": 4.900519376747902, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 12266 + }, + { + "epoch": 4.900918897323212, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12267 + }, + { + "epoch": 4.901318417898522, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12268 + }, + { + "epoch": 4.901717938473832, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12269 + }, + { + "epoch": 4.9021174590491405, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12270 + }, + { + "epoch": 4.90251697962445, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 12271 + }, + { + "epoch": 4.90291650019976, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12272 + }, + { + "epoch": 4.90331602077507, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12273 + }, + { + "epoch": 4.90371554135038, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12274 + }, + { + "epoch": 4.9041150619256895, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 12275 + }, + { + "epoch": 4.904514582500999, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12276 + }, + { + "epoch": 4.904914103076308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 12277 + }, + { + "epoch": 4.905313623651618, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12278 + }, + { + "epoch": 4.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12279 + }, + { + "epoch": 4.9061126648022375, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12280 + }, + { + "epoch": 4.906512185377547, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12281 + }, + { + "epoch": 4.906911705952856, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12282 + }, + { + "epoch": 4.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 12283 + }, + { + "epoch": 4.907710747103476, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12284 + }, + { + "epoch": 4.9081102676787856, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12285 + }, + { + "epoch": 4.908509788254095, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12286 + }, + { + "epoch": 4.908909308829404, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12287 + }, + { + "epoch": 4.909308829404714, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12288 + }, + { + "epoch": 4.909708349980024, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12289 + }, + { + "epoch": 4.910107870555334, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12290 + }, + { + "epoch": 4.910507391130643, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12291 + }, + { + "epoch": 4.910906911705953, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12292 + }, + { + "epoch": 4.911306432281263, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12293 + }, + { + "epoch": 4.911705952856572, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 12294 + }, + { + "epoch": 4.912105473431882, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12295 + }, + { + "epoch": 4.912504994007191, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12296 + }, + { + "epoch": 4.912904514582501, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12297 + }, + { + "epoch": 4.913304035157811, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12298 + }, + { + "epoch": 4.91370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12299 + }, + { + "epoch": 4.91410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12300 + }, + { + "epoch": 4.9145025968837395, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12301 + }, + { + "epoch": 4.914902117459049, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12302 + }, + { + "epoch": 4.915301638034359, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12303 + }, + { + "epoch": 4.915701158609668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12304 + }, + { + "epoch": 4.916100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12305 + }, + { + "epoch": 4.9165001997602875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12306 + }, + { + "epoch": 4.916899720335597, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 12307 + }, + { + "epoch": 4.917299240910907, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12308 + }, + { + "epoch": 4.917698761486217, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12309 + }, + { + "epoch": 4.918098282061527, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12310 + }, + { + "epoch": 4.918497802636836, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12311 + }, + { + "epoch": 4.918897323212145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12312 + }, + { + "epoch": 4.919296843787455, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12313 + }, + { + "epoch": 4.919696364362765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12314 + }, + { + "epoch": 4.920095884938075, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12315 + }, + { + "epoch": 4.920495405513384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12316 + }, + { + "epoch": 4.920894926088693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12317 + }, + { + "epoch": 4.921294446664003, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12318 + }, + { + "epoch": 4.921693967239313, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12319 + }, + { + "epoch": 4.922093487814623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12320 + }, + { + "epoch": 4.922493008389932, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12321 + }, + { + "epoch": 4.922892528965241, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12322 + }, + { + "epoch": 4.923292049540551, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12323 + }, + { + "epoch": 4.923691570115861, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12324 + }, + { + "epoch": 4.924091090691171, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12325 + }, + { + "epoch": 4.924490611266481, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12326 + }, + { + "epoch": 4.9248901318417895, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12327 + }, + { + "epoch": 4.925289652417099, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12328 + }, + { + "epoch": 4.925689172992409, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12329 + }, + { + "epoch": 4.926088693567719, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12330 + }, + { + "epoch": 4.926488214143029, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2973, + "step": 12331 + }, + { + "epoch": 4.926887734718338, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12332 + }, + { + "epoch": 4.927287255293647, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 12333 + }, + { + "epoch": 4.927686775868957, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12334 + }, + { + "epoch": 4.928086296444267, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12335 + }, + { + "epoch": 4.928485817019577, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12336 + }, + { + "epoch": 4.9288853375948865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12337 + }, + { + "epoch": 4.929284858170195, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 12338 + }, + { + "epoch": 4.929684378745505, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12339 + }, + { + "epoch": 4.930083899320815, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12340 + }, + { + "epoch": 4.930483419896125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12341 + }, + { + "epoch": 4.9308829404714345, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 12342 + }, + { + "epoch": 4.931282461046744, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 12343 + }, + { + "epoch": 4.931681981622053, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12344 + }, + { + "epoch": 4.932081502197363, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12345 + }, + { + "epoch": 4.932481022772673, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12346 + }, + { + "epoch": 4.9328805433479825, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12347 + }, + { + "epoch": 4.933280063923292, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12348 + }, + { + "epoch": 4.933679584498602, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12349 + }, + { + "epoch": 4.934079105073911, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12350 + }, + { + "epoch": 4.934478625649221, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12351 + }, + { + "epoch": 4.934878146224531, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12352 + }, + { + "epoch": 4.93527766679984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12353 + }, + { + "epoch": 4.93567718737515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 12354 + }, + { + "epoch": 4.936076707950459, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 12355 + }, + { + "epoch": 4.936476228525769, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12356 + }, + { + "epoch": 4.936875749101079, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12357 + }, + { + "epoch": 4.937275269676388, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12358 + }, + { + "epoch": 4.937674790251698, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 12359 + }, + { + "epoch": 4.938074310827007, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12360 + }, + { + "epoch": 4.938473831402317, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12361 + }, + { + "epoch": 4.938873351977627, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12362 + }, + { + "epoch": 4.9392728725529365, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12363 + }, + { + "epoch": 4.939672393128246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12364 + }, + { + "epoch": 4.940071913703556, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 12365 + }, + { + "epoch": 4.940471434278866, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12366 + }, + { + "epoch": 4.940870954854175, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12367 + }, + { + "epoch": 4.9412704754294845, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12368 + }, + { + "epoch": 4.941669996004794, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12369 + }, + { + "epoch": 4.942069516580104, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12370 + }, + { + "epoch": 4.942469037155414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12371 + }, + { + "epoch": 4.942868557730723, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12372 + }, + { + "epoch": 4.9432680783060325, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12373 + }, + { + "epoch": 4.943667598881342, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12374 + }, + { + "epoch": 4.944067119456652, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 12375 + }, + { + "epoch": 4.944466640031962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12376 + }, + { + "epoch": 4.944866160607271, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12377 + }, + { + "epoch": 4.945265681182581, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12378 + }, + { + "epoch": 4.94566520175789, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12379 + }, + { + "epoch": 4.9460647223332, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12380 + }, + { + "epoch": 4.94646424290851, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12381 + }, + { + "epoch": 4.94686376348382, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12382 + }, + { + "epoch": 4.9472632840591295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12383 + }, + { + "epoch": 4.947662804634438, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12384 + }, + { + "epoch": 4.948062325209748, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12385 + }, + { + "epoch": 4.948461845785058, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 12386 + }, + { + "epoch": 4.948861366360368, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12387 + }, + { + "epoch": 4.949260886935678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12388 + }, + { + "epoch": 4.9496604075109865, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12389 + }, + { + "epoch": 4.950059928086296, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12390 + }, + { + "epoch": 4.950459448661606, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12391 + }, + { + "epoch": 4.950858969236916, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12392 + }, + { + "epoch": 4.951258489812226, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 12393 + }, + { + "epoch": 4.9516580103875345, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12394 + }, + { + "epoch": 4.952057530962844, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 12395 + }, + { + "epoch": 4.952457051538154, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12396 + }, + { + "epoch": 4.952856572113464, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12397 + }, + { + "epoch": 4.953256092688774, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 12398 + }, + { + "epoch": 4.953655613264083, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12399 + }, + { + "epoch": 4.954055133839393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12400 + }, + { + "epoch": 4.954454654414702, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12401 + }, + { + "epoch": 4.954854174990012, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12402 + }, + { + "epoch": 4.955253695565322, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12403 + }, + { + "epoch": 4.9556532161406315, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12404 + }, + { + "epoch": 4.956052736715941, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12405 + }, + { + "epoch": 4.95645225729125, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12406 + }, + { + "epoch": 4.95685177786656, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 12407 + }, + { + "epoch": 4.95725129844187, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12408 + }, + { + "epoch": 4.9576508190171795, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 12409 + }, + { + "epoch": 4.958050339592489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 12410 + }, + { + "epoch": 4.958449860167798, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12411 + }, + { + "epoch": 4.958849380743108, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 12412 + }, + { + "epoch": 4.959248901318418, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12413 + }, + { + "epoch": 4.959648421893728, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12414 + }, + { + "epoch": 4.960047942469037, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12415 + }, + { + "epoch": 4.960447463044347, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12416 + }, + { + "epoch": 4.960846983619656, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12417 + }, + { + "epoch": 4.961246504194966, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12418 + }, + { + "epoch": 4.961646024770276, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12419 + }, + { + "epoch": 4.962045545345585, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12420 + }, + { + "epoch": 4.962445065920895, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12421 + }, + { + "epoch": 4.962844586496205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12422 + }, + { + "epoch": 4.963244107071514, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12423 + }, + { + "epoch": 4.963643627646824, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12424 + }, + { + "epoch": 4.964043148222133, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12425 + }, + { + "epoch": 4.964442668797443, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 12426 + }, + { + "epoch": 4.964842189372753, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 12427 + }, + { + "epoch": 4.965241709948062, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12428 + }, + { + "epoch": 4.965641230523372, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12429 + }, + { + "epoch": 4.9660407510986815, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3363, + "step": 12430 + }, + { + "epoch": 4.966440271673991, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12431 + }, + { + "epoch": 4.966839792249301, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12432 + }, + { + "epoch": 4.967239312824611, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12433 + }, + { + "epoch": 4.96763883339992, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12434 + }, + { + "epoch": 4.9680383539752295, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 12435 + }, + { + "epoch": 4.968437874550539, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12436 + }, + { + "epoch": 4.968837395125849, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12437 + }, + { + "epoch": 4.969236915701159, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 12438 + }, + { + "epoch": 4.969636436276469, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12439 + }, + { + "epoch": 4.970035956851778, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12440 + }, + { + "epoch": 4.970435477427087, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12441 + }, + { + "epoch": 4.970834998002397, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12442 + }, + { + "epoch": 4.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12443 + }, + { + "epoch": 4.971634039153017, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12444 + }, + { + "epoch": 4.972033559728326, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12445 + }, + { + "epoch": 4.972433080303635, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12446 + }, + { + "epoch": 4.972832600878945, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12447 + }, + { + "epoch": 4.973232121454255, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12448 + }, + { + "epoch": 4.973631642029565, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12449 + }, + { + "epoch": 4.974031162604874, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12450 + }, + { + "epoch": 4.974430683180183, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12451 + }, + { + "epoch": 4.974830203755493, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12452 + }, + { + "epoch": 4.975229724330803, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 12453 + }, + { + "epoch": 4.975629244906113, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12454 + }, + { + "epoch": 4.976028765481423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 12455 + }, + { + "epoch": 4.976428286056732, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12456 + }, + { + "epoch": 4.976827806632041, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12457 + }, + { + "epoch": 4.977227327207351, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12458 + }, + { + "epoch": 4.977626847782661, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12459 + }, + { + "epoch": 4.978026368357971, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 12460 + }, + { + "epoch": 4.97842588893328, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12461 + }, + { + "epoch": 4.978825409508589, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 12462 + }, + { + "epoch": 4.979224930083899, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12463 + }, + { + "epoch": 4.979624450659209, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 12464 + }, + { + "epoch": 4.980023971234519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 12465 + }, + { + "epoch": 4.9804234918098285, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12466 + }, + { + "epoch": 4.980823012385137, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12467 + }, + { + "epoch": 4.981222532960447, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12468 + }, + { + "epoch": 4.981622053535757, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12469 + }, + { + "epoch": 4.982021574111067, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12470 + }, + { + "epoch": 4.9824210946863765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12471 + }, + { + "epoch": 4.982820615261686, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12472 + }, + { + "epoch": 4.983220135836996, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12473 + }, + { + "epoch": 4.983619656412305, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12474 + }, + { + "epoch": 4.984019176987615, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12475 + }, + { + "epoch": 4.9844186975629245, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 12476 + }, + { + "epoch": 4.984818218138234, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 12477 + }, + { + "epoch": 4.985217738713544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12478 + }, + { + "epoch": 4.985617259288853, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12479 + }, + { + "epoch": 4.986016779864163, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12480 + }, + { + "epoch": 4.986416300439473, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 12481 + }, + { + "epoch": 4.986815821014782, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12482 + }, + { + "epoch": 4.987215341590092, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12483 + }, + { + "epoch": 4.987614862165401, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12484 + }, + { + "epoch": 4.988014382740711, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12485 + }, + { + "epoch": 4.988413903316021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12486 + }, + { + "epoch": 4.98881342389133, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12487 + }, + { + "epoch": 4.98921294446664, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12488 + }, + { + "epoch": 4.98961246504195, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12489 + }, + { + "epoch": 4.99001198561726, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12490 + }, + { + "epoch": 4.990411506192569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12491 + }, + { + "epoch": 4.9908110267678785, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12492 + }, + { + "epoch": 4.991210547343188, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12493 + }, + { + "epoch": 4.991610067918498, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12494 + }, + { + "epoch": 4.992009588493808, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3382, + "step": 12495 + }, + { + "epoch": 4.992409109069117, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12496 + }, + { + "epoch": 4.9928086296444265, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12497 + }, + { + "epoch": 4.993208150219736, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 12498 + }, + { + "epoch": 4.993607670795046, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12499 + }, + { + "epoch": 4.994007191370356, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 12500 + }, + { + "epoch": 4.994406711945665, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12501 + }, + { + "epoch": 4.9948062325209746, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 12502 + }, + { + "epoch": 4.995205753096284, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12503 + }, + { + "epoch": 4.995605273671594, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12504 + }, + { + "epoch": 4.996004794246904, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12505 + }, + { + "epoch": 4.996404314822214, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12506 + }, + { + "epoch": 4.996803835397523, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12507 + }, + { + "epoch": 4.997203355972832, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12508 + }, + { + "epoch": 4.997602876548142, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12509 + }, + { + "epoch": 4.998002397123452, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12510 + }, + { + "epoch": 4.998401917698762, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 12511 + }, + { + "epoch": 4.9988014382740715, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12512 + }, + { + "epoch": 4.99920095884938, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12513 + }, + { + "epoch": 4.99960047942469, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12514 + }, + { + "epoch": 5.0, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12515 + }, + { + "epoch": 5.00039952057531, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12516 + }, + { + "epoch": 5.00079904115062, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12517 + }, + { + "epoch": 5.0011985617259285, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12518 + }, + { + "epoch": 5.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12519 + }, + { + "epoch": 5.001997602876548, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12520 + }, + { + "epoch": 5.002397123451858, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12521 + }, + { + "epoch": 5.002796644027168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12522 + }, + { + "epoch": 5.003196164602477, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12523 + }, + { + "epoch": 5.003595685177786, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12524 + }, + { + "epoch": 5.003995205753096, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 12525 + }, + { + "epoch": 5.004394726328406, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12526 + }, + { + "epoch": 5.004794246903716, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12527 + }, + { + "epoch": 5.0051937674790254, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12528 + }, + { + "epoch": 5.005593288054335, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12529 + }, + { + "epoch": 5.005992808629644, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12530 + }, + { + "epoch": 5.006392329204954, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12531 + }, + { + "epoch": 5.006791849780264, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 12532 + }, + { + "epoch": 5.0071913703555735, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12533 + }, + { + "epoch": 5.007590890930883, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12534 + }, + { + "epoch": 5.007990411506192, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 12535 + }, + { + "epoch": 5.008389932081502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12536 + }, + { + "epoch": 5.008789452656812, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12537 + }, + { + "epoch": 5.0091889732321215, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12538 + }, + { + "epoch": 5.009588493807431, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12539 + }, + { + "epoch": 5.009988014382741, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12540 + }, + { + "epoch": 5.01038753495805, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12541 + }, + { + "epoch": 5.01078705553336, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12542 + }, + { + "epoch": 5.01118657610867, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12543 + }, + { + "epoch": 5.011586096683979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12544 + }, + { + "epoch": 5.011985617259289, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12545 + }, + { + "epoch": 5.012385137834598, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12546 + }, + { + "epoch": 5.012784658409908, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12547 + }, + { + "epoch": 5.013184178985218, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12548 + }, + { + "epoch": 5.013583699560527, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12549 + }, + { + "epoch": 5.013983220135837, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 12550 + }, + { + "epoch": 5.014382740711147, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12551 + }, + { + "epoch": 5.014782261286456, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12552 + }, + { + "epoch": 5.015181781861766, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 12553 + }, + { + "epoch": 5.0155813024370755, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12554 + }, + { + "epoch": 5.015980823012385, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12555 + }, + { + "epoch": 5.016380343587695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12556 + }, + { + "epoch": 5.016779864163005, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12557 + }, + { + "epoch": 5.017179384738314, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12558 + }, + { + "epoch": 5.0175789053136235, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12559 + }, + { + "epoch": 5.017978425888933, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12560 + }, + { + "epoch": 5.018377946464243, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12561 + }, + { + "epoch": 5.018777467039553, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12562 + }, + { + "epoch": 5.019176987614862, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12563 + }, + { + "epoch": 5.0195765081901715, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12564 + }, + { + "epoch": 5.019976028765481, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 12565 + }, + { + "epoch": 5.020375549340791, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12566 + }, + { + "epoch": 5.020775069916101, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12567 + }, + { + "epoch": 5.021174590491411, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12568 + }, + { + "epoch": 5.02157411106672, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12569 + }, + { + "epoch": 5.021973631642029, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 12570 + }, + { + "epoch": 5.022373152217339, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12571 + }, + { + "epoch": 5.022772672792649, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12572 + }, + { + "epoch": 5.023172193367959, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12573 + }, + { + "epoch": 5.0235717139432685, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12574 + }, + { + "epoch": 5.023971234518577, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12575 + }, + { + "epoch": 5.024370755093887, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12576 + }, + { + "epoch": 5.024770275669197, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12577 + }, + { + "epoch": 5.025169796244507, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12578 + }, + { + "epoch": 5.025569316819817, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12579 + }, + { + "epoch": 5.0259688373951255, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12580 + }, + { + "epoch": 5.026368357970435, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12581 + }, + { + "epoch": 5.026767878545745, + "grad_norm": 0.12109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12582 + }, + { + "epoch": 5.027167399121055, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12583 + }, + { + "epoch": 5.027566919696365, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12584 + }, + { + "epoch": 5.027966440271674, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12585 + }, + { + "epoch": 5.028365960846983, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12586 + }, + { + "epoch": 5.028765481422293, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12587 + }, + { + "epoch": 5.029165001997603, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 12588 + }, + { + "epoch": 5.029564522572913, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12589 + }, + { + "epoch": 5.029964043148222, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12590 + }, + { + "epoch": 5.030363563723531, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 12591 + }, + { + "epoch": 5.030763084298841, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12592 + }, + { + "epoch": 5.031162604874151, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12593 + }, + { + "epoch": 5.031562125449461, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12594 + }, + { + "epoch": 5.0319616460247705, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12595 + }, + { + "epoch": 5.03236116660008, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 12596 + }, + { + "epoch": 5.032760687175389, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12597 + }, + { + "epoch": 5.033160207750699, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 12598 + }, + { + "epoch": 5.033559728326009, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12599 + }, + { + "epoch": 5.0339592489013185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12600 + }, + { + "epoch": 5.034358769476628, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 12601 + }, + { + "epoch": 5.034758290051938, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12602 + }, + { + "epoch": 5.035157810627247, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12603 + }, + { + "epoch": 5.035557331202557, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 12604 + }, + { + "epoch": 5.035956851777867, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 12605 + }, + { + "epoch": 5.036356372353176, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12606 + }, + { + "epoch": 5.036755892928486, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12607 + }, + { + "epoch": 5.037155413503795, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 12608 + }, + { + "epoch": 5.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12609 + }, + { + "epoch": 5.037954454654415, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 12610 + }, + { + "epoch": 5.038353975229724, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12611 + }, + { + "epoch": 5.038753495805034, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12612 + }, + { + "epoch": 5.039153016380344, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12613 + }, + { + "epoch": 5.039552536955653, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12614 + }, + { + "epoch": 5.039952057530963, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12615 + }, + { + "epoch": 5.040351578106272, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 12616 + }, + { + "epoch": 5.040751098681582, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12617 + }, + { + "epoch": 5.041150619256892, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 12618 + }, + { + "epoch": 5.041550139832202, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12619 + }, + { + "epoch": 5.041949660407511, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12620 + }, + { + "epoch": 5.0423491809828205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12621 + }, + { + "epoch": 5.04274870155813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12622 + }, + { + "epoch": 5.04314822213344, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 12623 + }, + { + "epoch": 5.04354774270875, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12624 + }, + { + "epoch": 5.043947263284059, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 12625 + }, + { + "epoch": 5.0443467838593685, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 12626 + }, + { + "epoch": 5.044746304434678, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12627 + }, + { + "epoch": 5.045145825009988, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12628 + }, + { + "epoch": 5.045545345585298, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12629 + }, + { + "epoch": 5.045944866160608, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12630 + }, + { + "epoch": 5.046344386735917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12631 + }, + { + "epoch": 5.046743907311226, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12632 + }, + { + "epoch": 5.047143427886536, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12633 + }, + { + "epoch": 5.047542948461846, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12634 + }, + { + "epoch": 5.047942469037156, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12635 + }, + { + "epoch": 5.048341989612465, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12636 + }, + { + "epoch": 5.048741510187774, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12637 + }, + { + "epoch": 5.049141030763084, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12638 + }, + { + "epoch": 5.049540551338394, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12639 + }, + { + "epoch": 5.049940071913704, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12640 + }, + { + "epoch": 5.0503395924890135, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12641 + }, + { + "epoch": 5.050739113064322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12642 + }, + { + "epoch": 5.051138633639632, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12643 + }, + { + "epoch": 5.051538154214942, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 12644 + }, + { + "epoch": 5.051937674790252, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 12645 + }, + { + "epoch": 5.052337195365562, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12646 + }, + { + "epoch": 5.052736715940871, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12647 + }, + { + "epoch": 5.05313623651618, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12648 + }, + { + "epoch": 5.05353575709149, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12649 + }, + { + "epoch": 5.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12650 + }, + { + "epoch": 5.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12651 + }, + { + "epoch": 5.054734318817419, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12652 + }, + { + "epoch": 5.055133839392728, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12653 + }, + { + "epoch": 5.055533359968038, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12654 + }, + { + "epoch": 5.055932880543348, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12655 + }, + { + "epoch": 5.056332401118658, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12656 + }, + { + "epoch": 5.0567319216939675, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12657 + }, + { + "epoch": 5.057131442269277, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12658 + }, + { + "epoch": 5.057530962844586, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12659 + }, + { + "epoch": 5.057930483419896, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 12660 + }, + { + "epoch": 5.058330003995206, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12661 + }, + { + "epoch": 5.0587295245705155, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 12662 + }, + { + "epoch": 5.059129045145825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12663 + }, + { + "epoch": 5.059528565721135, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12664 + }, + { + "epoch": 5.059928086296444, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12665 + }, + { + "epoch": 5.060327606871754, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12666 + }, + { + "epoch": 5.0607271274470635, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12667 + }, + { + "epoch": 5.061126648022373, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12668 + }, + { + "epoch": 5.061526168597683, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 12669 + }, + { + "epoch": 5.061925689172992, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 12670 + }, + { + "epoch": 5.062325209748302, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 12671 + }, + { + "epoch": 5.062724730323612, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12672 + }, + { + "epoch": 5.063124250898921, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 12673 + }, + { + "epoch": 5.063523771474231, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 12674 + }, + { + "epoch": 5.063923292049541, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12675 + }, + { + "epoch": 5.06432281262485, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12676 + }, + { + "epoch": 5.06472233320016, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12677 + }, + { + "epoch": 5.065121853775469, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12678 + }, + { + "epoch": 5.065521374350779, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12679 + }, + { + "epoch": 5.065920894926089, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12680 + }, + { + "epoch": 5.066320415501398, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12681 + }, + { + "epoch": 5.066719936076708, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12682 + }, + { + "epoch": 5.0671194566520175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 12683 + }, + { + "epoch": 5.067518977227327, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12684 + }, + { + "epoch": 5.067918497802637, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 12685 + }, + { + "epoch": 5.068318018377947, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12686 + }, + { + "epoch": 5.068717538953256, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12687 + }, + { + "epoch": 5.0691170595285655, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12688 + }, + { + "epoch": 5.069516580103875, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12689 + }, + { + "epoch": 5.069916100679185, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12690 + }, + { + "epoch": 5.070315621254495, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 12691 + }, + { + "epoch": 5.070715141829805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12692 + }, + { + "epoch": 5.0711146624051135, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12693 + }, + { + "epoch": 5.071514182980423, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12694 + }, + { + "epoch": 5.071913703555733, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12695 + }, + { + "epoch": 5.072313224131043, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 12696 + }, + { + "epoch": 5.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12697 + }, + { + "epoch": 5.073112265281662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12698 + }, + { + "epoch": 5.073511785856971, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12699 + }, + { + "epoch": 5.073911306432281, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12700 + }, + { + "epoch": 5.074310827007591, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12701 + }, + { + "epoch": 5.074710347582901, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3348, + "step": 12702 + }, + { + "epoch": 5.0751098681582105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12703 + }, + { + "epoch": 5.075509388733519, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12704 + }, + { + "epoch": 5.075908909308829, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12705 + }, + { + "epoch": 5.076308429884139, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12706 + }, + { + "epoch": 5.076707950459449, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 12707 + }, + { + "epoch": 5.077107471034759, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12708 + }, + { + "epoch": 5.077506991610068, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12709 + }, + { + "epoch": 5.077906512185377, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 12710 + }, + { + "epoch": 5.078306032760687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12711 + }, + { + "epoch": 5.078705553335997, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12712 + }, + { + "epoch": 5.079105073911307, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12713 + }, + { + "epoch": 5.079504594486616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12714 + }, + { + "epoch": 5.079904115061925, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 12715 + }, + { + "epoch": 5.080303635637235, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12716 + }, + { + "epoch": 5.080703156212545, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12717 + }, + { + "epoch": 5.081102676787855, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 12718 + }, + { + "epoch": 5.081502197363164, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12719 + }, + { + "epoch": 5.081901717938474, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12720 + }, + { + "epoch": 5.082301238513783, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 12721 + }, + { + "epoch": 5.082700759089093, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12722 + }, + { + "epoch": 5.083100279664403, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 12723 + }, + { + "epoch": 5.0834998002397125, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12724 + }, + { + "epoch": 5.083899320815022, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 12725 + }, + { + "epoch": 5.084298841390331, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 12726 + }, + { + "epoch": 5.084698361965641, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 12727 + }, + { + "epoch": 5.085097882540951, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12728 + }, + { + "epoch": 5.0854974031162605, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12729 + }, + { + "epoch": 5.08589692369157, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12730 + }, + { + "epoch": 5.08629644426688, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 12731 + }, + { + "epoch": 5.086695964842189, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12732 + }, + { + "epoch": 5.087095485417499, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12733 + }, + { + "epoch": 5.087495005992809, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 12734 + }, + { + "epoch": 5.087894526568118, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12735 + }, + { + "epoch": 5.088294047143428, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12736 + }, + { + "epoch": 5.088693567718738, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 12737 + }, + { + "epoch": 5.089093088294047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 12738 + }, + { + "epoch": 5.089492608869357, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 12739 + }, + { + "epoch": 5.089892129444666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12740 + }, + { + "epoch": 5.090291650019976, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12741 + }, + { + "epoch": 5.090691170595286, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12742 + }, + { + "epoch": 5.091090691170595, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 12743 + }, + { + "epoch": 5.091490211745905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12744 + }, + { + "epoch": 5.0918897323212144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 12745 + }, + { + "epoch": 5.092289252896524, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12746 + }, + { + "epoch": 5.092688773471834, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12747 + }, + { + "epoch": 5.093088294047144, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 12748 + }, + { + "epoch": 5.093487814622453, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 12749 + }, + { + "epoch": 5.0938873351977625, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 12750 + }, + { + "epoch": 5.094286855773072, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 12751 + }, + { + "epoch": 5.094686376348382, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 12752 + }, + { + "epoch": 5.095085896923692, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12753 + }, + { + "epoch": 5.095485417499002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 12754 + }, + { + "epoch": 5.0958849380743105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 12755 + }, + { + "epoch": 5.09628445864962, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12756 + }, + { + "epoch": 5.09668397922493, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12757 + }, + { + "epoch": 5.09708349980024, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 12758 + }, + { + "epoch": 5.09748302037555, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12759 + }, + { + "epoch": 5.097882540950859, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12760 + }, + { + "epoch": 5.098282061526168, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 12761 + }, + { + "epoch": 5.098681582101478, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12762 + }, + { + "epoch": 5.099081102676788, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 12763 + }, + { + "epoch": 5.099480623252098, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12764 + }, + { + "epoch": 5.0998801438274075, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12765 + }, + { + "epoch": 5.100279664402716, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 12766 + }, + { + "epoch": 5.100679184978026, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12767 + }, + { + "epoch": 5.101078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 12768 + }, + { + "epoch": 5.101478226128646, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12769 + }, + { + "epoch": 5.1018777467039556, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12770 + }, + { + "epoch": 5.1022772672792644, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12771 + }, + { + "epoch": 5.102676787854574, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 12772 + }, + { + "epoch": 5.103076308429884, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12773 + }, + { + "epoch": 5.103475829005194, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 12774 + }, + { + "epoch": 5.103875349580504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12775 + }, + { + "epoch": 5.104274870155813, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 12776 + }, + { + "epoch": 5.104674390731122, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12777 + }, + { + "epoch": 5.105073911306432, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 12778 + }, + { + "epoch": 5.105473431881742, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12779 + }, + { + "epoch": 5.105872952457052, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 12780 + }, + { + "epoch": 5.106272473032361, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 12781 + }, + { + "epoch": 5.106671993607671, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12782 + }, + { + "epoch": 5.10707151418298, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 12783 + }, + { + "epoch": 5.10747103475829, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12784 + }, + { + "epoch": 5.1078705553336, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 12785 + }, + { + "epoch": 5.1082700759089095, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 12786 + }, + { + "epoch": 5.108669596484219, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12787 + }, + { + "epoch": 5.109069117059528, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12788 + }, + { + "epoch": 5.109468637634838, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12789 + }, + { + "epoch": 5.109868158210148, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12790 + }, + { + "epoch": 5.1102676787854575, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12791 + }, + { + "epoch": 5.110667199360767, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12792 + }, + { + "epoch": 5.111066719936077, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12793 + }, + { + "epoch": 5.111466240511386, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12794 + }, + { + "epoch": 5.111865761086696, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12795 + }, + { + "epoch": 5.112265281662006, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12796 + }, + { + "epoch": 5.112664802237315, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12797 + }, + { + "epoch": 5.113064322812625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 12798 + }, + { + "epoch": 5.113463843387935, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12799 + }, + { + "epoch": 5.113863363963244, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 12800 + }, + { + "epoch": 5.114262884538554, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12801 + }, + { + "epoch": 5.114662405113863, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 12802 + }, + { + "epoch": 5.115061925689173, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12803 + }, + { + "epoch": 5.115461446264483, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12804 + }, + { + "epoch": 5.115860966839792, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 12805 + }, + { + "epoch": 5.116260487415102, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 12806 + }, + { + "epoch": 5.116660007990411, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 12807 + }, + { + "epoch": 5.117059528565721, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 12808 + }, + { + "epoch": 5.117459049141031, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12809 + }, + { + "epoch": 5.117858569716341, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 12810 + }, + { + "epoch": 5.11825809029165, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12811 + }, + { + "epoch": 5.1186576108669595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 12812 + }, + { + "epoch": 5.119057131442269, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 12813 + }, + { + "epoch": 5.119456652017579, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12814 + }, + { + "epoch": 5.119856172592889, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 12815 + }, + { + "epoch": 5.120255693168198, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12816 + }, + { + "epoch": 5.1206552137435075, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12817 + }, + { + "epoch": 5.121054734318817, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12818 + }, + { + "epoch": 5.121454254894127, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 12819 + }, + { + "epoch": 5.121853775469437, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 12820 + }, + { + "epoch": 5.122253296044747, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3011, + "step": 12821 + }, + { + "epoch": 5.122652816620056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 12822 + }, + { + "epoch": 5.123052337195365, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12823 + }, + { + "epoch": 5.123451857770675, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12824 + }, + { + "epoch": 5.123851378345985, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12825 + }, + { + "epoch": 5.124250898921295, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 12826 + }, + { + "epoch": 5.1246504194966045, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12827 + }, + { + "epoch": 5.125049940071913, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12828 + }, + { + "epoch": 5.125449460647223, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 12829 + }, + { + "epoch": 5.125848981222533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12830 + }, + { + "epoch": 5.126248501797843, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12831 + }, + { + "epoch": 5.1266480223731525, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 12832 + }, + { + "epoch": 5.127047542948461, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 12833 + }, + { + "epoch": 5.127447063523771, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12834 + }, + { + "epoch": 5.127846584099081, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 12835 + }, + { + "epoch": 5.128246104674391, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12836 + }, + { + "epoch": 5.128645625249701, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12837 + }, + { + "epoch": 5.12904514582501, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12838 + }, + { + "epoch": 5.129444666400319, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 12839 + }, + { + "epoch": 5.129844186975629, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 12840 + }, + { + "epoch": 5.130243707550939, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 12841 + }, + { + "epoch": 5.130643228126249, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 12842 + }, + { + "epoch": 5.131042748701558, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 12843 + }, + { + "epoch": 5.131442269276867, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 12844 + }, + { + "epoch": 5.131841789852177, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 12845 + }, + { + "epoch": 5.132241310427487, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12846 + }, + { + "epoch": 5.132640831002797, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 12847 + }, + { + "epoch": 5.1330403515781065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 12848 + }, + { + "epoch": 5.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 12849 + }, + { + "epoch": 5.133839392728725, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 12850 + }, + { + "epoch": 5.134238913304035, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 12851 + }, + { + "epoch": 5.134638433879345, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 12852 + }, + { + "epoch": 5.1350379544546545, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12853 + }, + { + "epoch": 5.135437475029964, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 12854 + }, + { + "epoch": 5.135836995605274, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 12855 + }, + { + "epoch": 5.136236516180583, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12856 + }, + { + "epoch": 5.136636036755893, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 12857 + }, + { + "epoch": 5.1370355573312025, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12858 + }, + { + "epoch": 5.137435077906512, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12859 + }, + { + "epoch": 5.137834598481822, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12860 + }, + { + "epoch": 5.138234119057131, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 12861 + }, + { + "epoch": 5.138633639632441, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 12862 + }, + { + "epoch": 5.139033160207751, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 12863 + }, + { + "epoch": 5.13943268078306, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 12864 + }, + { + "epoch": 5.13983220135837, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 12865 + }, + { + "epoch": 5.14023172193368, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 12866 + }, + { + "epoch": 5.140631242508989, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 12867 + }, + { + "epoch": 5.141030763084299, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12868 + }, + { + "epoch": 5.141430283659608, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 12869 + }, + { + "epoch": 5.141829804234918, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 12870 + }, + { + "epoch": 5.142229324810228, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 12871 + }, + { + "epoch": 5.142628845385538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 12872 + }, + { + "epoch": 5.143028365960847, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12873 + }, + { + "epoch": 5.1434278865361565, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12874 + }, + { + "epoch": 5.143827407111466, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 12875 + }, + { + "epoch": 5.144226927686776, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 12876 + }, + { + "epoch": 5.144626448262086, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 12877 + }, + { + "epoch": 5.145025968837395, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12878 + }, + { + "epoch": 5.1454254894127045, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 12879 + }, + { + "epoch": 5.145825009988014, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 12880 + }, + { + "epoch": 5.146224530563324, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 12881 + }, + { + "epoch": 5.146624051138634, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 12882 + }, + { + "epoch": 5.147023571713944, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 12883 + }, + { + "epoch": 5.1474230922892525, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 12884 + }, + { + "epoch": 5.147822612864562, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 12885 + }, + { + "epoch": 5.148222133439872, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 12886 + }, + { + "epoch": 5.148621654015182, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12887 + }, + { + "epoch": 5.149021174590492, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 12888 + }, + { + "epoch": 5.1494206951658015, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 12889 + }, + { + "epoch": 5.14982021574111, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 12890 + }, + { + "epoch": 5.15021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 12891 + }, + { + "epoch": 5.15061925689173, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 12892 + }, + { + "epoch": 5.15101877746704, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12893 + }, + { + "epoch": 5.1514182980423495, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 12894 + }, + { + "epoch": 5.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12895 + }, + { + "epoch": 5.152217339192968, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12896 + }, + { + "epoch": 5.152616859768278, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 12897 + }, + { + "epoch": 5.153016380343588, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 12898 + }, + { + "epoch": 5.153415900918898, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 12899 + }, + { + "epoch": 5.153815421494207, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 12900 + }, + { + "epoch": 5.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 12901 + }, + { + "epoch": 5.154614462644826, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 12902 + }, + { + "epoch": 5.155013983220136, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12903 + }, + { + "epoch": 5.155413503795446, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 12904 + }, + { + "epoch": 5.155813024370755, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12905 + }, + { + "epoch": 5.156212544946065, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 12906 + }, + { + "epoch": 5.156612065521374, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12907 + }, + { + "epoch": 5.157011586096684, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 12908 + }, + { + "epoch": 5.157411106671994, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 12909 + }, + { + "epoch": 5.157810627247303, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12910 + }, + { + "epoch": 5.158210147822613, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12911 + }, + { + "epoch": 5.158609668397922, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12912 + }, + { + "epoch": 5.159009188973232, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 12913 + }, + { + "epoch": 5.159408709548542, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 12914 + }, + { + "epoch": 5.1598082301238515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 12915 + }, + { + "epoch": 5.160207750699161, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 12916 + }, + { + "epoch": 5.160607271274471, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 12917 + }, + { + "epoch": 5.16100679184978, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 12918 + }, + { + "epoch": 5.16140631242509, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12919 + }, + { + "epoch": 5.1618058330003995, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 12920 + }, + { + "epoch": 5.162205353575709, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 12921 + }, + { + "epoch": 5.162604874151019, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12922 + }, + { + "epoch": 5.163004394726328, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 12923 + }, + { + "epoch": 5.163403915301638, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 12924 + }, + { + "epoch": 5.163803435876948, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 12925 + }, + { + "epoch": 5.164202956452257, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 12926 + }, + { + "epoch": 5.164602477027567, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 12927 + }, + { + "epoch": 5.165001997602877, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12928 + }, + { + "epoch": 5.165401518178186, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 12929 + }, + { + "epoch": 5.165801038753496, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 12930 + }, + { + "epoch": 5.166200559328805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 12931 + }, + { + "epoch": 5.166600079904115, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 12932 + }, + { + "epoch": 5.166999600479425, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 12933 + }, + { + "epoch": 5.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12934 + }, + { + "epoch": 5.167798641630044, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 12935 + }, + { + "epoch": 5.168198162205353, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 12936 + }, + { + "epoch": 5.168597682780663, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12937 + }, + { + "epoch": 5.168997203355973, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 12938 + }, + { + "epoch": 5.169396723931283, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 12939 + }, + { + "epoch": 5.169796244506592, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 12940 + }, + { + "epoch": 5.1701957650819015, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12941 + }, + { + "epoch": 5.170595285657211, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 12942 + }, + { + "epoch": 5.170994806232521, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 12943 + }, + { + "epoch": 5.171394326807831, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3319, + "step": 12944 + }, + { + "epoch": 5.171793847383141, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 12945 + }, + { + "epoch": 5.1721933679584495, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 12946 + }, + { + "epoch": 5.172592888533759, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 12947 + }, + { + "epoch": 5.172992409109069, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 12948 + }, + { + "epoch": 5.173391929684379, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 12949 + }, + { + "epoch": 5.173791450259689, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 12950 + }, + { + "epoch": 5.174190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12951 + }, + { + "epoch": 5.174590491410307, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 12952 + }, + { + "epoch": 5.174990011985617, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12953 + }, + { + "epoch": 5.175389532560927, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 12954 + }, + { + "epoch": 5.175789053136237, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 12955 + }, + { + "epoch": 5.1761885737115465, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12956 + }, + { + "epoch": 5.176588094286855, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 12957 + }, + { + "epoch": 5.176987614862165, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12958 + }, + { + "epoch": 5.177387135437475, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 12959 + }, + { + "epoch": 5.177786656012785, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 12960 + }, + { + "epoch": 5.1781861765880945, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 12961 + }, + { + "epoch": 5.178585697163404, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 12962 + }, + { + "epoch": 5.178985217738713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 12963 + }, + { + "epoch": 5.179384738314023, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 12964 + }, + { + "epoch": 5.179784258889333, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 12965 + }, + { + "epoch": 5.180183779464643, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 12966 + }, + { + "epoch": 5.180583300039952, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 12967 + }, + { + "epoch": 5.180982820615261, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12968 + }, + { + "epoch": 5.181382341190571, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12969 + }, + { + "epoch": 5.181781861765881, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12970 + }, + { + "epoch": 5.182181382341191, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12971 + }, + { + "epoch": 5.1825809029165, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 12972 + }, + { + "epoch": 5.18298042349181, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 12973 + }, + { + "epoch": 5.183379944067119, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 12974 + }, + { + "epoch": 5.183779464642429, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 12975 + }, + { + "epoch": 5.184178985217739, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 12976 + }, + { + "epoch": 5.1845785057930485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12977 + }, + { + "epoch": 5.184978026368358, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 12978 + }, + { + "epoch": 5.185377546943668, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 12979 + }, + { + "epoch": 5.185777067518977, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 12980 + }, + { + "epoch": 5.186176588094287, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 12981 + }, + { + "epoch": 5.1865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12982 + }, + { + "epoch": 5.186975629244906, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 12983 + }, + { + "epoch": 5.187375149820216, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12984 + }, + { + "epoch": 5.187774670395525, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 12985 + }, + { + "epoch": 5.188174190970835, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 12986 + }, + { + "epoch": 5.1885737115461446, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 12987 + }, + { + "epoch": 5.188973232121454, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 12988 + }, + { + "epoch": 5.189372752696764, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 12989 + }, + { + "epoch": 5.189772273272074, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 12990 + }, + { + "epoch": 5.190171793847383, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 12991 + }, + { + "epoch": 5.190571314422693, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 12992 + }, + { + "epoch": 5.190970834998002, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 12993 + }, + { + "epoch": 5.191370355573312, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 12994 + }, + { + "epoch": 5.191769876148622, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 12995 + }, + { + "epoch": 5.192169396723932, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 12996 + }, + { + "epoch": 5.192568917299241, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 12997 + }, + { + "epoch": 5.19296843787455, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 12998 + }, + { + "epoch": 5.19336795844986, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 12999 + }, + { + "epoch": 5.19376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13000 + }, + { + "epoch": 5.19416699960048, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13001 + }, + { + "epoch": 5.194566520175789, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13002 + }, + { + "epoch": 5.1949660407510985, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13003 + }, + { + "epoch": 5.195365561326408, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13004 + }, + { + "epoch": 5.195765081901718, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13005 + }, + { + "epoch": 5.196164602477028, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13006 + }, + { + "epoch": 5.196564123052338, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13007 + }, + { + "epoch": 5.1969636436276465, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13008 + }, + { + "epoch": 5.197363164202956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13009 + }, + { + "epoch": 5.197762684778266, + "grad_norm": 0.1181640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13010 + }, + { + "epoch": 5.198162205353576, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13011 + }, + { + "epoch": 5.198561725928886, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13012 + }, + { + "epoch": 5.1989612465041946, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13013 + }, + { + "epoch": 5.199360767079504, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13014 + }, + { + "epoch": 5.199760287654814, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13015 + }, + { + "epoch": 5.200159808230124, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13016 + }, + { + "epoch": 5.200559328805434, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13017 + }, + { + "epoch": 5.2009588493807435, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13018 + }, + { + "epoch": 5.201358369956052, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13019 + }, + { + "epoch": 5.201757890531362, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13020 + }, + { + "epoch": 5.202157411106672, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13021 + }, + { + "epoch": 5.202556931681982, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13022 + }, + { + "epoch": 5.2029564522572915, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 13023 + }, + { + "epoch": 5.2033559728326, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13024 + }, + { + "epoch": 5.20375549340791, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13025 + }, + { + "epoch": 5.20415501398322, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13026 + }, + { + "epoch": 5.20455453455853, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 13027 + }, + { + "epoch": 5.20495405513384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13028 + }, + { + "epoch": 5.205353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13029 + }, + { + "epoch": 5.205753096284458, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13030 + }, + { + "epoch": 5.206152616859768, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13031 + }, + { + "epoch": 5.206552137435078, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13032 + }, + { + "epoch": 5.206951658010388, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13033 + }, + { + "epoch": 5.207351178585697, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13034 + }, + { + "epoch": 5.207750699161007, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13035 + }, + { + "epoch": 5.208150219736316, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13036 + }, + { + "epoch": 5.208549740311626, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13037 + }, + { + "epoch": 5.208949260886936, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 13038 + }, + { + "epoch": 5.2093487814622454, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13039 + }, + { + "epoch": 5.209748302037555, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13040 + }, + { + "epoch": 5.210147822612864, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13041 + }, + { + "epoch": 5.210547343188174, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13042 + }, + { + "epoch": 5.210946863763484, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13043 + }, + { + "epoch": 5.2113463843387935, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13044 + }, + { + "epoch": 5.211745904914103, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13045 + }, + { + "epoch": 5.212145425489413, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13046 + }, + { + "epoch": 5.212544946064722, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13047 + }, + { + "epoch": 5.212944466640032, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13048 + }, + { + "epoch": 5.2133439872153415, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3373, + "step": 13049 + }, + { + "epoch": 5.213743507790651, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13050 + }, + { + "epoch": 5.214143028365961, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13051 + }, + { + "epoch": 5.214542548941271, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13052 + }, + { + "epoch": 5.21494206951658, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13053 + }, + { + "epoch": 5.21534159009189, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13054 + }, + { + "epoch": 5.215741110667199, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13055 + }, + { + "epoch": 5.216140631242509, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 13056 + }, + { + "epoch": 5.216540151817819, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13057 + }, + { + "epoch": 5.216939672393128, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13058 + }, + { + "epoch": 5.217339192968438, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13059 + }, + { + "epoch": 5.217738713543747, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13060 + }, + { + "epoch": 5.218138234119057, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13061 + }, + { + "epoch": 5.218537754694367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 13062 + }, + { + "epoch": 5.218937275269677, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13063 + }, + { + "epoch": 5.219336795844986, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13064 + }, + { + "epoch": 5.2197363164202955, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13065 + }, + { + "epoch": 5.220135836995605, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13066 + }, + { + "epoch": 5.220535357570915, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13067 + }, + { + "epoch": 5.220934878146225, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13068 + }, + { + "epoch": 5.221334398721535, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13069 + }, + { + "epoch": 5.2217339192968435, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13070 + }, + { + "epoch": 5.222133439872153, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13071 + }, + { + "epoch": 5.222532960447463, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13072 + }, + { + "epoch": 5.222932481022773, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13073 + }, + { + "epoch": 5.223332001598083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13074 + }, + { + "epoch": 5.2237315221733915, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13075 + }, + { + "epoch": 5.224131042748701, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13076 + }, + { + "epoch": 5.224530563324011, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13077 + }, + { + "epoch": 5.224930083899321, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13078 + }, + { + "epoch": 5.225329604474631, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13079 + }, + { + "epoch": 5.2257291250499405, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13080 + }, + { + "epoch": 5.226128645625249, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13081 + }, + { + "epoch": 5.226528166200559, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13082 + }, + { + "epoch": 5.226927686775869, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13083 + }, + { + "epoch": 5.227327207351179, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13084 + }, + { + "epoch": 5.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13085 + }, + { + "epoch": 5.228126248501797, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13086 + }, + { + "epoch": 5.228525769077107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13087 + }, + { + "epoch": 5.228925289652417, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13088 + }, + { + "epoch": 5.229324810227727, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13089 + }, + { + "epoch": 5.229724330803037, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13090 + }, + { + "epoch": 5.230123851378346, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13091 + }, + { + "epoch": 5.230523371953655, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13092 + }, + { + "epoch": 5.230922892528965, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13093 + }, + { + "epoch": 5.231322413104275, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13094 + }, + { + "epoch": 5.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13095 + }, + { + "epoch": 5.232121454254894, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13096 + }, + { + "epoch": 5.232520974830204, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13097 + }, + { + "epoch": 5.232920495405513, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13098 + }, + { + "epoch": 5.233320015980823, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13099 + }, + { + "epoch": 5.233719536556133, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13100 + }, + { + "epoch": 5.234119057131442, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13101 + }, + { + "epoch": 5.234518577706752, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13102 + }, + { + "epoch": 5.234918098282061, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13103 + }, + { + "epoch": 5.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13104 + }, + { + "epoch": 5.235717139432681, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13105 + }, + { + "epoch": 5.2361166600079905, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13106 + }, + { + "epoch": 5.2365161805833, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13107 + }, + { + "epoch": 5.23691570115861, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13108 + }, + { + "epoch": 5.237315221733919, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13109 + }, + { + "epoch": 5.237714742309229, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13110 + }, + { + "epoch": 5.2381142628845385, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13111 + }, + { + "epoch": 5.238513783459848, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13112 + }, + { + "epoch": 5.238913304035158, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13113 + }, + { + "epoch": 5.239312824610467, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13114 + }, + { + "epoch": 5.239712345185777, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13115 + }, + { + "epoch": 5.240111865761087, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13116 + }, + { + "epoch": 5.240511386336396, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13117 + }, + { + "epoch": 5.240910906911706, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13118 + }, + { + "epoch": 5.241310427487016, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13119 + }, + { + "epoch": 5.241709948062325, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13120 + }, + { + "epoch": 5.242109468637635, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13121 + }, + { + "epoch": 5.242508989212944, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13122 + }, + { + "epoch": 5.242908509788254, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13123 + }, + { + "epoch": 5.243308030363564, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13124 + }, + { + "epoch": 5.243707550938874, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13125 + }, + { + "epoch": 5.244107071514183, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13126 + }, + { + "epoch": 5.244506592089492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13127 + }, + { + "epoch": 5.244906112664802, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13128 + }, + { + "epoch": 5.245305633240112, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 13129 + }, + { + "epoch": 5.245705153815422, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13130 + }, + { + "epoch": 5.246104674390731, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13131 + }, + { + "epoch": 5.2465041949660405, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13132 + }, + { + "epoch": 5.24690371554135, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13133 + }, + { + "epoch": 5.24730323611666, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13134 + }, + { + "epoch": 5.24770275669197, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13135 + }, + { + "epoch": 5.24810227726728, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13136 + }, + { + "epoch": 5.2485017978425885, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13137 + }, + { + "epoch": 5.248901318417898, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13138 + }, + { + "epoch": 5.249300838993208, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13139 + }, + { + "epoch": 5.249700359568518, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13140 + }, + { + "epoch": 5.250099880143828, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13141 + }, + { + "epoch": 5.2504994007191375, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13142 + }, + { + "epoch": 5.250898921294446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13143 + }, + { + "epoch": 5.251298441869756, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13144 + }, + { + "epoch": 5.251697962445066, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13145 + }, + { + "epoch": 5.252097483020376, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13146 + }, + { + "epoch": 5.2524970035956855, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13147 + }, + { + "epoch": 5.252896524170994, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13148 + }, + { + "epoch": 5.253296044746304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13149 + }, + { + "epoch": 5.253695565321614, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13150 + }, + { + "epoch": 5.254095085896924, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13151 + }, + { + "epoch": 5.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13152 + }, + { + "epoch": 5.254894127047543, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13153 + }, + { + "epoch": 5.255293647622852, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13154 + }, + { + "epoch": 5.255693168198162, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13155 + }, + { + "epoch": 5.256092688773472, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13156 + }, + { + "epoch": 5.256492209348782, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13157 + }, + { + "epoch": 5.256891729924091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13158 + }, + { + "epoch": 5.257291250499401, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13159 + }, + { + "epoch": 5.25769077107471, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13160 + }, + { + "epoch": 5.25809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13161 + }, + { + "epoch": 5.25848981222533, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13162 + }, + { + "epoch": 5.258889332800639, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13163 + }, + { + "epoch": 5.259288853375949, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13164 + }, + { + "epoch": 5.259688373951258, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13165 + }, + { + "epoch": 5.260087894526568, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13166 + }, + { + "epoch": 5.260487415101878, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13167 + }, + { + "epoch": 5.2608869356771875, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13168 + }, + { + "epoch": 5.261286456252497, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13169 + }, + { + "epoch": 5.261685976827807, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13170 + }, + { + "epoch": 5.262085497403116, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13171 + }, + { + "epoch": 5.262485017978426, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13172 + }, + { + "epoch": 5.2628845385537355, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13173 + }, + { + "epoch": 5.263284059129045, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13174 + }, + { + "epoch": 5.263683579704355, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13175 + }, + { + "epoch": 5.264083100279665, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13176 + }, + { + "epoch": 5.264482620854974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13177 + }, + { + "epoch": 5.2648821414302835, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13178 + }, + { + "epoch": 5.265281662005593, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13179 + }, + { + "epoch": 5.265681182580903, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13180 + }, + { + "epoch": 5.266080703156213, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13181 + }, + { + "epoch": 5.266480223731522, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13182 + }, + { + "epoch": 5.266879744306832, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13183 + }, + { + "epoch": 5.267279264882141, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13184 + }, + { + "epoch": 5.267678785457451, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13185 + }, + { + "epoch": 5.268078306032761, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13186 + }, + { + "epoch": 5.26847782660807, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13187 + }, + { + "epoch": 5.26887734718338, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13188 + }, + { + "epoch": 5.269276867758689, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13189 + }, + { + "epoch": 5.269676388333999, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13190 + }, + { + "epoch": 5.270075908909309, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13191 + }, + { + "epoch": 5.270475429484619, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13192 + }, + { + "epoch": 5.270874950059928, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13193 + }, + { + "epoch": 5.2712744706352375, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13194 + }, + { + "epoch": 5.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13195 + }, + { + "epoch": 5.272073511785857, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13196 + }, + { + "epoch": 5.272473032361167, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13197 + }, + { + "epoch": 5.272872552936477, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13198 + }, + { + "epoch": 5.2732720735117855, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13199 + }, + { + "epoch": 5.273671594087095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13200 + }, + { + "epoch": 5.274071114662405, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13201 + }, + { + "epoch": 5.274470635237715, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13202 + }, + { + "epoch": 5.274870155813025, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13203 + }, + { + "epoch": 5.2752696763883336, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 13204 + }, + { + "epoch": 5.275669196963643, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13205 + }, + { + "epoch": 5.276068717538953, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13206 + }, + { + "epoch": 5.276468238114263, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13207 + }, + { + "epoch": 5.276867758689573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13208 + }, + { + "epoch": 5.2772672792648825, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13209 + }, + { + "epoch": 5.277666799840191, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13210 + }, + { + "epoch": 5.278066320415501, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13211 + }, + { + "epoch": 5.278465840990811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 13212 + }, + { + "epoch": 5.278865361566121, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13213 + }, + { + "epoch": 5.2792648821414305, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13214 + }, + { + "epoch": 5.27966440271674, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13215 + }, + { + "epoch": 5.280063923292049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13216 + }, + { + "epoch": 5.280463443867359, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13217 + }, + { + "epoch": 5.280862964442669, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13218 + }, + { + "epoch": 5.281262485017979, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13219 + }, + { + "epoch": 5.281662005593288, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13220 + }, + { + "epoch": 5.282061526168597, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13221 + }, + { + "epoch": 5.282461046743907, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13222 + }, + { + "epoch": 5.282860567319217, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13223 + }, + { + "epoch": 5.283260087894527, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13224 + }, + { + "epoch": 5.283659608469836, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13225 + }, + { + "epoch": 5.284059129045146, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13226 + }, + { + "epoch": 5.284458649620455, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13227 + }, + { + "epoch": 5.284858170195765, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 13228 + }, + { + "epoch": 5.285257690771075, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 13229 + }, + { + "epoch": 5.285657211346384, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13230 + }, + { + "epoch": 5.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13231 + }, + { + "epoch": 5.286456252497004, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13232 + }, + { + "epoch": 5.286855773072313, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 13233 + }, + { + "epoch": 5.287255293647623, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13234 + }, + { + "epoch": 5.2876548142229325, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13235 + }, + { + "epoch": 5.288054334798242, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13236 + }, + { + "epoch": 5.288453855373552, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13237 + }, + { + "epoch": 5.288853375948861, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13238 + }, + { + "epoch": 5.289252896524171, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13239 + }, + { + "epoch": 5.2896524170994805, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13240 + }, + { + "epoch": 5.29005193767479, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13241 + }, + { + "epoch": 5.2904514582501, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13242 + }, + { + "epoch": 5.29085097882541, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13243 + }, + { + "epoch": 5.291250499400719, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 13244 + }, + { + "epoch": 5.291650019976029, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 13245 + }, + { + "epoch": 5.292049540551338, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13246 + }, + { + "epoch": 5.292449061126648, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13247 + }, + { + "epoch": 5.292848581701958, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13248 + }, + { + "epoch": 5.293248102277268, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13249 + }, + { + "epoch": 5.293647622852577, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13250 + }, + { + "epoch": 5.294047143427886, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13251 + }, + { + "epoch": 5.294446664003196, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13252 + }, + { + "epoch": 5.294846184578506, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13253 + }, + { + "epoch": 5.295245705153816, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13254 + }, + { + "epoch": 5.295645225729125, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13255 + }, + { + "epoch": 5.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13256 + }, + { + "epoch": 5.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.2965, + "step": 13257 + }, + { + "epoch": 5.296843787455054, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13258 + }, + { + "epoch": 5.297243308030364, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13259 + }, + { + "epoch": 5.297642828605674, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13260 + }, + { + "epoch": 5.2980423491809825, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13261 + }, + { + "epoch": 5.298441869756292, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13262 + }, + { + "epoch": 5.298841390331602, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13263 + }, + { + "epoch": 5.299240910906912, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13264 + }, + { + "epoch": 5.299640431482222, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13265 + }, + { + "epoch": 5.300039952057531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13266 + }, + { + "epoch": 5.30043947263284, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13267 + }, + { + "epoch": 5.30083899320815, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 13268 + }, + { + "epoch": 5.30123851378346, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13269 + }, + { + "epoch": 5.30163803435877, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13270 + }, + { + "epoch": 5.3020375549340795, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13271 + }, + { + "epoch": 5.302437075509388, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13272 + }, + { + "epoch": 5.302836596084698, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 13273 + }, + { + "epoch": 5.303236116660008, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13274 + }, + { + "epoch": 5.303635637235318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13275 + }, + { + "epoch": 5.3040351578106275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13276 + }, + { + "epoch": 5.304434678385936, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13277 + }, + { + "epoch": 5.304834198961246, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13278 + }, + { + "epoch": 5.305233719536556, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13279 + }, + { + "epoch": 5.305633240111866, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13280 + }, + { + "epoch": 5.3060327606871756, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13281 + }, + { + "epoch": 5.306432281262485, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13282 + }, + { + "epoch": 5.306831801837794, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13283 + }, + { + "epoch": 5.307231322413104, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13284 + }, + { + "epoch": 5.307630842988414, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13285 + }, + { + "epoch": 5.308030363563724, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13286 + }, + { + "epoch": 5.308429884139033, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13287 + }, + { + "epoch": 5.308829404714343, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13288 + }, + { + "epoch": 5.309228925289652, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13289 + }, + { + "epoch": 5.309628445864962, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13290 + }, + { + "epoch": 5.310027966440272, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13291 + }, + { + "epoch": 5.310427487015581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 13292 + }, + { + "epoch": 5.310827007590891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13293 + }, + { + "epoch": 5.3112265281662, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13294 + }, + { + "epoch": 5.31162604874151, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13295 + }, + { + "epoch": 5.31202556931682, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13296 + }, + { + "epoch": 5.3124250898921295, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13297 + }, + { + "epoch": 5.312824610467439, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13298 + }, + { + "epoch": 5.313224131042749, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13299 + }, + { + "epoch": 5.313623651618058, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13300 + }, + { + "epoch": 5.314023172193368, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13301 + }, + { + "epoch": 5.3144226927686775, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13302 + }, + { + "epoch": 5.314822213343987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13303 + }, + { + "epoch": 5.315221733919297, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13304 + }, + { + "epoch": 5.315621254494607, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13305 + }, + { + "epoch": 5.316020775069916, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13306 + }, + { + "epoch": 5.316420295645226, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 13307 + }, + { + "epoch": 5.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13308 + }, + { + "epoch": 5.317219336795845, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13309 + }, + { + "epoch": 5.317618857371155, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13310 + }, + { + "epoch": 5.318018377946464, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13311 + }, + { + "epoch": 5.318417898521774, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13312 + }, + { + "epoch": 5.318817419097083, + "grad_norm": 0.140625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 13313 + }, + { + "epoch": 5.319216939672393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 13314 + }, + { + "epoch": 5.319616460247703, + "grad_norm": 0.146484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13315 + }, + { + "epoch": 5.320015980823013, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 13316 + }, + { + "epoch": 5.320415501398322, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13317 + }, + { + "epoch": 5.320815021973631, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 13318 + }, + { + "epoch": 5.321214542548941, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13319 + }, + { + "epoch": 5.321614063124251, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13320 + }, + { + "epoch": 5.322013583699561, + "grad_norm": 0.1689453125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13321 + }, + { + "epoch": 5.322413104274871, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13322 + }, + { + "epoch": 5.3228126248501795, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13323 + }, + { + "epoch": 5.323212145425489, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13324 + }, + { + "epoch": 5.323611666000799, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13325 + }, + { + "epoch": 5.324011186576109, + "grad_norm": 0.142578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13326 + }, + { + "epoch": 5.324410707151419, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13327 + }, + { + "epoch": 5.3248102277267275, + "grad_norm": 0.1474609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13328 + }, + { + "epoch": 5.325209748302037, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13329 + }, + { + "epoch": 5.325609268877347, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13330 + }, + { + "epoch": 5.326008789452657, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13331 + }, + { + "epoch": 5.326408310027967, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13332 + }, + { + "epoch": 5.3268078306032765, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13333 + }, + { + "epoch": 5.327207351178585, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13334 + }, + { + "epoch": 5.327606871753895, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13335 + }, + { + "epoch": 5.328006392329205, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3381, + "step": 13336 + }, + { + "epoch": 5.328405912904515, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13337 + }, + { + "epoch": 5.3288054334798245, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13338 + }, + { + "epoch": 5.329204954055134, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13339 + }, + { + "epoch": 5.329604474630443, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13340 + }, + { + "epoch": 5.330003995205753, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13341 + }, + { + "epoch": 5.330403515781063, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13342 + }, + { + "epoch": 5.3308030363563725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13343 + }, + { + "epoch": 5.331202556931682, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13344 + }, + { + "epoch": 5.331602077506991, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 13345 + }, + { + "epoch": 5.332001598082301, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13346 + }, + { + "epoch": 5.332401118657611, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13347 + }, + { + "epoch": 5.332800639232921, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13348 + }, + { + "epoch": 5.33320015980823, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13349 + }, + { + "epoch": 5.33359968038354, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13350 + }, + { + "epoch": 5.333999200958849, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13351 + }, + { + "epoch": 5.334398721534159, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13352 + }, + { + "epoch": 5.334798242109469, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13353 + }, + { + "epoch": 5.335197762684778, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 13354 + }, + { + "epoch": 5.335597283260088, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13355 + }, + { + "epoch": 5.335996803835398, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13356 + }, + { + "epoch": 5.336396324410707, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13357 + }, + { + "epoch": 5.336795844986017, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13358 + }, + { + "epoch": 5.3371953655613265, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13359 + }, + { + "epoch": 5.337594886136636, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13360 + }, + { + "epoch": 5.337994406711946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13361 + }, + { + "epoch": 5.338393927287255, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13362 + }, + { + "epoch": 5.338793447862565, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13363 + }, + { + "epoch": 5.3391929684378745, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13364 + }, + { + "epoch": 5.339592489013184, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13365 + }, + { + "epoch": 5.339992009588494, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13366 + }, + { + "epoch": 5.340391530163803, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13367 + }, + { + "epoch": 5.340791050739113, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 13368 + }, + { + "epoch": 5.3411905713144225, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13369 + }, + { + "epoch": 5.341590091889732, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13370 + }, + { + "epoch": 5.341989612465042, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13371 + }, + { + "epoch": 5.342389133040352, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13372 + }, + { + "epoch": 5.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13373 + }, + { + "epoch": 5.343188174190971, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 13374 + }, + { + "epoch": 5.34358769476628, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 13375 + }, + { + "epoch": 5.34398721534159, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13376 + }, + { + "epoch": 5.3443867359169, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13377 + }, + { + "epoch": 5.34478625649221, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13378 + }, + { + "epoch": 5.345185777067519, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13379 + }, + { + "epoch": 5.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13380 + }, + { + "epoch": 5.345984818218138, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13381 + }, + { + "epoch": 5.346384338793448, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 13382 + }, + { + "epoch": 5.346783859368758, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13383 + }, + { + "epoch": 5.347183379944067, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13384 + }, + { + "epoch": 5.3475829005193765, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13385 + }, + { + "epoch": 5.347982421094686, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13386 + }, + { + "epoch": 5.348381941669996, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13387 + }, + { + "epoch": 5.348781462245306, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13388 + }, + { + "epoch": 5.349180982820616, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13389 + }, + { + "epoch": 5.3495805033959245, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13390 + }, + { + "epoch": 5.349980023971234, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13391 + }, + { + "epoch": 5.350379544546544, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13392 + }, + { + "epoch": 5.350779065121854, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13393 + }, + { + "epoch": 5.351178585697164, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13394 + }, + { + "epoch": 5.351578106272473, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 13395 + }, + { + "epoch": 5.351977626847782, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13396 + }, + { + "epoch": 5.352377147423092, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 13397 + }, + { + "epoch": 5.352776667998402, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13398 + }, + { + "epoch": 5.353176188573712, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13399 + }, + { + "epoch": 5.3535757091490215, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13400 + }, + { + "epoch": 5.35397522972433, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13401 + }, + { + "epoch": 5.35437475029964, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13402 + }, + { + "epoch": 5.35477427087495, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13403 + }, + { + "epoch": 5.35517379145026, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13404 + }, + { + "epoch": 5.3555733120255695, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13405 + }, + { + "epoch": 5.355972832600879, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13406 + }, + { + "epoch": 5.356372353176188, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13407 + }, + { + "epoch": 5.356771873751498, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 13408 + }, + { + "epoch": 5.357171394326808, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13409 + }, + { + "epoch": 5.357570914902118, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13410 + }, + { + "epoch": 5.357970435477427, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13411 + }, + { + "epoch": 5.358369956052737, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13412 + }, + { + "epoch": 5.358769476628046, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13413 + }, + { + "epoch": 5.359168997203356, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13414 + }, + { + "epoch": 5.359568517778666, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13415 + }, + { + "epoch": 5.359968038353975, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 13416 + }, + { + "epoch": 5.360367558929285, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13417 + }, + { + "epoch": 5.360767079504594, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13418 + }, + { + "epoch": 5.361166600079904, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13419 + }, + { + "epoch": 5.361566120655214, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13420 + }, + { + "epoch": 5.361965641230523, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13421 + }, + { + "epoch": 5.362365161805833, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13422 + }, + { + "epoch": 5.362764682381143, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13423 + }, + { + "epoch": 5.363164202956452, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13424 + }, + { + "epoch": 5.363563723531762, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 13425 + }, + { + "epoch": 5.3639632441070715, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13426 + }, + { + "epoch": 5.364362764682381, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13427 + }, + { + "epoch": 5.364762285257691, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13428 + }, + { + "epoch": 5.365161805833001, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13429 + }, + { + "epoch": 5.36556132640831, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 13430 + }, + { + "epoch": 5.3659608469836195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13431 + }, + { + "epoch": 5.366360367558929, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13432 + }, + { + "epoch": 5.366759888134239, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13433 + }, + { + "epoch": 5.367159408709549, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13434 + }, + { + "epoch": 5.367558929284858, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 13435 + }, + { + "epoch": 5.367958449860168, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13436 + }, + { + "epoch": 5.368357970435477, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 13437 + }, + { + "epoch": 5.368757491010787, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13438 + }, + { + "epoch": 5.369157011586097, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13439 + }, + { + "epoch": 5.369556532161407, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 13440 + }, + { + "epoch": 5.369956052736716, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13441 + }, + { + "epoch": 5.370355573312025, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 13442 + }, + { + "epoch": 5.370755093887335, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 13443 + }, + { + "epoch": 5.371154614462645, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13444 + }, + { + "epoch": 5.371554135037955, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13445 + }, + { + "epoch": 5.3719536556132645, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13446 + }, + { + "epoch": 5.372353176188573, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13447 + }, + { + "epoch": 5.372752696763883, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13448 + }, + { + "epoch": 5.373152217339193, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13449 + }, + { + "epoch": 5.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13450 + }, + { + "epoch": 5.373951258489813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13451 + }, + { + "epoch": 5.3743507790651215, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13452 + }, + { + "epoch": 5.374750299640431, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13453 + }, + { + "epoch": 5.375149820215741, + "grad_norm": 0.109375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 13454 + }, + { + "epoch": 5.375549340791051, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13455 + }, + { + "epoch": 5.375948861366361, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13456 + }, + { + "epoch": 5.3763483819416695, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13457 + }, + { + "epoch": 5.376747902516979, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13458 + }, + { + "epoch": 5.377147423092289, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13459 + }, + { + "epoch": 5.377546943667599, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13460 + }, + { + "epoch": 5.377946464242909, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13461 + }, + { + "epoch": 5.3783459848182185, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13462 + }, + { + "epoch": 5.378745505393527, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13463 + }, + { + "epoch": 5.379145025968837, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13464 + }, + { + "epoch": 5.379544546544147, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13465 + }, + { + "epoch": 5.379944067119457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13466 + }, + { + "epoch": 5.3803435876947665, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 13467 + }, + { + "epoch": 5.380743108270076, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13468 + }, + { + "epoch": 5.381142628845385, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13469 + }, + { + "epoch": 5.381542149420695, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13470 + }, + { + "epoch": 5.381941669996005, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13471 + }, + { + "epoch": 5.3823411905713145, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13472 + }, + { + "epoch": 5.382740711146624, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13473 + }, + { + "epoch": 5.383140231721933, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13474 + }, + { + "epoch": 5.383539752297243, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13475 + }, + { + "epoch": 5.383939272872553, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13476 + }, + { + "epoch": 5.384338793447863, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13477 + }, + { + "epoch": 5.384738314023172, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 13478 + }, + { + "epoch": 5.385137834598482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13479 + }, + { + "epoch": 5.385537355173791, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 13480 + }, + { + "epoch": 5.385936875749101, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13481 + }, + { + "epoch": 5.386336396324411, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13482 + }, + { + "epoch": 5.38673591689972, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13483 + }, + { + "epoch": 5.38713543747503, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13484 + }, + { + "epoch": 5.38753495805034, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13485 + }, + { + "epoch": 5.387934478625649, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 13486 + }, + { + "epoch": 5.388333999200959, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 13487 + }, + { + "epoch": 5.3887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13488 + }, + { + "epoch": 5.389133040351578, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 13489 + }, + { + "epoch": 5.389532560926888, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13490 + }, + { + "epoch": 5.389932081502197, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13491 + }, + { + "epoch": 5.390331602077507, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13492 + }, + { + "epoch": 5.3907311226528165, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13493 + }, + { + "epoch": 5.391130643228126, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13494 + }, + { + "epoch": 5.391530163803436, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13495 + }, + { + "epoch": 5.391929684378746, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13496 + }, + { + "epoch": 5.392329204954055, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13497 + }, + { + "epoch": 5.3927287255293646, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 13498 + }, + { + "epoch": 5.393128246104674, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13499 + }, + { + "epoch": 5.393527766679984, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13500 + }, + { + "epoch": 5.393927287255294, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13501 + }, + { + "epoch": 5.394326807830604, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13502 + }, + { + "epoch": 5.394726328405913, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13503 + }, + { + "epoch": 5.395125848981222, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 13504 + }, + { + "epoch": 5.395525369556532, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13505 + }, + { + "epoch": 5.395924890131842, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13506 + }, + { + "epoch": 5.396324410707152, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13507 + }, + { + "epoch": 5.396723931282461, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 13508 + }, + { + "epoch": 5.39712345185777, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13509 + }, + { + "epoch": 5.39752297243308, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13510 + }, + { + "epoch": 5.39792249300839, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13511 + }, + { + "epoch": 5.3983220135837, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13512 + }, + { + "epoch": 5.39872153415901, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13513 + }, + { + "epoch": 5.3991210547343185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13514 + }, + { + "epoch": 5.399520575309628, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13515 + }, + { + "epoch": 5.399920095884938, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13516 + }, + { + "epoch": 5.400319616460248, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13517 + }, + { + "epoch": 5.400719137035558, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13518 + }, + { + "epoch": 5.401118657610867, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13519 + }, + { + "epoch": 5.401518178186176, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13520 + }, + { + "epoch": 5.401917698761486, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13521 + }, + { + "epoch": 5.402317219336796, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13522 + }, + { + "epoch": 5.402716739912106, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 13523 + }, + { + "epoch": 5.4031162604874154, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13524 + }, + { + "epoch": 5.403515781062724, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13525 + }, + { + "epoch": 5.403915301638034, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13526 + }, + { + "epoch": 5.404314822213344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13527 + }, + { + "epoch": 5.404714342788654, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13528 + }, + { + "epoch": 5.4051138633639635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13529 + }, + { + "epoch": 5.405513383939273, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.2988, + "step": 13530 + }, + { + "epoch": 5.405912904514582, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13531 + }, + { + "epoch": 5.406312425089892, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13532 + }, + { + "epoch": 5.406711945665202, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13533 + }, + { + "epoch": 5.4071114662405115, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13534 + }, + { + "epoch": 5.407510986815821, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13535 + }, + { + "epoch": 5.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13536 + }, + { + "epoch": 5.40831002796644, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 13537 + }, + { + "epoch": 5.40870954854175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13538 + }, + { + "epoch": 5.40910906911706, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13539 + }, + { + "epoch": 5.409508589692369, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 13540 + }, + { + "epoch": 5.409908110267679, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13541 + }, + { + "epoch": 5.410307630842988, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13542 + }, + { + "epoch": 5.410707151418298, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13543 + }, + { + "epoch": 5.411106671993608, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13544 + }, + { + "epoch": 5.411506192568917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13545 + }, + { + "epoch": 5.411905713144227, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13546 + }, + { + "epoch": 5.412305233719536, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13547 + }, + { + "epoch": 5.412704754294846, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13548 + }, + { + "epoch": 5.413104274870156, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 13549 + }, + { + "epoch": 5.4135037954454654, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13550 + }, + { + "epoch": 5.413903316020775, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13551 + }, + { + "epoch": 5.414302836596085, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13552 + }, + { + "epoch": 5.414702357171394, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 13553 + }, + { + "epoch": 5.415101877746704, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13554 + }, + { + "epoch": 5.4155013983220135, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13555 + }, + { + "epoch": 5.415900918897323, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 13556 + }, + { + "epoch": 5.416300439472633, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13557 + }, + { + "epoch": 5.416699960047943, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13558 + }, + { + "epoch": 5.417099480623252, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13559 + }, + { + "epoch": 5.4174990011985615, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 13560 + }, + { + "epoch": 5.417898521773871, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13561 + }, + { + "epoch": 5.418298042349181, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 13562 + }, + { + "epoch": 5.418697562924491, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13563 + }, + { + "epoch": 5.4190970834998, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13564 + }, + { + "epoch": 5.41949660407511, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13565 + }, + { + "epoch": 5.419896124650419, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13566 + }, + { + "epoch": 5.420295645225729, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13567 + }, + { + "epoch": 5.420695165801039, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13568 + }, + { + "epoch": 5.421094686376349, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13569 + }, + { + "epoch": 5.421494206951658, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13570 + }, + { + "epoch": 5.421893727526967, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 13571 + }, + { + "epoch": 5.422293248102277, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13572 + }, + { + "epoch": 5.422692768677587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13573 + }, + { + "epoch": 5.423092289252897, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13574 + }, + { + "epoch": 5.4234918098282066, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13575 + }, + { + "epoch": 5.4238913304035155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13576 + }, + { + "epoch": 5.424290850978825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13577 + }, + { + "epoch": 5.424690371554135, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13578 + }, + { + "epoch": 5.425089892129445, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13579 + }, + { + "epoch": 5.425489412704755, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13580 + }, + { + "epoch": 5.4258889332800635, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13581 + }, + { + "epoch": 5.426288453855373, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13582 + }, + { + "epoch": 5.426687974430683, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13583 + }, + { + "epoch": 5.427087495005993, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13584 + }, + { + "epoch": 5.427487015581303, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13585 + }, + { + "epoch": 5.427886536156612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 13586 + }, + { + "epoch": 5.428286056731921, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13587 + }, + { + "epoch": 5.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 13588 + }, + { + "epoch": 5.429085097882541, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 13589 + }, + { + "epoch": 5.429484618457851, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13590 + }, + { + "epoch": 5.4298841390331605, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13591 + }, + { + "epoch": 5.43028365960847, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13592 + }, + { + "epoch": 5.430683180183779, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13593 + }, + { + "epoch": 5.431082700759089, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13594 + }, + { + "epoch": 5.431482221334399, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13595 + }, + { + "epoch": 5.4318817419097085, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13596 + }, + { + "epoch": 5.432281262485018, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13597 + }, + { + "epoch": 5.432680783060327, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13598 + }, + { + "epoch": 5.433080303635637, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13599 + }, + { + "epoch": 5.433479824210947, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13600 + }, + { + "epoch": 5.433879344786257, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13601 + }, + { + "epoch": 5.434278865361566, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13602 + }, + { + "epoch": 5.434678385936876, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 13603 + }, + { + "epoch": 5.435077906512185, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13604 + }, + { + "epoch": 5.435477427087495, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13605 + }, + { + "epoch": 5.435876947662805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13606 + }, + { + "epoch": 5.436276468238114, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13607 + }, + { + "epoch": 5.436675988813424, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 13608 + }, + { + "epoch": 5.437075509388734, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13609 + }, + { + "epoch": 5.437475029964043, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 13610 + }, + { + "epoch": 5.437874550539353, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13611 + }, + { + "epoch": 5.438274071114662, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13612 + }, + { + "epoch": 5.438673591689972, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13613 + }, + { + "epoch": 5.439073112265282, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 13614 + }, + { + "epoch": 5.439472632840591, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13615 + }, + { + "epoch": 5.439872153415901, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13616 + }, + { + "epoch": 5.4402716739912105, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3344, + "step": 13617 + }, + { + "epoch": 5.44067119456652, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13618 + }, + { + "epoch": 5.44107071514183, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 13619 + }, + { + "epoch": 5.44147023571714, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13620 + }, + { + "epoch": 5.441869756292449, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13621 + }, + { + "epoch": 5.4422692768677585, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13622 + }, + { + "epoch": 5.442668797443068, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13623 + }, + { + "epoch": 5.443068318018378, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13624 + }, + { + "epoch": 5.443467838593688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13625 + }, + { + "epoch": 5.443867359168998, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13626 + }, + { + "epoch": 5.444266879744307, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13627 + }, + { + "epoch": 5.444666400319616, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13628 + }, + { + "epoch": 5.445065920894926, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13629 + }, + { + "epoch": 5.445465441470236, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13630 + }, + { + "epoch": 5.445864962045546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13631 + }, + { + "epoch": 5.446264482620855, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 13632 + }, + { + "epoch": 5.446664003196164, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13633 + }, + { + "epoch": 5.447063523771474, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13634 + }, + { + "epoch": 5.447463044346784, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13635 + }, + { + "epoch": 5.447862564922094, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13636 + }, + { + "epoch": 5.448262085497403, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13637 + }, + { + "epoch": 5.448661606072712, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13638 + }, + { + "epoch": 5.449061126648022, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13639 + }, + { + "epoch": 5.449460647223332, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 13640 + }, + { + "epoch": 5.449860167798642, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13641 + }, + { + "epoch": 5.450259688373952, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13642 + }, + { + "epoch": 5.4506592089492605, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 13643 + }, + { + "epoch": 5.45105872952457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 13644 + }, + { + "epoch": 5.45145825009988, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 13645 + }, + { + "epoch": 5.45185777067519, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13646 + }, + { + "epoch": 5.4522572912505, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13647 + }, + { + "epoch": 5.452656811825809, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13648 + }, + { + "epoch": 5.453056332401118, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 13649 + }, + { + "epoch": 5.453455852976428, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13650 + }, + { + "epoch": 5.453855373551738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 13651 + }, + { + "epoch": 5.454254894127048, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.333, + "step": 13652 + }, + { + "epoch": 5.4546544147023575, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3313, + "step": 13653 + }, + { + "epoch": 5.455053935277666, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13654 + }, + { + "epoch": 5.455453455852976, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 13655 + }, + { + "epoch": 5.455852976428286, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 13656 + }, + { + "epoch": 5.456252497003596, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13657 + }, + { + "epoch": 5.4566520175789055, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13658 + }, + { + "epoch": 5.457051538154215, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13659 + }, + { + "epoch": 5.457451058729524, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13660 + }, + { + "epoch": 5.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13661 + }, + { + "epoch": 5.458250099880144, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 13662 + }, + { + "epoch": 5.4586496204554535, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 13663 + }, + { + "epoch": 5.459049141030763, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 13664 + }, + { + "epoch": 5.459448661606073, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13665 + }, + { + "epoch": 5.459848182181382, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13666 + }, + { + "epoch": 5.460247702756692, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13667 + }, + { + "epoch": 5.460647223332002, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 13668 + }, + { + "epoch": 5.461046743907311, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 13669 + }, + { + "epoch": 5.461446264482621, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13670 + }, + { + "epoch": 5.46184578505793, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13671 + }, + { + "epoch": 5.46224530563324, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13672 + }, + { + "epoch": 5.46264482620855, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13673 + }, + { + "epoch": 5.463044346783859, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13674 + }, + { + "epoch": 5.463443867359169, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13675 + }, + { + "epoch": 5.463843387934479, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 13676 + }, + { + "epoch": 5.464242908509788, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13677 + }, + { + "epoch": 5.464642429085098, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13678 + }, + { + "epoch": 5.4650419496604075, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13679 + }, + { + "epoch": 5.465441470235717, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 13680 + }, + { + "epoch": 5.465840990811027, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13681 + }, + { + "epoch": 5.466240511386337, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13682 + }, + { + "epoch": 5.466640031961646, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 13683 + }, + { + "epoch": 5.4670395525369555, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13684 + }, + { + "epoch": 5.467439073112265, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13685 + }, + { + "epoch": 5.467838593687575, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13686 + }, + { + "epoch": 5.468238114262885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13687 + }, + { + "epoch": 5.468637634838194, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13688 + }, + { + "epoch": 5.4690371554135035, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13689 + }, + { + "epoch": 5.469436675988813, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 13690 + }, + { + "epoch": 5.469836196564123, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13691 + }, + { + "epoch": 5.470235717139433, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13692 + }, + { + "epoch": 5.470635237714743, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 13693 + }, + { + "epoch": 5.471034758290052, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13694 + }, + { + "epoch": 5.471434278865361, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 13695 + }, + { + "epoch": 5.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13696 + }, + { + "epoch": 5.472233320015981, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13697 + }, + { + "epoch": 5.472632840591291, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 13698 + }, + { + "epoch": 5.4730323611666005, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13699 + }, + { + "epoch": 5.473431881741909, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13700 + }, + { + "epoch": 5.473831402317219, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13701 + }, + { + "epoch": 5.474230922892529, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13702 + }, + { + "epoch": 5.474630443467839, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 13703 + }, + { + "epoch": 5.475029964043149, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 13704 + }, + { + "epoch": 5.4754294846184575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13705 + }, + { + "epoch": 5.475829005193767, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13706 + }, + { + "epoch": 5.476228525769077, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13707 + }, + { + "epoch": 5.476628046344387, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13708 + }, + { + "epoch": 5.477027566919697, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13709 + }, + { + "epoch": 5.477427087495006, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 13710 + }, + { + "epoch": 5.477826608070315, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13711 + }, + { + "epoch": 5.478226128645625, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13712 + }, + { + "epoch": 5.478625649220935, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 13713 + }, + { + "epoch": 5.479025169796245, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3374, + "step": 13714 + }, + { + "epoch": 5.479424690371554, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13715 + }, + { + "epoch": 5.479824210946864, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13716 + }, + { + "epoch": 5.480223731522173, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 13717 + }, + { + "epoch": 5.480623252097483, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13718 + }, + { + "epoch": 5.481022772672793, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13719 + }, + { + "epoch": 5.4814222932481025, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13720 + }, + { + "epoch": 5.481821813823412, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13721 + }, + { + "epoch": 5.482221334398721, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13722 + }, + { + "epoch": 5.482620854974031, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13723 + }, + { + "epoch": 5.483020375549341, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 13724 + }, + { + "epoch": 5.4834198961246505, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13725 + }, + { + "epoch": 5.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 13726 + }, + { + "epoch": 5.484218937275269, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13727 + }, + { + "epoch": 5.484618457850579, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13728 + }, + { + "epoch": 5.485017978425889, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 13729 + }, + { + "epoch": 5.485417499001199, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13730 + }, + { + "epoch": 5.485817019576508, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 13731 + }, + { + "epoch": 5.486216540151818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13732 + }, + { + "epoch": 5.486616060727127, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 13733 + }, + { + "epoch": 5.487015581302437, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13734 + }, + { + "epoch": 5.487415101877747, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 13735 + }, + { + "epoch": 5.487814622453056, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13736 + }, + { + "epoch": 5.488214143028366, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13737 + }, + { + "epoch": 5.488613663603676, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 13738 + }, + { + "epoch": 5.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 13739 + }, + { + "epoch": 5.489412704754295, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 13740 + }, + { + "epoch": 5.4898122253296044, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13741 + }, + { + "epoch": 5.490211745904914, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13742 + }, + { + "epoch": 5.490611266480224, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13743 + }, + { + "epoch": 5.491010787055533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 13744 + }, + { + "epoch": 5.491410307630843, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 13745 + }, + { + "epoch": 5.4918098282061525, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 13746 + }, + { + "epoch": 5.492209348781462, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13747 + }, + { + "epoch": 5.492608869356772, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13748 + }, + { + "epoch": 5.493008389932082, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13749 + }, + { + "epoch": 5.493407910507391, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13750 + }, + { + "epoch": 5.4938074310827005, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13751 + }, + { + "epoch": 5.49420695165801, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13752 + }, + { + "epoch": 5.49460647223332, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13753 + }, + { + "epoch": 5.49500599280863, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13754 + }, + { + "epoch": 5.49540551338394, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13755 + }, + { + "epoch": 5.495805033959249, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13756 + }, + { + "epoch": 5.496204554534558, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 13757 + }, + { + "epoch": 5.496604075109868, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13758 + }, + { + "epoch": 5.497003595685178, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13759 + }, + { + "epoch": 5.497403116260488, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13760 + }, + { + "epoch": 5.497802636835797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3335, + "step": 13761 + }, + { + "epoch": 5.498202157411106, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13762 + }, + { + "epoch": 5.498601677986416, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 13763 + }, + { + "epoch": 5.499001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13764 + }, + { + "epoch": 5.499400719137036, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 13765 + }, + { + "epoch": 5.4998002397123456, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13766 + }, + { + "epoch": 5.5001997602876544, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13767 + }, + { + "epoch": 5.500599280862964, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 13768 + }, + { + "epoch": 5.500998801438274, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 13769 + }, + { + "epoch": 5.501398322013584, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 13770 + }, + { + "epoch": 5.501797842588894, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 13771 + }, + { + "epoch": 5.502197363164203, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 13772 + }, + { + "epoch": 5.502596883739512, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13773 + }, + { + "epoch": 5.502996404314822, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 13774 + }, + { + "epoch": 5.503395924890132, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13775 + }, + { + "epoch": 5.503795445465442, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 13776 + }, + { + "epoch": 5.504194966040751, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 13777 + }, + { + "epoch": 5.50459448661606, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13778 + }, + { + "epoch": 5.50499400719137, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13779 + }, + { + "epoch": 5.50539352776668, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13780 + }, + { + "epoch": 5.50579304834199, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13781 + }, + { + "epoch": 5.5061925689172995, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13782 + }, + { + "epoch": 5.506592089492609, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 13783 + }, + { + "epoch": 5.506991610067918, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13784 + }, + { + "epoch": 5.507391130643228, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13785 + }, + { + "epoch": 5.507790651218538, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13786 + }, + { + "epoch": 5.5081901717938475, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13787 + }, + { + "epoch": 5.508589692369157, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13788 + }, + { + "epoch": 5.508989212944467, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13789 + }, + { + "epoch": 5.509388733519776, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13790 + }, + { + "epoch": 5.509788254095086, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13791 + }, + { + "epoch": 5.5101877746703956, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13792 + }, + { + "epoch": 5.510587295245705, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 13793 + }, + { + "epoch": 5.510986815821015, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13794 + }, + { + "epoch": 5.511386336396324, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 13795 + }, + { + "epoch": 5.511785856971634, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13796 + }, + { + "epoch": 5.512185377546944, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 13797 + }, + { + "epoch": 5.512584898122253, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13798 + }, + { + "epoch": 5.512984418697563, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 13799 + }, + { + "epoch": 5.513383939272872, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13800 + }, + { + "epoch": 5.513783459848182, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 13801 + }, + { + "epoch": 5.514182980423492, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 13802 + }, + { + "epoch": 5.514582500998801, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13803 + }, + { + "epoch": 5.514982021574111, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13804 + }, + { + "epoch": 5.515381542149421, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 13805 + }, + { + "epoch": 5.515781062724731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 13806 + }, + { + "epoch": 5.51618058330004, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 13807 + }, + { + "epoch": 5.5165801038753495, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 13808 + }, + { + "epoch": 5.516979624450659, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13809 + }, + { + "epoch": 5.517379145025969, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13810 + }, + { + "epoch": 5.517778665601279, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13811 + }, + { + "epoch": 5.518178186176588, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 13812 + }, + { + "epoch": 5.5185777067518975, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 13813 + }, + { + "epoch": 5.518977227327207, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 13814 + }, + { + "epoch": 5.519376747902517, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13815 + }, + { + "epoch": 5.519776268477827, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13816 + }, + { + "epoch": 5.520175789053136, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 13817 + }, + { + "epoch": 5.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 13818 + }, + { + "epoch": 5.520974830203755, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 13819 + }, + { + "epoch": 5.521374350779065, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 13820 + }, + { + "epoch": 5.521773871354375, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13821 + }, + { + "epoch": 5.522173391929685, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 13822 + }, + { + "epoch": 5.5225729125049945, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 13823 + }, + { + "epoch": 5.522972433080303, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13824 + }, + { + "epoch": 5.523371953655613, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13825 + }, + { + "epoch": 5.523771474230923, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 13826 + }, + { + "epoch": 5.524170994806233, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13827 + }, + { + "epoch": 5.5245705153815425, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 13828 + }, + { + "epoch": 5.524970035956851, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 13829 + }, + { + "epoch": 5.525369556532161, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13830 + }, + { + "epoch": 5.525769077107471, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 13831 + }, + { + "epoch": 5.526168597682781, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 13832 + }, + { + "epoch": 5.526568118258091, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 13833 + }, + { + "epoch": 5.5269676388333995, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13834 + }, + { + "epoch": 5.527367159408709, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 13835 + }, + { + "epoch": 5.527766679984019, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13836 + }, + { + "epoch": 5.528166200559329, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13837 + }, + { + "epoch": 5.528565721134639, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 13838 + }, + { + "epoch": 5.528965241709948, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 13839 + }, + { + "epoch": 5.529364762285257, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 13840 + }, + { + "epoch": 5.529764282860567, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 13841 + }, + { + "epoch": 5.530163803435877, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 13842 + }, + { + "epoch": 5.530563324011187, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 13843 + }, + { + "epoch": 5.5309628445864965, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13844 + }, + { + "epoch": 5.531362365161806, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 13845 + }, + { + "epoch": 5.531761885737115, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13846 + }, + { + "epoch": 5.532161406312425, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 13847 + }, + { + "epoch": 5.532560926887735, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 13848 + }, + { + "epoch": 5.5329604474630445, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 13849 + }, + { + "epoch": 5.533359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 13850 + }, + { + "epoch": 5.533759488613663, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13851 + }, + { + "epoch": 5.534159009188973, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 13852 + }, + { + "epoch": 5.534558529764283, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 13853 + }, + { + "epoch": 5.5349580503395925, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13854 + }, + { + "epoch": 5.535357570914902, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13855 + }, + { + "epoch": 5.535757091490212, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 13856 + }, + { + "epoch": 5.536156612065521, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 13857 + }, + { + "epoch": 5.536556132640831, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13858 + }, + { + "epoch": 5.536955653216141, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13859 + }, + { + "epoch": 5.53735517379145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 13860 + }, + { + "epoch": 5.53775469436676, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 13861 + }, + { + "epoch": 5.53815421494207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 13862 + }, + { + "epoch": 5.538553735517379, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 13863 + }, + { + "epoch": 5.538953256092689, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 13864 + }, + { + "epoch": 5.539352776667998, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13865 + }, + { + "epoch": 5.539752297243308, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 13866 + }, + { + "epoch": 5.540151817818618, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 13867 + }, + { + "epoch": 5.540551338393927, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 13868 + }, + { + "epoch": 5.540950858969237, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 13869 + }, + { + "epoch": 5.5413503795445465, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 13870 + }, + { + "epoch": 5.541749900119856, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13871 + }, + { + "epoch": 5.542149420695166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13872 + }, + { + "epoch": 5.542548941270476, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13873 + }, + { + "epoch": 5.542948461845785, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 13874 + }, + { + "epoch": 5.5433479824210945, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 13875 + }, + { + "epoch": 5.543747502996404, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 13876 + }, + { + "epoch": 5.544147023571714, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13877 + }, + { + "epoch": 5.544546544147024, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 13878 + }, + { + "epoch": 5.544946064722334, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13879 + }, + { + "epoch": 5.5453455852976425, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13880 + }, + { + "epoch": 5.545745105872952, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 13881 + }, + { + "epoch": 5.546144626448262, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13882 + }, + { + "epoch": 5.546544147023572, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 13883 + }, + { + "epoch": 5.546943667598882, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 13884 + }, + { + "epoch": 5.547343188174191, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 13885 + }, + { + "epoch": 5.5477427087495, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13886 + }, + { + "epoch": 5.54814222932481, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 13887 + }, + { + "epoch": 5.54854174990012, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 13888 + }, + { + "epoch": 5.54894127047543, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 13889 + }, + { + "epoch": 5.549340791050739, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 13890 + }, + { + "epoch": 5.549740311626048, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13891 + }, + { + "epoch": 5.550139832201358, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13892 + }, + { + "epoch": 5.550539352776668, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13893 + }, + { + "epoch": 5.550938873351978, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 13894 + }, + { + "epoch": 5.551338393927288, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 13895 + }, + { + "epoch": 5.551737914502597, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3353, + "step": 13896 + }, + { + "epoch": 5.552137435077906, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 13897 + }, + { + "epoch": 5.552536955653216, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 13898 + }, + { + "epoch": 5.552936476228526, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 13899 + }, + { + "epoch": 5.553335996803836, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13900 + }, + { + "epoch": 5.553735517379145, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13901 + }, + { + "epoch": 5.554135037954454, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 13902 + }, + { + "epoch": 5.554534558529764, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13903 + }, + { + "epoch": 5.554934079105074, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13904 + }, + { + "epoch": 5.555333599680384, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 13905 + }, + { + "epoch": 5.555733120255693, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13906 + }, + { + "epoch": 5.556132640831002, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13907 + }, + { + "epoch": 5.556532161406312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 13908 + }, + { + "epoch": 5.556931681981622, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 13909 + }, + { + "epoch": 5.557331202556932, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 13910 + }, + { + "epoch": 5.5577307231322415, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13911 + }, + { + "epoch": 5.558130243707551, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13912 + }, + { + "epoch": 5.558529764282861, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 13913 + }, + { + "epoch": 5.55892928485817, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 13914 + }, + { + "epoch": 5.55932880543348, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 13915 + }, + { + "epoch": 5.5597283260087895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 13916 + }, + { + "epoch": 5.560127846584099, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 13917 + }, + { + "epoch": 5.560527367159409, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13918 + }, + { + "epoch": 5.560926887734718, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13919 + }, + { + "epoch": 5.561326408310028, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13920 + }, + { + "epoch": 5.561725928885338, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13921 + }, + { + "epoch": 5.562125449460647, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 13922 + }, + { + "epoch": 5.562524970035957, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 13923 + }, + { + "epoch": 5.562924490611266, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13924 + }, + { + "epoch": 5.563324011186576, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 13925 + }, + { + "epoch": 5.563723531761886, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13926 + }, + { + "epoch": 5.564123052337195, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 13927 + }, + { + "epoch": 5.564522572912505, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13928 + }, + { + "epoch": 5.564922093487815, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13929 + }, + { + "epoch": 5.565321614063124, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 13930 + }, + { + "epoch": 5.565721134638434, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 13931 + }, + { + "epoch": 5.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13932 + }, + { + "epoch": 5.566520175789053, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 13933 + }, + { + "epoch": 5.566919696364363, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13934 + }, + { + "epoch": 5.567319216939673, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 13935 + }, + { + "epoch": 5.567718737514982, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13936 + }, + { + "epoch": 5.5681182580902915, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 13937 + }, + { + "epoch": 5.568517778665601, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 13938 + }, + { + "epoch": 5.568917299240911, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 13939 + }, + { + "epoch": 5.569316819816221, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 13940 + }, + { + "epoch": 5.56971634039153, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 13941 + }, + { + "epoch": 5.5701158609668395, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 13942 + }, + { + "epoch": 5.570515381542149, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 13943 + }, + { + "epoch": 5.570914902117459, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 13944 + }, + { + "epoch": 5.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 13945 + }, + { + "epoch": 5.571713943268079, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 13946 + }, + { + "epoch": 5.572113463843388, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 13947 + }, + { + "epoch": 5.572512984418697, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 13948 + }, + { + "epoch": 5.572912504994007, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 13949 + }, + { + "epoch": 5.573312025569317, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 13950 + }, + { + "epoch": 5.573711546144627, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 13951 + }, + { + "epoch": 5.5741110667199365, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 13952 + }, + { + "epoch": 5.574510587295245, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 13953 + }, + { + "epoch": 5.574910107870555, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13954 + }, + { + "epoch": 5.575309628445865, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 13955 + }, + { + "epoch": 5.575709149021175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 13956 + }, + { + "epoch": 5.5761086695964845, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13957 + }, + { + "epoch": 5.5765081901717934, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 13958 + }, + { + "epoch": 5.576907710747103, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 13959 + }, + { + "epoch": 5.577307231322413, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 13960 + }, + { + "epoch": 5.577706751897723, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 13961 + }, + { + "epoch": 5.578106272473033, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 13962 + }, + { + "epoch": 5.578505793048342, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 13963 + }, + { + "epoch": 5.578905313623651, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 13964 + }, + { + "epoch": 5.579304834198961, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 13965 + }, + { + "epoch": 5.579704354774271, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13966 + }, + { + "epoch": 5.580103875349581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13967 + }, + { + "epoch": 5.58050339592489, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 13968 + }, + { + "epoch": 5.5809029165002, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 13969 + }, + { + "epoch": 5.581302437075509, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 13970 + }, + { + "epoch": 5.581701957650819, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 13971 + }, + { + "epoch": 5.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 13972 + }, + { + "epoch": 5.5825009988014385, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 13973 + }, + { + "epoch": 5.582900519376748, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 13974 + }, + { + "epoch": 5.583300039952057, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 13975 + }, + { + "epoch": 5.583699560527367, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 13976 + }, + { + "epoch": 5.584099081102677, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 13977 + }, + { + "epoch": 5.5844986016779865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 13978 + }, + { + "epoch": 5.584898122253296, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 13979 + }, + { + "epoch": 5.585297642828605, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 13980 + }, + { + "epoch": 5.585697163403915, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 13981 + }, + { + "epoch": 5.586096683979225, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 13982 + }, + { + "epoch": 5.5864962045545346, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 13983 + }, + { + "epoch": 5.586895725129844, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 13984 + }, + { + "epoch": 5.587295245705154, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 13985 + }, + { + "epoch": 5.587694766280464, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 13986 + }, + { + "epoch": 5.588094286855773, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 13987 + }, + { + "epoch": 5.588493807431083, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 13988 + }, + { + "epoch": 5.588893328006392, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 13989 + }, + { + "epoch": 5.589292848581702, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 13990 + }, + { + "epoch": 5.589692369157012, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 13991 + }, + { + "epoch": 5.590091889732321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 13992 + }, + { + "epoch": 5.590491410307631, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 13993 + }, + { + "epoch": 5.59089093088294, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 13994 + }, + { + "epoch": 5.59129045145825, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 13995 + }, + { + "epoch": 5.59168997203356, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 13996 + }, + { + "epoch": 5.592089492608869, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 13997 + }, + { + "epoch": 5.592489013184179, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 13998 + }, + { + "epoch": 5.5928885337594885, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 13999 + }, + { + "epoch": 5.593288054334798, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14000 + }, + { + "epoch": 5.593687574910108, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14001 + }, + { + "epoch": 5.594087095485418, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14002 + }, + { + "epoch": 5.594486616060728, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 14003 + }, + { + "epoch": 5.5948861366360365, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14004 + }, + { + "epoch": 5.595285657211346, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14005 + }, + { + "epoch": 5.595685177786656, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14006 + }, + { + "epoch": 5.596084698361966, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14007 + }, + { + "epoch": 5.596484218937276, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14008 + }, + { + "epoch": 5.5968837395125846, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14009 + }, + { + "epoch": 5.597283260087894, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.2923, + "step": 14010 + }, + { + "epoch": 5.597682780663204, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14011 + }, + { + "epoch": 5.598082301238514, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14012 + }, + { + "epoch": 5.598481821813824, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14013 + }, + { + "epoch": 5.598881342389133, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14014 + }, + { + "epoch": 5.599280862964442, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14015 + }, + { + "epoch": 5.599680383539752, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14016 + }, + { + "epoch": 5.600079904115062, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3283, + "step": 14017 + }, + { + "epoch": 5.600479424690372, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14018 + }, + { + "epoch": 5.6008789452656815, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14019 + }, + { + "epoch": 5.60127846584099, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14020 + }, + { + "epoch": 5.6016779864163, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14021 + }, + { + "epoch": 5.60207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14022 + }, + { + "epoch": 5.60247702756692, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14023 + }, + { + "epoch": 5.60287654814223, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 14024 + }, + { + "epoch": 5.603276068717539, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14025 + }, + { + "epoch": 5.603675589292848, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14026 + }, + { + "epoch": 5.604075109868158, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14027 + }, + { + "epoch": 5.604474630443468, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14028 + }, + { + "epoch": 5.604874151018778, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14029 + }, + { + "epoch": 5.605273671594087, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 14030 + }, + { + "epoch": 5.605673192169396, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14031 + }, + { + "epoch": 5.606072712744706, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14032 + }, + { + "epoch": 5.606472233320016, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14033 + }, + { + "epoch": 5.606871753895326, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14034 + }, + { + "epoch": 5.6072712744706354, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14035 + }, + { + "epoch": 5.607670795045945, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14036 + }, + { + "epoch": 5.608070315621254, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14037 + }, + { + "epoch": 5.608469836196564, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14038 + }, + { + "epoch": 5.608869356771874, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14039 + }, + { + "epoch": 5.6092688773471835, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14040 + }, + { + "epoch": 5.609668397922493, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14041 + }, + { + "epoch": 5.610067918497803, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14042 + }, + { + "epoch": 5.610467439073112, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14043 + }, + { + "epoch": 5.610866959648422, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14044 + }, + { + "epoch": 5.6112664802237315, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14045 + }, + { + "epoch": 5.611666000799041, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14046 + }, + { + "epoch": 5.612065521374351, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14047 + }, + { + "epoch": 5.61246504194966, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14048 + }, + { + "epoch": 5.61286456252497, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14049 + }, + { + "epoch": 5.61326408310028, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14050 + }, + { + "epoch": 5.613663603675589, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14051 + }, + { + "epoch": 5.614063124250899, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14052 + }, + { + "epoch": 5.614462644826209, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14053 + }, + { + "epoch": 5.614862165401518, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14054 + }, + { + "epoch": 5.615261685976828, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14055 + }, + { + "epoch": 5.615661206552137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14056 + }, + { + "epoch": 5.616060727127447, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14057 + }, + { + "epoch": 5.616460247702757, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14058 + }, + { + "epoch": 5.616859768278067, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14059 + }, + { + "epoch": 5.617259288853376, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14060 + }, + { + "epoch": 5.6176588094286855, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14061 + }, + { + "epoch": 5.618058330003995, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14062 + }, + { + "epoch": 5.618457850579305, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14063 + }, + { + "epoch": 5.618857371154615, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14064 + }, + { + "epoch": 5.619256891729924, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14065 + }, + { + "epoch": 5.6196564123052335, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14066 + }, + { + "epoch": 5.620055932880543, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14067 + }, + { + "epoch": 5.620455453455853, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14068 + }, + { + "epoch": 5.620854974031163, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14069 + }, + { + "epoch": 5.621254494606472, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14070 + }, + { + "epoch": 5.6216540151817815, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14071 + }, + { + "epoch": 5.622053535757091, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14072 + }, + { + "epoch": 5.622453056332401, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14073 + }, + { + "epoch": 5.622852576907711, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14074 + }, + { + "epoch": 5.623252097483021, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14075 + }, + { + "epoch": 5.6236516180583305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14076 + }, + { + "epoch": 5.624051138633639, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14077 + }, + { + "epoch": 5.624450659208949, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14078 + }, + { + "epoch": 5.624850179784259, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14079 + }, + { + "epoch": 5.625249700359569, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.2954, + "step": 14080 + }, + { + "epoch": 5.6256492209348785, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14081 + }, + { + "epoch": 5.626048741510187, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14082 + }, + { + "epoch": 5.626448262085497, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14083 + }, + { + "epoch": 5.626847782660807, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14084 + }, + { + "epoch": 5.627247303236117, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 14085 + }, + { + "epoch": 5.627646823811427, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14086 + }, + { + "epoch": 5.6280463443867355, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14087 + }, + { + "epoch": 5.628445864962045, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14088 + }, + { + "epoch": 5.628845385537355, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14089 + }, + { + "epoch": 5.629244906112665, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14090 + }, + { + "epoch": 5.629644426687975, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14091 + }, + { + "epoch": 5.630043947263284, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14092 + }, + { + "epoch": 5.630443467838594, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14093 + }, + { + "epoch": 5.630842988413903, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14094 + }, + { + "epoch": 5.631242508989213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14095 + }, + { + "epoch": 5.631642029564523, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14096 + }, + { + "epoch": 5.632041550139832, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14097 + }, + { + "epoch": 5.632441070715142, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14098 + }, + { + "epoch": 5.632840591290451, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14099 + }, + { + "epoch": 5.633240111865761, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14100 + }, + { + "epoch": 5.633639632441071, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14101 + }, + { + "epoch": 5.6340391530163805, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14102 + }, + { + "epoch": 5.63443867359169, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14103 + }, + { + "epoch": 5.634838194166999, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14104 + }, + { + "epoch": 5.635237714742309, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14105 + }, + { + "epoch": 5.635637235317619, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14106 + }, + { + "epoch": 5.6360367558929285, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14107 + }, + { + "epoch": 5.636436276468238, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14108 + }, + { + "epoch": 5.636835797043548, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14109 + }, + { + "epoch": 5.637235317618857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3315, + "step": 14110 + }, + { + "epoch": 5.637634838194167, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14111 + }, + { + "epoch": 5.638034358769477, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14112 + }, + { + "epoch": 5.638433879344786, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14113 + }, + { + "epoch": 5.638833399920096, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14114 + }, + { + "epoch": 5.639232920495406, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14115 + }, + { + "epoch": 5.639632441070715, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14116 + }, + { + "epoch": 5.640031961646025, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14117 + }, + { + "epoch": 5.640431482221334, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14118 + }, + { + "epoch": 5.640831002796644, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14119 + }, + { + "epoch": 5.641230523371954, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14120 + }, + { + "epoch": 5.641630043947263, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14121 + }, + { + "epoch": 5.642029564522573, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14122 + }, + { + "epoch": 5.642429085097882, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 14123 + }, + { + "epoch": 5.642828605673192, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14124 + }, + { + "epoch": 5.643228126248502, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14125 + }, + { + "epoch": 5.643627646823812, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14126 + }, + { + "epoch": 5.644027167399121, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14127 + }, + { + "epoch": 5.6444266879744305, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14128 + }, + { + "epoch": 5.64482620854974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14129 + }, + { + "epoch": 5.64522572912505, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14130 + }, + { + "epoch": 5.64562524970036, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14131 + }, + { + "epoch": 5.64602477027567, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14132 + }, + { + "epoch": 5.6464242908509785, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14133 + }, + { + "epoch": 5.646823811426288, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14134 + }, + { + "epoch": 5.647223332001598, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14135 + }, + { + "epoch": 5.647622852576908, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14136 + }, + { + "epoch": 5.648022373152218, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14137 + }, + { + "epoch": 5.648421893727527, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14138 + }, + { + "epoch": 5.648821414302836, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14139 + }, + { + "epoch": 5.649220934878146, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14140 + }, + { + "epoch": 5.649620455453456, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14141 + }, + { + "epoch": 5.650019976028766, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14142 + }, + { + "epoch": 5.6504194966040755, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14143 + }, + { + "epoch": 5.650819017179384, + "grad_norm": 0.046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14144 + }, + { + "epoch": 5.651218537754694, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14145 + }, + { + "epoch": 5.651618058330004, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14146 + }, + { + "epoch": 5.652017578905314, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14147 + }, + { + "epoch": 5.6524170994806235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14148 + }, + { + "epoch": 5.652816620055933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14149 + }, + { + "epoch": 5.653216140631242, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 14150 + }, + { + "epoch": 5.653615661206552, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14151 + }, + { + "epoch": 5.654015181781862, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14152 + }, + { + "epoch": 5.654414702357172, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14153 + }, + { + "epoch": 5.654814222932481, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14154 + }, + { + "epoch": 5.65521374350779, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14155 + }, + { + "epoch": 5.6556132640831, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14156 + }, + { + "epoch": 5.65601278465841, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14157 + }, + { + "epoch": 5.65641230523372, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14158 + }, + { + "epoch": 5.656811825809029, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14159 + }, + { + "epoch": 5.657211346384338, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14160 + }, + { + "epoch": 5.657610866959648, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 14161 + }, + { + "epoch": 5.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14162 + }, + { + "epoch": 5.658409908110268, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14163 + }, + { + "epoch": 5.6588094286855775, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14164 + }, + { + "epoch": 5.659208949260887, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14165 + }, + { + "epoch": 5.659608469836197, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14166 + }, + { + "epoch": 5.660007990411506, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14167 + }, + { + "epoch": 5.660407510986816, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14168 + }, + { + "epoch": 5.6608070315621255, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14169 + }, + { + "epoch": 5.661206552137435, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14170 + }, + { + "epoch": 5.661606072712745, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14171 + }, + { + "epoch": 5.662005593288054, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14172 + }, + { + "epoch": 5.662405113863364, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14173 + }, + { + "epoch": 5.6628046344386735, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14174 + }, + { + "epoch": 5.663204155013983, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14175 + }, + { + "epoch": 5.663603675589293, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3367, + "step": 14176 + }, + { + "epoch": 5.664003196164602, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 14177 + }, + { + "epoch": 5.664402716739912, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14178 + }, + { + "epoch": 5.664802237315222, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14179 + }, + { + "epoch": 5.665201757890531, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14180 + }, + { + "epoch": 5.665601278465841, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14181 + }, + { + "epoch": 5.666000799041151, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14182 + }, + { + "epoch": 5.666400319616461, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14183 + }, + { + "epoch": 5.66679984019177, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14184 + }, + { + "epoch": 5.667199360767079, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14185 + }, + { + "epoch": 5.667598881342389, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14186 + }, + { + "epoch": 5.667998401917699, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14187 + }, + { + "epoch": 5.668397922493009, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14188 + }, + { + "epoch": 5.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14189 + }, + { + "epoch": 5.6691969636436275, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14190 + }, + { + "epoch": 5.669596484218937, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14191 + }, + { + "epoch": 5.669996004794247, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14192 + }, + { + "epoch": 5.670395525369557, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14193 + }, + { + "epoch": 5.670795045944866, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14194 + }, + { + "epoch": 5.6711945665201755, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14195 + }, + { + "epoch": 5.671594087095485, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14196 + }, + { + "epoch": 5.671993607670795, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14197 + }, + { + "epoch": 5.672393128246105, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14198 + }, + { + "epoch": 5.672792648821415, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14199 + }, + { + "epoch": 5.6731921693967235, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14200 + }, + { + "epoch": 5.673591689972033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14201 + }, + { + "epoch": 5.673991210547343, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14202 + }, + { + "epoch": 5.674390731122653, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14203 + }, + { + "epoch": 5.674790251697963, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14204 + }, + { + "epoch": 5.6751897722732725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 14205 + }, + { + "epoch": 5.675589292848581, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14206 + }, + { + "epoch": 5.675988813423891, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14207 + }, + { + "epoch": 5.676388333999201, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14208 + }, + { + "epoch": 5.676787854574511, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14209 + }, + { + "epoch": 5.6771873751498205, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14210 + }, + { + "epoch": 5.677586895725129, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14211 + }, + { + "epoch": 5.677986416300439, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14212 + }, + { + "epoch": 5.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14213 + }, + { + "epoch": 5.678785457451059, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14214 + }, + { + "epoch": 5.679184978026369, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14215 + }, + { + "epoch": 5.679584498601678, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14216 + }, + { + "epoch": 5.679984019176987, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 14217 + }, + { + "epoch": 5.680383539752297, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14218 + }, + { + "epoch": 5.680783060327607, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14219 + }, + { + "epoch": 5.681182580902917, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14220 + }, + { + "epoch": 5.681582101478226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14221 + }, + { + "epoch": 5.681981622053536, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14222 + }, + { + "epoch": 5.682381142628845, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14223 + }, + { + "epoch": 5.682780663204155, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14224 + }, + { + "epoch": 5.683180183779465, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14225 + }, + { + "epoch": 5.683579704354774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 14226 + }, + { + "epoch": 5.683979224930084, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14227 + }, + { + "epoch": 5.684378745505393, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14228 + }, + { + "epoch": 5.684778266080703, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14229 + }, + { + "epoch": 5.685177786656013, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14230 + }, + { + "epoch": 5.6855773072313225, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14231 + }, + { + "epoch": 5.685976827806632, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14232 + }, + { + "epoch": 5.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14233 + }, + { + "epoch": 5.686775868957251, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14234 + }, + { + "epoch": 5.687175389532561, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14235 + }, + { + "epoch": 5.6875749101078705, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14236 + }, + { + "epoch": 5.68797443068318, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14237 + }, + { + "epoch": 5.68837395125849, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14238 + }, + { + "epoch": 5.6887734718338, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14239 + }, + { + "epoch": 5.689172992409109, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14240 + }, + { + "epoch": 5.689572512984419, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14241 + }, + { + "epoch": 5.689972033559728, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 14242 + }, + { + "epoch": 5.690371554135038, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 14243 + }, + { + "epoch": 5.690771074710348, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14244 + }, + { + "epoch": 5.691170595285657, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14245 + }, + { + "epoch": 5.691570115860967, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14246 + }, + { + "epoch": 5.691969636436276, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14247 + }, + { + "epoch": 5.692369157011586, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14248 + }, + { + "epoch": 5.692768677586896, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14249 + }, + { + "epoch": 5.693168198162205, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14250 + }, + { + "epoch": 5.693567718737515, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 14251 + }, + { + "epoch": 5.6939672393128244, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14252 + }, + { + "epoch": 5.694366759888134, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14253 + }, + { + "epoch": 5.694766280463444, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14254 + }, + { + "epoch": 5.695165801038754, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14255 + }, + { + "epoch": 5.695565321614064, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14256 + }, + { + "epoch": 5.6959648421893725, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14257 + }, + { + "epoch": 5.696364362764682, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14258 + }, + { + "epoch": 5.696763883339992, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14259 + }, + { + "epoch": 5.697163403915302, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14260 + }, + { + "epoch": 5.697562924490612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14261 + }, + { + "epoch": 5.6979624450659205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14262 + }, + { + "epoch": 5.69836196564123, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14263 + }, + { + "epoch": 5.69876148621654, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14264 + }, + { + "epoch": 5.69916100679185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14265 + }, + { + "epoch": 5.69956052736716, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14266 + }, + { + "epoch": 5.699960047942469, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14267 + }, + { + "epoch": 5.700359568517778, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 14268 + }, + { + "epoch": 5.700759089093088, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14269 + }, + { + "epoch": 5.701158609668398, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14270 + }, + { + "epoch": 5.701558130243708, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14271 + }, + { + "epoch": 5.7019576508190175, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 14272 + }, + { + "epoch": 5.702357171394327, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14273 + }, + { + "epoch": 5.702756691969636, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14274 + }, + { + "epoch": 5.703156212544946, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14275 + }, + { + "epoch": 5.703555733120256, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14276 + }, + { + "epoch": 5.7039552536955656, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14277 + }, + { + "epoch": 5.704354774270875, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14278 + }, + { + "epoch": 5.704754294846184, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 14279 + }, + { + "epoch": 5.705153815421494, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14280 + }, + { + "epoch": 5.705553335996804, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14281 + }, + { + "epoch": 5.705952856572114, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14282 + }, + { + "epoch": 5.706352377147423, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14283 + }, + { + "epoch": 5.706751897722732, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14284 + }, + { + "epoch": 5.707151418298042, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14285 + }, + { + "epoch": 5.707550938873352, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14286 + }, + { + "epoch": 5.707950459448662, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14287 + }, + { + "epoch": 5.708349980023971, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14288 + }, + { + "epoch": 5.708749500599281, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14289 + }, + { + "epoch": 5.70914902117459, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 14290 + }, + { + "epoch": 5.7095485417499, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14291 + }, + { + "epoch": 5.70994806232521, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14292 + }, + { + "epoch": 5.7103475829005195, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14293 + }, + { + "epoch": 5.710747103475829, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14294 + }, + { + "epoch": 5.711146624051139, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14295 + }, + { + "epoch": 5.711546144626448, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14296 + }, + { + "epoch": 5.711945665201758, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 14297 + }, + { + "epoch": 5.7123451857770675, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14298 + }, + { + "epoch": 5.712744706352377, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 14299 + }, + { + "epoch": 5.713144226927687, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14300 + }, + { + "epoch": 5.713543747502996, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14301 + }, + { + "epoch": 5.713943268078306, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14302 + }, + { + "epoch": 5.714342788653616, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14303 + }, + { + "epoch": 5.714742309228925, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14304 + }, + { + "epoch": 5.715141829804235, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14305 + }, + { + "epoch": 5.715541350379545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14306 + }, + { + "epoch": 5.715940870954854, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14307 + }, + { + "epoch": 5.716340391530164, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14308 + }, + { + "epoch": 5.716739912105473, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14309 + }, + { + "epoch": 5.717139432680783, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14310 + }, + { + "epoch": 5.717538953256093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14311 + }, + { + "epoch": 5.717938473831403, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14312 + }, + { + "epoch": 5.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14313 + }, + { + "epoch": 5.718737514982021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14314 + }, + { + "epoch": 5.719137035557331, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14315 + }, + { + "epoch": 5.719536556132641, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14316 + }, + { + "epoch": 5.719936076707951, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 14317 + }, + { + "epoch": 5.72033559728326, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14318 + }, + { + "epoch": 5.7207351178585695, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14319 + }, + { + "epoch": 5.721134638433879, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14320 + }, + { + "epoch": 5.721534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14321 + }, + { + "epoch": 5.721933679584499, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14322 + }, + { + "epoch": 5.722333200159808, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14323 + }, + { + "epoch": 5.7227327207351175, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 14324 + }, + { + "epoch": 5.723132241310427, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14325 + }, + { + "epoch": 5.723531761885737, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14326 + }, + { + "epoch": 5.723931282461047, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14327 + }, + { + "epoch": 5.724330803036357, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14328 + }, + { + "epoch": 5.7247303236116664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14329 + }, + { + "epoch": 5.725129844186975, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14330 + }, + { + "epoch": 5.725529364762285, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14331 + }, + { + "epoch": 5.725928885337595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.301, + "step": 14332 + }, + { + "epoch": 5.726328405912905, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14333 + }, + { + "epoch": 5.7267279264882145, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14334 + }, + { + "epoch": 5.727127447063523, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14335 + }, + { + "epoch": 5.727526967638833, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14336 + }, + { + "epoch": 5.727926488214143, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14337 + }, + { + "epoch": 5.728326008789453, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 14338 + }, + { + "epoch": 5.7287255293647625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14339 + }, + { + "epoch": 5.729125049940071, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14340 + }, + { + "epoch": 5.729524570515381, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14341 + }, + { + "epoch": 5.729924091090691, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14342 + }, + { + "epoch": 5.730323611666001, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14343 + }, + { + "epoch": 5.730723132241311, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14344 + }, + { + "epoch": 5.73112265281662, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14345 + }, + { + "epoch": 5.73152217339193, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14346 + }, + { + "epoch": 5.731921693967239, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14347 + }, + { + "epoch": 5.732321214542549, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14348 + }, + { + "epoch": 5.732720735117859, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14349 + }, + { + "epoch": 5.733120255693168, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14350 + }, + { + "epoch": 5.733519776268478, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14351 + }, + { + "epoch": 5.733919296843787, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 14352 + }, + { + "epoch": 5.734318817419097, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14353 + }, + { + "epoch": 5.734718337994407, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 14354 + }, + { + "epoch": 5.7351178585697165, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14355 + }, + { + "epoch": 5.735517379145026, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14356 + }, + { + "epoch": 5.735916899720335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14357 + }, + { + "epoch": 5.736316420295645, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14358 + }, + { + "epoch": 5.736715940870955, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 14359 + }, + { + "epoch": 5.7371154614462645, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14360 + }, + { + "epoch": 5.737514982021574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14361 + }, + { + "epoch": 5.737914502596884, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14362 + }, + { + "epoch": 5.738314023172194, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14363 + }, + { + "epoch": 5.738713543747503, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14364 + }, + { + "epoch": 5.7391130643228125, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14365 + }, + { + "epoch": 5.739512584898122, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14366 + }, + { + "epoch": 5.739912105473432, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 14367 + }, + { + "epoch": 5.740311626048742, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14368 + }, + { + "epoch": 5.740711146624051, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14369 + }, + { + "epoch": 5.741110667199361, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14370 + }, + { + "epoch": 5.74151018777467, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14371 + }, + { + "epoch": 5.74190970834998, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14372 + }, + { + "epoch": 5.74230922892529, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14373 + }, + { + "epoch": 5.742708749500599, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 14374 + }, + { + "epoch": 5.743108270075909, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 14375 + }, + { + "epoch": 5.743507790651218, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14376 + }, + { + "epoch": 5.743907311226528, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14377 + }, + { + "epoch": 5.744306831801838, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14378 + }, + { + "epoch": 5.744706352377148, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14379 + }, + { + "epoch": 5.745105872952457, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14380 + }, + { + "epoch": 5.7455053935277665, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 14381 + }, + { + "epoch": 5.745904914103076, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14382 + }, + { + "epoch": 5.746304434678386, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 14383 + }, + { + "epoch": 5.746703955253696, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14384 + }, + { + "epoch": 5.747103475829006, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14385 + }, + { + "epoch": 5.7475029964043145, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14386 + }, + { + "epoch": 5.747902516979624, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14387 + }, + { + "epoch": 5.748302037554934, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14388 + }, + { + "epoch": 5.748701558130244, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14389 + }, + { + "epoch": 5.749101078705554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14390 + }, + { + "epoch": 5.7495005992808625, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14391 + }, + { + "epoch": 5.749900119856172, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14392 + }, + { + "epoch": 5.750299640431482, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14393 + }, + { + "epoch": 5.750699161006792, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 14394 + }, + { + "epoch": 5.751098681582102, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 14395 + }, + { + "epoch": 5.7514982021574115, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14396 + }, + { + "epoch": 5.75189772273272, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 14397 + }, + { + "epoch": 5.75229724330803, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14398 + }, + { + "epoch": 5.75269676388334, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14399 + }, + { + "epoch": 5.75309628445865, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14400 + }, + { + "epoch": 5.7534958050339595, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14401 + }, + { + "epoch": 5.753895325609269, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14402 + }, + { + "epoch": 5.754294846184578, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14403 + }, + { + "epoch": 5.754694366759888, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14404 + }, + { + "epoch": 5.755093887335198, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14405 + }, + { + "epoch": 5.755493407910508, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14406 + }, + { + "epoch": 5.755892928485817, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14407 + }, + { + "epoch": 5.756292449061126, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14408 + }, + { + "epoch": 5.756691969636436, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14409 + }, + { + "epoch": 5.757091490211746, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14410 + }, + { + "epoch": 5.757491010787056, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14411 + }, + { + "epoch": 5.757890531362365, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14412 + }, + { + "epoch": 5.758290051937674, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14413 + }, + { + "epoch": 5.758689572512984, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14414 + }, + { + "epoch": 5.759089093088294, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14415 + }, + { + "epoch": 5.759488613663604, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14416 + }, + { + "epoch": 5.759888134238913, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14417 + }, + { + "epoch": 5.760287654814223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14418 + }, + { + "epoch": 5.760687175389533, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14419 + }, + { + "epoch": 5.761086695964842, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 14420 + }, + { + "epoch": 5.761486216540152, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14421 + }, + { + "epoch": 5.7618857371154615, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14422 + }, + { + "epoch": 5.762285257690771, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14423 + }, + { + "epoch": 5.762684778266081, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14424 + }, + { + "epoch": 5.76308429884139, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14425 + }, + { + "epoch": 5.7634838194167, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14426 + }, + { + "epoch": 5.7638833399920095, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14427 + }, + { + "epoch": 5.764282860567319, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14428 + }, + { + "epoch": 5.764682381142629, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14429 + }, + { + "epoch": 5.765081901717938, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14430 + }, + { + "epoch": 5.765481422293248, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14431 + }, + { + "epoch": 5.765880942868558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14432 + }, + { + "epoch": 5.766280463443867, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14433 + }, + { + "epoch": 5.766679984019177, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14434 + }, + { + "epoch": 5.767079504594487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14435 + }, + { + "epoch": 5.767479025169797, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14436 + }, + { + "epoch": 5.767878545745106, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 14437 + }, + { + "epoch": 5.768278066320415, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14438 + }, + { + "epoch": 5.768677586895725, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14439 + }, + { + "epoch": 5.769077107471035, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14440 + }, + { + "epoch": 5.769476628046345, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14441 + }, + { + "epoch": 5.769876148621654, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14442 + }, + { + "epoch": 5.770275669196963, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14443 + }, + { + "epoch": 5.770675189772273, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14444 + }, + { + "epoch": 5.771074710347583, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14445 + }, + { + "epoch": 5.771474230922893, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14446 + }, + { + "epoch": 5.771873751498202, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 14447 + }, + { + "epoch": 5.7722732720735115, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14448 + }, + { + "epoch": 5.772672792648821, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14449 + }, + { + "epoch": 5.773072313224131, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14450 + }, + { + "epoch": 5.773471833799441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14451 + }, + { + "epoch": 5.773871354374751, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14452 + }, + { + "epoch": 5.77427087495006, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14453 + }, + { + "epoch": 5.774670395525369, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14454 + }, + { + "epoch": 5.775069916100679, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14455 + }, + { + "epoch": 5.775469436675989, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14456 + }, + { + "epoch": 5.775868957251299, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14457 + }, + { + "epoch": 5.7762684778266085, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14458 + }, + { + "epoch": 5.776667998401917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14459 + }, + { + "epoch": 5.777067518977227, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14460 + }, + { + "epoch": 5.777467039552537, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.331, + "step": 14461 + }, + { + "epoch": 5.777866560127847, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14462 + }, + { + "epoch": 5.7782660807031565, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14463 + }, + { + "epoch": 5.778665601278465, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14464 + }, + { + "epoch": 5.779065121853775, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 14465 + }, + { + "epoch": 5.779464642429085, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14466 + }, + { + "epoch": 5.779864163004395, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14467 + }, + { + "epoch": 5.7802636835797045, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14468 + }, + { + "epoch": 5.780663204155014, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14469 + }, + { + "epoch": 5.781062724730323, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14470 + }, + { + "epoch": 5.781462245305633, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14471 + }, + { + "epoch": 5.781861765880943, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14472 + }, + { + "epoch": 5.782261286456253, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14473 + }, + { + "epoch": 5.782660807031562, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14474 + }, + { + "epoch": 5.783060327606872, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14475 + }, + { + "epoch": 5.783459848182181, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14476 + }, + { + "epoch": 5.783859368757491, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14477 + }, + { + "epoch": 5.784258889332801, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14478 + }, + { + "epoch": 5.78465840990811, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14479 + }, + { + "epoch": 5.78505793048342, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14480 + }, + { + "epoch": 5.785457451058729, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 14481 + }, + { + "epoch": 5.785856971634039, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 14482 + }, + { + "epoch": 5.786256492209349, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14483 + }, + { + "epoch": 5.7866560127846585, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14484 + }, + { + "epoch": 5.787055533359968, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 14485 + }, + { + "epoch": 5.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 14486 + }, + { + "epoch": 5.787854574510587, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14487 + }, + { + "epoch": 5.788254095085897, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14488 + }, + { + "epoch": 5.7886536156612065, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14489 + }, + { + "epoch": 5.789053136236516, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14490 + }, + { + "epoch": 5.789452656811826, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14491 + }, + { + "epoch": 5.789852177387136, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14492 + }, + { + "epoch": 5.790251697962445, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14493 + }, + { + "epoch": 5.7906512185377546, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14494 + }, + { + "epoch": 5.791050739113064, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14495 + }, + { + "epoch": 5.791450259688374, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 14496 + }, + { + "epoch": 5.791849780263684, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14497 + }, + { + "epoch": 5.792249300838993, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14498 + }, + { + "epoch": 5.792648821414303, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14499 + }, + { + "epoch": 5.793048341989612, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 14500 + }, + { + "epoch": 5.793447862564922, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 14501 + }, + { + "epoch": 5.793847383140232, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14502 + }, + { + "epoch": 5.794246903715541, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14503 + }, + { + "epoch": 5.794646424290851, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.2985, + "step": 14504 + }, + { + "epoch": 5.79504594486616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14505 + }, + { + "epoch": 5.79544546544147, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14506 + }, + { + "epoch": 5.79584498601678, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14507 + }, + { + "epoch": 5.79624450659209, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14508 + }, + { + "epoch": 5.7966440271674, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14509 + }, + { + "epoch": 5.7970435477427085, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14510 + }, + { + "epoch": 5.797443068318018, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14511 + }, + { + "epoch": 5.797842588893328, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14512 + }, + { + "epoch": 5.798242109468638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14513 + }, + { + "epoch": 5.798641630043948, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14514 + }, + { + "epoch": 5.7990411506192565, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14515 + }, + { + "epoch": 5.799440671194566, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14516 + }, + { + "epoch": 5.799840191769876, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3329, + "step": 14517 + }, + { + "epoch": 5.800239712345186, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14518 + }, + { + "epoch": 5.800639232920496, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 14519 + }, + { + "epoch": 5.801038753495805, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14520 + }, + { + "epoch": 5.801438274071114, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14521 + }, + { + "epoch": 5.801837794646424, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14522 + }, + { + "epoch": 5.802237315221734, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14523 + }, + { + "epoch": 5.802636835797044, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 14524 + }, + { + "epoch": 5.8030363563723535, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14525 + }, + { + "epoch": 5.803435876947663, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14526 + }, + { + "epoch": 5.803835397522972, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14527 + }, + { + "epoch": 5.804234918098282, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14528 + }, + { + "epoch": 5.804634438673592, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14529 + }, + { + "epoch": 5.8050339592489015, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14530 + }, + { + "epoch": 5.805433479824211, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14531 + }, + { + "epoch": 5.80583300039952, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 14532 + }, + { + "epoch": 5.80623252097483, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14533 + }, + { + "epoch": 5.80663204155014, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14534 + }, + { + "epoch": 5.80703156212545, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14535 + }, + { + "epoch": 5.807431082700759, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3368, + "step": 14536 + }, + { + "epoch": 5.807830603276068, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14537 + }, + { + "epoch": 5.808230123851378, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14538 + }, + { + "epoch": 5.808629644426688, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14539 + }, + { + "epoch": 5.809029165001998, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14540 + }, + { + "epoch": 5.809428685577307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14541 + }, + { + "epoch": 5.809828206152617, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14542 + }, + { + "epoch": 5.810227726727927, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14543 + }, + { + "epoch": 5.810627247303236, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14544 + }, + { + "epoch": 5.811026767878546, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14545 + }, + { + "epoch": 5.8114262884538554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14546 + }, + { + "epoch": 5.811825809029165, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14547 + }, + { + "epoch": 5.812225329604475, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14548 + }, + { + "epoch": 5.812624850179784, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14549 + }, + { + "epoch": 5.813024370755094, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14550 + }, + { + "epoch": 5.8134238913304035, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 14551 + }, + { + "epoch": 5.813823411905713, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14552 + }, + { + "epoch": 5.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14553 + }, + { + "epoch": 5.814622453056332, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14554 + }, + { + "epoch": 5.815021973631642, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14555 + }, + { + "epoch": 5.8154214942069515, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 14556 + }, + { + "epoch": 5.815821014782261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14557 + }, + { + "epoch": 5.816220535357571, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14558 + }, + { + "epoch": 5.816620055932881, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14559 + }, + { + "epoch": 5.81701957650819, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14560 + }, + { + "epoch": 5.8174190970835, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14561 + }, + { + "epoch": 5.817818617658809, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14562 + }, + { + "epoch": 5.818218138234119, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 14563 + }, + { + "epoch": 5.818617658809429, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 14564 + }, + { + "epoch": 5.819017179384739, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14565 + }, + { + "epoch": 5.819416699960048, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 14566 + }, + { + "epoch": 5.819816220535357, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14567 + }, + { + "epoch": 5.820215741110667, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14568 + }, + { + "epoch": 5.820615261685977, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14569 + }, + { + "epoch": 5.821014782261287, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14570 + }, + { + "epoch": 5.821414302836596, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14571 + }, + { + "epoch": 5.8218138234119055, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14572 + }, + { + "epoch": 5.822213343987215, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14573 + }, + { + "epoch": 5.822612864562525, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14574 + }, + { + "epoch": 5.823012385137835, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14575 + }, + { + "epoch": 5.823411905713145, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14576 + }, + { + "epoch": 5.8238114262884535, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14577 + }, + { + "epoch": 5.824210946863763, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14578 + }, + { + "epoch": 5.824610467439073, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14579 + }, + { + "epoch": 5.825009988014383, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14580 + }, + { + "epoch": 5.825409508589693, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14581 + }, + { + "epoch": 5.825809029165002, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14582 + }, + { + "epoch": 5.826208549740311, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14583 + }, + { + "epoch": 5.826608070315621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14584 + }, + { + "epoch": 5.827007590890931, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14585 + }, + { + "epoch": 5.827407111466241, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14586 + }, + { + "epoch": 5.8278066320415505, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14587 + }, + { + "epoch": 5.828206152616859, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 14588 + }, + { + "epoch": 5.828605673192169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14589 + }, + { + "epoch": 5.829005193767479, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14590 + }, + { + "epoch": 5.829404714342789, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14591 + }, + { + "epoch": 5.8298042349180985, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14592 + }, + { + "epoch": 5.830203755493407, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14593 + }, + { + "epoch": 5.830603276068717, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14594 + }, + { + "epoch": 5.831002796644027, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14595 + }, + { + "epoch": 5.831402317219337, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14596 + }, + { + "epoch": 5.831801837794647, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 14597 + }, + { + "epoch": 5.832201358369956, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14598 + }, + { + "epoch": 5.832600878945266, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14599 + }, + { + "epoch": 5.833000399520575, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 14600 + }, + { + "epoch": 5.833399920095885, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14601 + }, + { + "epoch": 5.833799440671195, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14602 + }, + { + "epoch": 5.834198961246504, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14603 + }, + { + "epoch": 5.834598481821814, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14604 + }, + { + "epoch": 5.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 14605 + }, + { + "epoch": 5.835397522972433, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 14606 + }, + { + "epoch": 5.835797043547743, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14607 + }, + { + "epoch": 5.836196564123052, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14608 + }, + { + "epoch": 5.836596084698362, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14609 + }, + { + "epoch": 5.836995605273671, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14610 + }, + { + "epoch": 5.837395125848981, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14611 + }, + { + "epoch": 5.837794646424291, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14612 + }, + { + "epoch": 5.8381941669996005, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14613 + }, + { + "epoch": 5.83859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14614 + }, + { + "epoch": 5.83899320815022, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14615 + }, + { + "epoch": 5.83939272872553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14616 + }, + { + "epoch": 5.839792249300839, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14617 + }, + { + "epoch": 5.8401917698761485, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 14618 + }, + { + "epoch": 5.840591290451458, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 14619 + }, + { + "epoch": 5.840990811026768, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 14620 + }, + { + "epoch": 5.841390331602078, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14621 + }, + { + "epoch": 5.841789852177387, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14622 + }, + { + "epoch": 5.842189372752697, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14623 + }, + { + "epoch": 5.842588893328006, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 14624 + }, + { + "epoch": 5.842988413903316, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 14625 + }, + { + "epoch": 5.843387934478626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14626 + }, + { + "epoch": 5.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 14627 + }, + { + "epoch": 5.844186975629245, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14628 + }, + { + "epoch": 5.844586496204554, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14629 + }, + { + "epoch": 5.844986016779864, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14630 + }, + { + "epoch": 5.845385537355174, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14631 + }, + { + "epoch": 5.845785057930484, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14632 + }, + { + "epoch": 5.8461845785057935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 14633 + }, + { + "epoch": 5.846584099081102, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14634 + }, + { + "epoch": 5.846983619656412, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 14635 + }, + { + "epoch": 5.847383140231722, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14636 + }, + { + "epoch": 5.847782660807032, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14637 + }, + { + "epoch": 5.848182181382342, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14638 + }, + { + "epoch": 5.8485817019576505, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14639 + }, + { + "epoch": 5.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14640 + }, + { + "epoch": 5.84938074310827, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14641 + }, + { + "epoch": 5.84978026368358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 14642 + }, + { + "epoch": 5.85017978425889, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14643 + }, + { + "epoch": 5.8505793048341985, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14644 + }, + { + "epoch": 5.850978825409508, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14645 + }, + { + "epoch": 5.851378345984818, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14646 + }, + { + "epoch": 5.851777866560128, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 14647 + }, + { + "epoch": 5.852177387135438, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14648 + }, + { + "epoch": 5.8525769077107475, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 14649 + }, + { + "epoch": 5.852976428286056, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14650 + }, + { + "epoch": 5.853375948861366, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 14651 + }, + { + "epoch": 5.853775469436676, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14652 + }, + { + "epoch": 5.854174990011986, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 14653 + }, + { + "epoch": 5.8545745105872955, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14654 + }, + { + "epoch": 5.854974031162605, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 14655 + }, + { + "epoch": 5.855373551737914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14656 + }, + { + "epoch": 5.855773072313224, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 14657 + }, + { + "epoch": 5.856172592888534, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14658 + }, + { + "epoch": 5.8565721134638435, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14659 + }, + { + "epoch": 5.856971634039153, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14660 + }, + { + "epoch": 5.857371154614462, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14661 + }, + { + "epoch": 5.857770675189772, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14662 + }, + { + "epoch": 5.858170195765082, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14663 + }, + { + "epoch": 5.858569716340392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14664 + }, + { + "epoch": 5.858969236915701, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 14665 + }, + { + "epoch": 5.859368757491011, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14666 + }, + { + "epoch": 5.85976827806632, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14667 + }, + { + "epoch": 5.86016779864163, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 14668 + }, + { + "epoch": 5.86056731921694, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 14669 + }, + { + "epoch": 5.860966839792249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14670 + }, + { + "epoch": 5.861366360367559, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14671 + }, + { + "epoch": 5.861765880942869, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14672 + }, + { + "epoch": 5.862165401518178, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14673 + }, + { + "epoch": 5.862564922093488, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 14674 + }, + { + "epoch": 5.8629644426687975, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14675 + }, + { + "epoch": 5.863363963244107, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 14676 + }, + { + "epoch": 5.863763483819417, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 14677 + }, + { + "epoch": 5.864163004394726, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14678 + }, + { + "epoch": 5.864562524970036, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14679 + }, + { + "epoch": 5.8649620455453455, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3349, + "step": 14680 + }, + { + "epoch": 5.865361566120655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14681 + }, + { + "epoch": 5.865761086695965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14682 + }, + { + "epoch": 5.866160607271274, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 14683 + }, + { + "epoch": 5.866560127846584, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14684 + }, + { + "epoch": 5.8669596484218935, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14685 + }, + { + "epoch": 5.867359168997203, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 14686 + }, + { + "epoch": 5.867758689572513, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14687 + }, + { + "epoch": 5.868158210147823, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14688 + }, + { + "epoch": 5.868557730723133, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14689 + }, + { + "epoch": 5.868957251298442, + "grad_norm": 0.12255859375, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 14690 + }, + { + "epoch": 5.869356771873751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14691 + }, + { + "epoch": 5.869756292449061, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14692 + }, + { + "epoch": 5.870155813024371, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 14693 + }, + { + "epoch": 5.870555333599681, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14694 + }, + { + "epoch": 5.87095485417499, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14695 + }, + { + "epoch": 5.871354374750299, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 14696 + }, + { + "epoch": 5.871753895325609, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14697 + }, + { + "epoch": 5.872153415900919, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14698 + }, + { + "epoch": 5.872552936476229, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 14699 + }, + { + "epoch": 5.872952457051538, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 14700 + }, + { + "epoch": 5.8733519776268475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14701 + }, + { + "epoch": 5.873751498202157, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 14702 + }, + { + "epoch": 5.874151018777467, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 14703 + }, + { + "epoch": 5.874550539352777, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14704 + }, + { + "epoch": 5.874950059928087, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14705 + }, + { + "epoch": 5.875349580503396, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14706 + }, + { + "epoch": 5.875749101078705, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14707 + }, + { + "epoch": 5.876148621654015, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14708 + }, + { + "epoch": 5.876548142229325, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 14709 + }, + { + "epoch": 5.876947662804635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 14710 + }, + { + "epoch": 5.877347183379944, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14711 + }, + { + "epoch": 5.877746703955253, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14712 + }, + { + "epoch": 5.878146224530563, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 14713 + }, + { + "epoch": 5.878545745105873, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 14714 + }, + { + "epoch": 5.878945265681183, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14715 + }, + { + "epoch": 5.8793447862564925, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 14716 + }, + { + "epoch": 5.879744306831801, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 14717 + }, + { + "epoch": 5.880143827407111, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14718 + }, + { + "epoch": 5.880543347982421, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14719 + }, + { + "epoch": 5.880942868557731, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 14720 + }, + { + "epoch": 5.8813423891330405, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14721 + }, + { + "epoch": 5.88174190970835, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14722 + }, + { + "epoch": 5.88214143028366, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 14723 + }, + { + "epoch": 5.882540950858969, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 14724 + }, + { + "epoch": 5.882940471434279, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14725 + }, + { + "epoch": 5.883339992009589, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 14726 + }, + { + "epoch": 5.883739512584898, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 14727 + }, + { + "epoch": 5.884139033160208, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14728 + }, + { + "epoch": 5.884538553735517, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14729 + }, + { + "epoch": 5.884938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 14730 + }, + { + "epoch": 5.885337594886137, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 14731 + }, + { + "epoch": 5.885737115461446, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14732 + }, + { + "epoch": 5.886136636036756, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14733 + }, + { + "epoch": 5.886536156612065, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 14734 + }, + { + "epoch": 5.886935677187375, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14735 + }, + { + "epoch": 5.887335197762685, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 14736 + }, + { + "epoch": 5.887734718337994, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 14737 + }, + { + "epoch": 5.888134238913304, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 14738 + }, + { + "epoch": 5.888533759488614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 14739 + }, + { + "epoch": 5.888933280063923, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14740 + }, + { + "epoch": 5.889332800639233, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3318, + "step": 14741 + }, + { + "epoch": 5.8897323212145425, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 14742 + }, + { + "epoch": 5.890131841789852, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 14743 + }, + { + "epoch": 5.890531362365162, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14744 + }, + { + "epoch": 5.890930882940472, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14745 + }, + { + "epoch": 5.891330403515781, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 14746 + }, + { + "epoch": 5.8917299240910905, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14747 + }, + { + "epoch": 5.8921294446664, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14748 + }, + { + "epoch": 5.89252896524171, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 14749 + }, + { + "epoch": 5.89292848581702, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14750 + }, + { + "epoch": 5.893328006392329, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14751 + }, + { + "epoch": 5.893727526967639, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14752 + }, + { + "epoch": 5.894127047542948, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 14753 + }, + { + "epoch": 5.894526568118258, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14754 + }, + { + "epoch": 5.894926088693568, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 14755 + }, + { + "epoch": 5.895325609268878, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 14756 + }, + { + "epoch": 5.895725129844187, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14757 + }, + { + "epoch": 5.896124650419496, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3371, + "step": 14758 + }, + { + "epoch": 5.896524170994806, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14759 + }, + { + "epoch": 5.896923691570116, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14760 + }, + { + "epoch": 5.897323212145426, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 14761 + }, + { + "epoch": 5.8977227327207356, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14762 + }, + { + "epoch": 5.8981222532960444, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14763 + }, + { + "epoch": 5.898521773871354, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14764 + }, + { + "epoch": 5.898921294446664, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14765 + }, + { + "epoch": 5.899320815021974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 14766 + }, + { + "epoch": 5.899720335597284, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14767 + }, + { + "epoch": 5.9001198561725925, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 14768 + }, + { + "epoch": 5.900519376747902, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 14769 + }, + { + "epoch": 5.900918897323212, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14770 + }, + { + "epoch": 5.901318417898522, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 14771 + }, + { + "epoch": 5.901717938473832, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14772 + }, + { + "epoch": 5.9021174590491405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 14773 + }, + { + "epoch": 5.90251697962445, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14774 + }, + { + "epoch": 5.90291650019976, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 14775 + }, + { + "epoch": 5.90331602077507, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 14776 + }, + { + "epoch": 5.90371554135038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 14777 + }, + { + "epoch": 5.9041150619256895, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 14778 + }, + { + "epoch": 5.904514582500999, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14779 + }, + { + "epoch": 5.904914103076308, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 14780 + }, + { + "epoch": 5.905313623651618, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14781 + }, + { + "epoch": 5.905713144226928, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 14782 + }, + { + "epoch": 5.9061126648022375, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14783 + }, + { + "epoch": 5.906512185377547, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14784 + }, + { + "epoch": 5.906911705952856, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14785 + }, + { + "epoch": 5.907311226528166, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14786 + }, + { + "epoch": 5.907710747103476, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 14787 + }, + { + "epoch": 5.9081102676787856, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 14788 + }, + { + "epoch": 5.908509788254095, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 14789 + }, + { + "epoch": 5.908909308829404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14790 + }, + { + "epoch": 5.909308829404714, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14791 + }, + { + "epoch": 5.909708349980024, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14792 + }, + { + "epoch": 5.910107870555334, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 14793 + }, + { + "epoch": 5.910507391130643, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 14794 + }, + { + "epoch": 5.910906911705953, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14795 + }, + { + "epoch": 5.911306432281263, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3376, + "step": 14796 + }, + { + "epoch": 5.911705952856572, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 14797 + }, + { + "epoch": 5.912105473431882, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 14798 + }, + { + "epoch": 5.912504994007191, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 14799 + }, + { + "epoch": 5.912904514582501, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 14800 + }, + { + "epoch": 5.913304035157811, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 14801 + }, + { + "epoch": 5.91370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 14802 + }, + { + "epoch": 5.91410307630843, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14803 + }, + { + "epoch": 5.9145025968837395, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14804 + }, + { + "epoch": 5.914902117459049, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 14805 + }, + { + "epoch": 5.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 14806 + }, + { + "epoch": 5.915701158609668, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14807 + }, + { + "epoch": 5.916100679184978, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 14808 + }, + { + "epoch": 5.9165001997602875, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 14809 + }, + { + "epoch": 5.916899720335597, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 14810 + }, + { + "epoch": 5.917299240910907, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14811 + }, + { + "epoch": 5.917698761486217, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14812 + }, + { + "epoch": 5.918098282061527, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14813 + }, + { + "epoch": 5.918497802636836, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 14814 + }, + { + "epoch": 5.918897323212145, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 14815 + }, + { + "epoch": 5.919296843787455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14816 + }, + { + "epoch": 5.919696364362765, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 14817 + }, + { + "epoch": 5.920095884938075, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14818 + }, + { + "epoch": 5.920495405513384, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14819 + }, + { + "epoch": 5.920894926088693, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 14820 + }, + { + "epoch": 5.921294446664003, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 14821 + }, + { + "epoch": 5.921693967239313, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 14822 + }, + { + "epoch": 5.922093487814623, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 14823 + }, + { + "epoch": 5.922493008389932, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 14824 + }, + { + "epoch": 5.922892528965241, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14825 + }, + { + "epoch": 5.923292049540551, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14826 + }, + { + "epoch": 5.923691570115861, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14827 + }, + { + "epoch": 5.924091090691171, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14828 + }, + { + "epoch": 5.924490611266481, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 14829 + }, + { + "epoch": 5.9248901318417895, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 14830 + }, + { + "epoch": 5.925289652417099, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14831 + }, + { + "epoch": 5.925689172992409, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14832 + }, + { + "epoch": 5.926088693567719, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14833 + }, + { + "epoch": 5.926488214143029, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14834 + }, + { + "epoch": 5.926887734718338, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 14835 + }, + { + "epoch": 5.927287255293647, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 14836 + }, + { + "epoch": 5.927686775868957, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 14837 + }, + { + "epoch": 5.928086296444267, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 14838 + }, + { + "epoch": 5.928485817019577, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14839 + }, + { + "epoch": 5.9288853375948865, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 14840 + }, + { + "epoch": 5.929284858170195, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 14841 + }, + { + "epoch": 5.929684378745505, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 14842 + }, + { + "epoch": 5.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14843 + }, + { + "epoch": 5.930483419896125, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 14844 + }, + { + "epoch": 5.9308829404714345, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 14845 + }, + { + "epoch": 5.931282461046744, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14846 + }, + { + "epoch": 5.931681981622053, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 14847 + }, + { + "epoch": 5.932081502197363, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 14848 + }, + { + "epoch": 5.932481022772673, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 14849 + }, + { + "epoch": 5.9328805433479825, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14850 + }, + { + "epoch": 5.933280063923292, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14851 + }, + { + "epoch": 5.933679584498602, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14852 + }, + { + "epoch": 5.934079105073911, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 14853 + }, + { + "epoch": 5.934478625649221, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14854 + }, + { + "epoch": 5.934878146224531, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14855 + }, + { + "epoch": 5.93527766679984, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 14856 + }, + { + "epoch": 5.93567718737515, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 14857 + }, + { + "epoch": 5.936076707950459, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 14858 + }, + { + "epoch": 5.936476228525769, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14859 + }, + { + "epoch": 5.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 14860 + }, + { + "epoch": 5.937275269676388, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 14861 + }, + { + "epoch": 5.937674790251698, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14862 + }, + { + "epoch": 5.938074310827007, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 14863 + }, + { + "epoch": 5.938473831402317, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 14864 + }, + { + "epoch": 5.938873351977627, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14865 + }, + { + "epoch": 5.9392728725529365, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 14866 + }, + { + "epoch": 5.939672393128246, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14867 + }, + { + "epoch": 5.940071913703556, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 14868 + }, + { + "epoch": 5.940471434278866, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 14869 + }, + { + "epoch": 5.940870954854175, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 14870 + }, + { + "epoch": 5.9412704754294845, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14871 + }, + { + "epoch": 5.941669996004794, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 14872 + }, + { + "epoch": 5.942069516580104, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14873 + }, + { + "epoch": 5.942469037155414, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14874 + }, + { + "epoch": 5.942868557730723, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 14875 + }, + { + "epoch": 5.9432680783060325, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 14876 + }, + { + "epoch": 5.943667598881342, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 14877 + }, + { + "epoch": 5.944067119456652, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14878 + }, + { + "epoch": 5.944466640031962, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 14879 + }, + { + "epoch": 5.944866160607271, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.337, + "step": 14880 + }, + { + "epoch": 5.945265681182581, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14881 + }, + { + "epoch": 5.94566520175789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 14882 + }, + { + "epoch": 5.9460647223332, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14883 + }, + { + "epoch": 5.94646424290851, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14884 + }, + { + "epoch": 5.94686376348382, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 14885 + }, + { + "epoch": 5.9472632840591295, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 14886 + }, + { + "epoch": 5.947662804634438, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 14887 + }, + { + "epoch": 5.948062325209748, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14888 + }, + { + "epoch": 5.948461845785058, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14889 + }, + { + "epoch": 5.948861366360368, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14890 + }, + { + "epoch": 5.949260886935678, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 14891 + }, + { + "epoch": 5.9496604075109865, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 14892 + }, + { + "epoch": 5.950059928086296, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 14893 + }, + { + "epoch": 5.950459448661606, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 14894 + }, + { + "epoch": 5.950858969236916, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14895 + }, + { + "epoch": 5.951258489812226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.336, + "step": 14896 + }, + { + "epoch": 5.9516580103875345, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 14897 + }, + { + "epoch": 5.952057530962844, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14898 + }, + { + "epoch": 5.952457051538154, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 14899 + }, + { + "epoch": 5.952856572113464, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14900 + }, + { + "epoch": 5.953256092688774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 14901 + }, + { + "epoch": 5.953655613264083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14902 + }, + { + "epoch": 5.954055133839393, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 14903 + }, + { + "epoch": 5.954454654414702, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 14904 + }, + { + "epoch": 5.954854174990012, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3345, + "step": 14905 + }, + { + "epoch": 5.955253695565322, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 14906 + }, + { + "epoch": 5.9556532161406315, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14907 + }, + { + "epoch": 5.956052736715941, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 14908 + }, + { + "epoch": 5.95645225729125, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 14909 + }, + { + "epoch": 5.95685177786656, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 14910 + }, + { + "epoch": 5.95725129844187, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 14911 + }, + { + "epoch": 5.9576508190171795, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 14912 + }, + { + "epoch": 5.958050339592489, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14913 + }, + { + "epoch": 5.958449860167798, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 14914 + }, + { + "epoch": 5.958849380743108, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14915 + }, + { + "epoch": 5.959248901318418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 14916 + }, + { + "epoch": 5.959648421893728, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14917 + }, + { + "epoch": 5.960047942469037, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14918 + }, + { + "epoch": 5.960447463044347, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 14919 + }, + { + "epoch": 5.960846983619656, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 14920 + }, + { + "epoch": 5.961246504194966, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 14921 + }, + { + "epoch": 5.961646024770276, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14922 + }, + { + "epoch": 5.962045545345585, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 14923 + }, + { + "epoch": 5.962445065920895, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 14924 + }, + { + "epoch": 5.962844586496205, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14925 + }, + { + "epoch": 5.963244107071514, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14926 + }, + { + "epoch": 5.963643627646824, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14927 + }, + { + "epoch": 5.964043148222133, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14928 + }, + { + "epoch": 5.964442668797443, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 14929 + }, + { + "epoch": 5.964842189372753, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 14930 + }, + { + "epoch": 5.965241709948062, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14931 + }, + { + "epoch": 5.965641230523372, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 14932 + }, + { + "epoch": 5.9660407510986815, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14933 + }, + { + "epoch": 5.966440271673991, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 14934 + }, + { + "epoch": 5.966839792249301, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 14935 + }, + { + "epoch": 5.967239312824611, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 14936 + }, + { + "epoch": 5.96763883339992, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 14937 + }, + { + "epoch": 5.9680383539752295, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 14938 + }, + { + "epoch": 5.968437874550539, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 14939 + }, + { + "epoch": 5.968837395125849, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 14940 + }, + { + "epoch": 5.969236915701159, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 14941 + }, + { + "epoch": 5.969636436276469, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14942 + }, + { + "epoch": 5.970035956851778, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 14943 + }, + { + "epoch": 5.970435477427087, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 14944 + }, + { + "epoch": 5.970834998002397, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3341, + "step": 14945 + }, + { + "epoch": 5.971234518577707, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14946 + }, + { + "epoch": 5.971634039153017, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 14947 + }, + { + "epoch": 5.972033559728326, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14948 + }, + { + "epoch": 5.972433080303635, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 14949 + }, + { + "epoch": 5.972832600878945, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 14950 + }, + { + "epoch": 5.973232121454255, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 14951 + }, + { + "epoch": 5.973631642029565, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14952 + }, + { + "epoch": 5.974031162604874, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 14953 + }, + { + "epoch": 5.974430683180183, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 14954 + }, + { + "epoch": 5.974830203755493, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14955 + }, + { + "epoch": 5.975229724330803, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 14956 + }, + { + "epoch": 5.975629244906113, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 14957 + }, + { + "epoch": 5.976028765481423, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14958 + }, + { + "epoch": 5.976428286056732, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 14959 + }, + { + "epoch": 5.976827806632041, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 14960 + }, + { + "epoch": 5.977227327207351, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 14961 + }, + { + "epoch": 5.977626847782661, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 14962 + }, + { + "epoch": 5.978026368357971, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 14963 + }, + { + "epoch": 5.97842588893328, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 14964 + }, + { + "epoch": 5.978825409508589, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 14965 + }, + { + "epoch": 5.979224930083899, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14966 + }, + { + "epoch": 5.979624450659209, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 14967 + }, + { + "epoch": 5.980023971234519, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 14968 + }, + { + "epoch": 5.9804234918098285, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 14969 + }, + { + "epoch": 5.980823012385137, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 14970 + }, + { + "epoch": 5.981222532960447, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14971 + }, + { + "epoch": 5.981622053535757, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 14972 + }, + { + "epoch": 5.982021574111067, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 14973 + }, + { + "epoch": 5.9824210946863765, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 14974 + }, + { + "epoch": 5.982820615261686, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 14975 + }, + { + "epoch": 5.983220135836996, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 14976 + }, + { + "epoch": 5.983619656412305, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 14977 + }, + { + "epoch": 5.984019176987615, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 14978 + }, + { + "epoch": 5.9844186975629245, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 14979 + }, + { + "epoch": 5.984818218138234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 14980 + }, + { + "epoch": 5.985217738713544, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 14981 + }, + { + "epoch": 5.985617259288853, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 14982 + }, + { + "epoch": 5.986016779864163, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 14983 + }, + { + "epoch": 5.986416300439473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 14984 + }, + { + "epoch": 5.986815821014782, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 14985 + }, + { + "epoch": 5.987215341590092, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 14986 + }, + { + "epoch": 5.987614862165401, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 14987 + }, + { + "epoch": 5.988014382740711, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 14988 + }, + { + "epoch": 5.988413903316021, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 14989 + }, + { + "epoch": 5.98881342389133, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 14990 + }, + { + "epoch": 5.98921294446664, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 14991 + }, + { + "epoch": 5.98961246504195, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 14992 + }, + { + "epoch": 5.99001198561726, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 14993 + }, + { + "epoch": 5.990411506192569, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 14994 + }, + { + "epoch": 5.9908110267678785, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 14995 + }, + { + "epoch": 5.991210547343188, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 14996 + }, + { + "epoch": 5.991610067918498, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 14997 + }, + { + "epoch": 5.992009588493808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 14998 + }, + { + "epoch": 5.992409109069117, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 14999 + }, + { + "epoch": 5.9928086296444265, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15000 + }, + { + "epoch": 5.993208150219736, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15001 + }, + { + "epoch": 5.993607670795046, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 15002 + }, + { + "epoch": 5.994007191370356, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15003 + }, + { + "epoch": 5.994406711945665, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15004 + }, + { + "epoch": 5.9948062325209746, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15005 + }, + { + "epoch": 5.995205753096284, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3332, + "step": 15006 + }, + { + "epoch": 5.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15007 + }, + { + "epoch": 5.996004794246904, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 15008 + }, + { + "epoch": 5.996404314822214, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15009 + }, + { + "epoch": 5.996803835397523, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15010 + }, + { + "epoch": 5.997203355972832, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15011 + }, + { + "epoch": 5.997602876548142, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15012 + }, + { + "epoch": 5.998002397123452, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15013 + }, + { + "epoch": 5.998401917698762, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15014 + }, + { + "epoch": 5.9988014382740715, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15015 + }, + { + "epoch": 5.99920095884938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15016 + }, + { + "epoch": 5.99960047942469, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 15017 + }, + { + "epoch": 6.0, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15018 + }, + { + "epoch": 6.00039952057531, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15019 + }, + { + "epoch": 6.00079904115062, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15020 + }, + { + "epoch": 6.0011985617259285, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15021 + }, + { + "epoch": 6.001598082301238, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15022 + }, + { + "epoch": 6.001997602876548, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15023 + }, + { + "epoch": 6.002397123451858, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15024 + }, + { + "epoch": 6.002796644027168, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 15025 + }, + { + "epoch": 6.003196164602477, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15026 + }, + { + "epoch": 6.003595685177786, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15027 + }, + { + "epoch": 6.003995205753096, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15028 + }, + { + "epoch": 6.004394726328406, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15029 + }, + { + "epoch": 6.004794246903716, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15030 + }, + { + "epoch": 6.0051937674790254, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15031 + }, + { + "epoch": 6.005593288054335, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15032 + }, + { + "epoch": 6.005992808629644, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15033 + }, + { + "epoch": 6.006392329204954, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15034 + }, + { + "epoch": 6.006791849780264, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15035 + }, + { + "epoch": 6.0071913703555735, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3361, + "step": 15036 + }, + { + "epoch": 6.007590890930883, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15037 + }, + { + "epoch": 6.007990411506192, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15038 + }, + { + "epoch": 6.008389932081502, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15039 + }, + { + "epoch": 6.008789452656812, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15040 + }, + { + "epoch": 6.0091889732321215, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15041 + }, + { + "epoch": 6.009588493807431, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15042 + }, + { + "epoch": 6.009988014382741, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15043 + }, + { + "epoch": 6.01038753495805, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15044 + }, + { + "epoch": 6.01078705553336, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15045 + }, + { + "epoch": 6.01118657610867, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15046 + }, + { + "epoch": 6.011586096683979, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15047 + }, + { + "epoch": 6.011985617259289, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15048 + }, + { + "epoch": 6.012385137834598, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15049 + }, + { + "epoch": 6.012784658409908, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15050 + }, + { + "epoch": 6.013184178985218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15051 + }, + { + "epoch": 6.013583699560527, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15052 + }, + { + "epoch": 6.013983220135837, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15053 + }, + { + "epoch": 6.014382740711147, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15054 + }, + { + "epoch": 6.014782261286456, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15055 + }, + { + "epoch": 6.015181781861766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15056 + }, + { + "epoch": 6.0155813024370755, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15057 + }, + { + "epoch": 6.015980823012385, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 15058 + }, + { + "epoch": 6.016380343587695, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15059 + }, + { + "epoch": 6.016779864163005, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 15060 + }, + { + "epoch": 6.017179384738314, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15061 + }, + { + "epoch": 6.0175789053136235, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15062 + }, + { + "epoch": 6.017978425888933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15063 + }, + { + "epoch": 6.018377946464243, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15064 + }, + { + "epoch": 6.018777467039553, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15065 + }, + { + "epoch": 6.019176987614862, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15066 + }, + { + "epoch": 6.0195765081901715, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15067 + }, + { + "epoch": 6.019976028765481, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15068 + }, + { + "epoch": 6.020375549340791, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15069 + }, + { + "epoch": 6.020775069916101, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 15070 + }, + { + "epoch": 6.021174590491411, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15071 + }, + { + "epoch": 6.02157411106672, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15072 + }, + { + "epoch": 6.021973631642029, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15073 + }, + { + "epoch": 6.022373152217339, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15074 + }, + { + "epoch": 6.022772672792649, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 15075 + }, + { + "epoch": 6.023172193367959, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15076 + }, + { + "epoch": 6.0235717139432685, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15077 + }, + { + "epoch": 6.023971234518577, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15078 + }, + { + "epoch": 6.024370755093887, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15079 + }, + { + "epoch": 6.024770275669197, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15080 + }, + { + "epoch": 6.025169796244507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15081 + }, + { + "epoch": 6.025569316819817, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3314, + "step": 15082 + }, + { + "epoch": 6.0259688373951255, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15083 + }, + { + "epoch": 6.026368357970435, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15084 + }, + { + "epoch": 6.026767878545745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15085 + }, + { + "epoch": 6.027167399121055, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15086 + }, + { + "epoch": 6.027566919696365, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15087 + }, + { + "epoch": 6.027966440271674, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15088 + }, + { + "epoch": 6.028365960846983, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15089 + }, + { + "epoch": 6.028765481422293, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15090 + }, + { + "epoch": 6.029165001997603, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15091 + }, + { + "epoch": 6.029564522572913, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15092 + }, + { + "epoch": 6.029964043148222, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15093 + }, + { + "epoch": 6.030363563723531, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15094 + }, + { + "epoch": 6.030763084298841, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15095 + }, + { + "epoch": 6.031162604874151, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15096 + }, + { + "epoch": 6.031562125449461, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15097 + }, + { + "epoch": 6.0319616460247705, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15098 + }, + { + "epoch": 6.03236116660008, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15099 + }, + { + "epoch": 6.032760687175389, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15100 + }, + { + "epoch": 6.033160207750699, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15101 + }, + { + "epoch": 6.033559728326009, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15102 + }, + { + "epoch": 6.0339592489013185, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15103 + }, + { + "epoch": 6.034358769476628, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15104 + }, + { + "epoch": 6.034758290051938, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15105 + }, + { + "epoch": 6.035157810627247, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15106 + }, + { + "epoch": 6.035557331202557, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15107 + }, + { + "epoch": 6.035956851777867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 15108 + }, + { + "epoch": 6.036356372353176, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15109 + }, + { + "epoch": 6.036755892928486, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15110 + }, + { + "epoch": 6.037155413503795, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15111 + }, + { + "epoch": 6.037554934079105, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15112 + }, + { + "epoch": 6.037954454654415, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15113 + }, + { + "epoch": 6.038353975229724, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15114 + }, + { + "epoch": 6.038753495805034, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15115 + }, + { + "epoch": 6.039153016380344, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15116 + }, + { + "epoch": 6.039552536955653, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15117 + }, + { + "epoch": 6.039952057530963, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15118 + }, + { + "epoch": 6.040351578106272, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15119 + }, + { + "epoch": 6.040751098681582, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15120 + }, + { + "epoch": 6.041150619256892, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15121 + }, + { + "epoch": 6.041550139832202, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15122 + }, + { + "epoch": 6.041949660407511, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15123 + }, + { + "epoch": 6.0423491809828205, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 15124 + }, + { + "epoch": 6.04274870155813, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15125 + }, + { + "epoch": 6.04314822213344, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15126 + }, + { + "epoch": 6.04354774270875, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15127 + }, + { + "epoch": 6.043947263284059, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15128 + }, + { + "epoch": 6.0443467838593685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15129 + }, + { + "epoch": 6.044746304434678, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15130 + }, + { + "epoch": 6.045145825009988, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15131 + }, + { + "epoch": 6.045545345585298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15132 + }, + { + "epoch": 6.045944866160608, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15133 + }, + { + "epoch": 6.046344386735917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15134 + }, + { + "epoch": 6.046743907311226, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15135 + }, + { + "epoch": 6.047143427886536, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15136 + }, + { + "epoch": 6.047542948461846, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15137 + }, + { + "epoch": 6.047942469037156, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15138 + }, + { + "epoch": 6.048341989612465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15139 + }, + { + "epoch": 6.048741510187774, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15140 + }, + { + "epoch": 6.049141030763084, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15141 + }, + { + "epoch": 6.049540551338394, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15142 + }, + { + "epoch": 6.049940071913704, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15143 + }, + { + "epoch": 6.0503395924890135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15144 + }, + { + "epoch": 6.050739113064322, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15145 + }, + { + "epoch": 6.051138633639632, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15146 + }, + { + "epoch": 6.051538154214942, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15147 + }, + { + "epoch": 6.051937674790252, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15148 + }, + { + "epoch": 6.052337195365562, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15149 + }, + { + "epoch": 6.052736715940871, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15150 + }, + { + "epoch": 6.05313623651618, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15151 + }, + { + "epoch": 6.05353575709149, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15152 + }, + { + "epoch": 6.0539352776668, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3333, + "step": 15153 + }, + { + "epoch": 6.05433479824211, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 15154 + }, + { + "epoch": 6.054734318817419, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15155 + }, + { + "epoch": 6.055133839392728, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15156 + }, + { + "epoch": 6.055533359968038, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15157 + }, + { + "epoch": 6.055932880543348, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15158 + }, + { + "epoch": 6.056332401118658, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15159 + }, + { + "epoch": 6.0567319216939675, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15160 + }, + { + "epoch": 6.057131442269277, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15161 + }, + { + "epoch": 6.057530962844586, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15162 + }, + { + "epoch": 6.057930483419896, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15163 + }, + { + "epoch": 6.058330003995206, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15164 + }, + { + "epoch": 6.0587295245705155, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15165 + }, + { + "epoch": 6.059129045145825, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15166 + }, + { + "epoch": 6.059528565721135, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15167 + }, + { + "epoch": 6.059928086296444, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15168 + }, + { + "epoch": 6.060327606871754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15169 + }, + { + "epoch": 6.0607271274470635, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15170 + }, + { + "epoch": 6.061126648022373, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15171 + }, + { + "epoch": 6.061526168597683, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15172 + }, + { + "epoch": 6.061925689172992, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15173 + }, + { + "epoch": 6.062325209748302, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15174 + }, + { + "epoch": 6.062724730323612, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 15175 + }, + { + "epoch": 6.063124250898921, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15176 + }, + { + "epoch": 6.063523771474231, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15177 + }, + { + "epoch": 6.063923292049541, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15178 + }, + { + "epoch": 6.06432281262485, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3324, + "step": 15179 + }, + { + "epoch": 6.06472233320016, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15180 + }, + { + "epoch": 6.065121853775469, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15181 + }, + { + "epoch": 6.065521374350779, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15182 + }, + { + "epoch": 6.065920894926089, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15183 + }, + { + "epoch": 6.066320415501398, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15184 + }, + { + "epoch": 6.066719936076708, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15185 + }, + { + "epoch": 6.0671194566520175, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15186 + }, + { + "epoch": 6.067518977227327, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15187 + }, + { + "epoch": 6.067918497802637, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15188 + }, + { + "epoch": 6.068318018377947, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15189 + }, + { + "epoch": 6.068717538953256, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15190 + }, + { + "epoch": 6.0691170595285655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15191 + }, + { + "epoch": 6.069516580103875, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15192 + }, + { + "epoch": 6.069916100679185, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15193 + }, + { + "epoch": 6.070315621254495, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 15194 + }, + { + "epoch": 6.070715141829805, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15195 + }, + { + "epoch": 6.0711146624051135, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15196 + }, + { + "epoch": 6.071514182980423, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15197 + }, + { + "epoch": 6.071913703555733, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15198 + }, + { + "epoch": 6.072313224131043, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 15199 + }, + { + "epoch": 6.072712744706353, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15200 + }, + { + "epoch": 6.073112265281662, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15201 + }, + { + "epoch": 6.073511785856971, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15202 + }, + { + "epoch": 6.073911306432281, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 15203 + }, + { + "epoch": 6.074310827007591, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15204 + }, + { + "epoch": 6.074710347582901, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15205 + }, + { + "epoch": 6.0751098681582105, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15206 + }, + { + "epoch": 6.075509388733519, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15207 + }, + { + "epoch": 6.075908909308829, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15208 + }, + { + "epoch": 6.076308429884139, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15209 + }, + { + "epoch": 6.076707950459449, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15210 + }, + { + "epoch": 6.077107471034759, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15211 + }, + { + "epoch": 6.077506991610068, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15212 + }, + { + "epoch": 6.077906512185377, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15213 + }, + { + "epoch": 6.078306032760687, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15214 + }, + { + "epoch": 6.078705553335997, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15215 + }, + { + "epoch": 6.079105073911307, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15216 + }, + { + "epoch": 6.079504594486616, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15217 + }, + { + "epoch": 6.079904115061925, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15218 + }, + { + "epoch": 6.080303635637235, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15219 + }, + { + "epoch": 6.080703156212545, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 15220 + }, + { + "epoch": 6.081102676787855, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15221 + }, + { + "epoch": 6.081502197363164, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15222 + }, + { + "epoch": 6.081901717938474, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15223 + }, + { + "epoch": 6.082301238513783, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15224 + }, + { + "epoch": 6.082700759089093, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15225 + }, + { + "epoch": 6.083100279664403, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15226 + }, + { + "epoch": 6.0834998002397125, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15227 + }, + { + "epoch": 6.083899320815022, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15228 + }, + { + "epoch": 6.084298841390331, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15229 + }, + { + "epoch": 6.084698361965641, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15230 + }, + { + "epoch": 6.085097882540951, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15231 + }, + { + "epoch": 6.0854974031162605, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15232 + }, + { + "epoch": 6.08589692369157, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15233 + }, + { + "epoch": 6.08629644426688, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15234 + }, + { + "epoch": 6.086695964842189, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15235 + }, + { + "epoch": 6.087095485417499, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15236 + }, + { + "epoch": 6.087495005992809, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15237 + }, + { + "epoch": 6.087894526568118, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15238 + }, + { + "epoch": 6.088294047143428, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15239 + }, + { + "epoch": 6.088693567718738, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15240 + }, + { + "epoch": 6.089093088294047, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15241 + }, + { + "epoch": 6.089492608869357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15242 + }, + { + "epoch": 6.089892129444666, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15243 + }, + { + "epoch": 6.090291650019976, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15244 + }, + { + "epoch": 6.090691170595286, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15245 + }, + { + "epoch": 6.091090691170595, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15246 + }, + { + "epoch": 6.091490211745905, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15247 + }, + { + "epoch": 6.0918897323212144, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15248 + }, + { + "epoch": 6.092289252896524, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15249 + }, + { + "epoch": 6.092688773471834, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15250 + }, + { + "epoch": 6.093088294047144, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15251 + }, + { + "epoch": 6.093487814622453, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15252 + }, + { + "epoch": 6.0938873351977625, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15253 + }, + { + "epoch": 6.094286855773072, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15254 + }, + { + "epoch": 6.094686376348382, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15255 + }, + { + "epoch": 6.095085896923692, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15256 + }, + { + "epoch": 6.095485417499002, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15257 + }, + { + "epoch": 6.0958849380743105, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15258 + }, + { + "epoch": 6.09628445864962, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15259 + }, + { + "epoch": 6.09668397922493, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15260 + }, + { + "epoch": 6.09708349980024, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15261 + }, + { + "epoch": 6.09748302037555, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15262 + }, + { + "epoch": 6.097882540950859, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15263 + }, + { + "epoch": 6.098282061526168, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15264 + }, + { + "epoch": 6.098681582101478, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15265 + }, + { + "epoch": 6.099081102676788, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15266 + }, + { + "epoch": 6.099480623252098, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15267 + }, + { + "epoch": 6.0998801438274075, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15268 + }, + { + "epoch": 6.100279664402716, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15269 + }, + { + "epoch": 6.100679184978026, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15270 + }, + { + "epoch": 6.101078705553336, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15271 + }, + { + "epoch": 6.101478226128646, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15272 + }, + { + "epoch": 6.1018777467039556, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15273 + }, + { + "epoch": 6.1022772672792644, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15274 + }, + { + "epoch": 6.102676787854574, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15275 + }, + { + "epoch": 6.103076308429884, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15276 + }, + { + "epoch": 6.103475829005194, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15277 + }, + { + "epoch": 6.103875349580504, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15278 + }, + { + "epoch": 6.104274870155813, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15279 + }, + { + "epoch": 6.104674390731122, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15280 + }, + { + "epoch": 6.105073911306432, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15281 + }, + { + "epoch": 6.105473431881742, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15282 + }, + { + "epoch": 6.105872952457052, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15283 + }, + { + "epoch": 6.106272473032361, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15284 + }, + { + "epoch": 6.106671993607671, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15285 + }, + { + "epoch": 6.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15286 + }, + { + "epoch": 6.10747103475829, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15287 + }, + { + "epoch": 6.1078705553336, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15288 + }, + { + "epoch": 6.1082700759089095, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15289 + }, + { + "epoch": 6.108669596484219, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15290 + }, + { + "epoch": 6.109069117059528, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15291 + }, + { + "epoch": 6.109468637634838, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15292 + }, + { + "epoch": 6.109868158210148, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15293 + }, + { + "epoch": 6.1102676787854575, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15294 + }, + { + "epoch": 6.110667199360767, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15295 + }, + { + "epoch": 6.111066719936077, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15296 + }, + { + "epoch": 6.111466240511386, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15297 + }, + { + "epoch": 6.111865761086696, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15298 + }, + { + "epoch": 6.112265281662006, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15299 + }, + { + "epoch": 6.112664802237315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15300 + }, + { + "epoch": 6.113064322812625, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15301 + }, + { + "epoch": 6.113463843387935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15302 + }, + { + "epoch": 6.113863363963244, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15303 + }, + { + "epoch": 6.114262884538554, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15304 + }, + { + "epoch": 6.114662405113863, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15305 + }, + { + "epoch": 6.115061925689173, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15306 + }, + { + "epoch": 6.115461446264483, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15307 + }, + { + "epoch": 6.115860966839792, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15308 + }, + { + "epoch": 6.116260487415102, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15309 + }, + { + "epoch": 6.116660007990411, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15310 + }, + { + "epoch": 6.117059528565721, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15311 + }, + { + "epoch": 6.117459049141031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 15312 + }, + { + "epoch": 6.117858569716341, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15313 + }, + { + "epoch": 6.11825809029165, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15314 + }, + { + "epoch": 6.1186576108669595, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15315 + }, + { + "epoch": 6.119057131442269, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 15316 + }, + { + "epoch": 6.119456652017579, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15317 + }, + { + "epoch": 6.119856172592889, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15318 + }, + { + "epoch": 6.120255693168198, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15319 + }, + { + "epoch": 6.1206552137435075, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15320 + }, + { + "epoch": 6.121054734318817, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15321 + }, + { + "epoch": 6.121454254894127, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15322 + }, + { + "epoch": 6.121853775469437, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15323 + }, + { + "epoch": 6.122253296044747, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15324 + }, + { + "epoch": 6.122652816620056, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15325 + }, + { + "epoch": 6.123052337195365, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15326 + }, + { + "epoch": 6.123451857770675, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15327 + }, + { + "epoch": 6.123851378345985, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15328 + }, + { + "epoch": 6.124250898921295, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15329 + }, + { + "epoch": 6.1246504194966045, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15330 + }, + { + "epoch": 6.125049940071913, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 15331 + }, + { + "epoch": 6.125449460647223, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15332 + }, + { + "epoch": 6.125848981222533, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15333 + }, + { + "epoch": 6.126248501797843, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15334 + }, + { + "epoch": 6.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15335 + }, + { + "epoch": 6.127047542948461, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15336 + }, + { + "epoch": 6.127447063523771, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15337 + }, + { + "epoch": 6.127846584099081, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15338 + }, + { + "epoch": 6.128246104674391, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 15339 + }, + { + "epoch": 6.128645625249701, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15340 + }, + { + "epoch": 6.12904514582501, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15341 + }, + { + "epoch": 6.129444666400319, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15342 + }, + { + "epoch": 6.129844186975629, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15343 + }, + { + "epoch": 6.130243707550939, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15344 + }, + { + "epoch": 6.130643228126249, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15345 + }, + { + "epoch": 6.131042748701558, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15346 + }, + { + "epoch": 6.131442269276867, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15347 + }, + { + "epoch": 6.131841789852177, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15348 + }, + { + "epoch": 6.132241310427487, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15349 + }, + { + "epoch": 6.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15350 + }, + { + "epoch": 6.1330403515781065, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15351 + }, + { + "epoch": 6.133439872153416, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15352 + }, + { + "epoch": 6.133839392728725, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15353 + }, + { + "epoch": 6.134238913304035, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15354 + }, + { + "epoch": 6.134638433879345, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15355 + }, + { + "epoch": 6.1350379544546545, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15356 + }, + { + "epoch": 6.135437475029964, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15357 + }, + { + "epoch": 6.135836995605274, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15358 + }, + { + "epoch": 6.136236516180583, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15359 + }, + { + "epoch": 6.136636036755893, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15360 + }, + { + "epoch": 6.1370355573312025, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15361 + }, + { + "epoch": 6.137435077906512, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15362 + }, + { + "epoch": 6.137834598481822, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15363 + }, + { + "epoch": 6.138234119057131, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15364 + }, + { + "epoch": 6.138633639632441, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 15365 + }, + { + "epoch": 6.139033160207751, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15366 + }, + { + "epoch": 6.13943268078306, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15367 + }, + { + "epoch": 6.13983220135837, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15368 + }, + { + "epoch": 6.14023172193368, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15369 + }, + { + "epoch": 6.140631242508989, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15370 + }, + { + "epoch": 6.141030763084299, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15371 + }, + { + "epoch": 6.141430283659608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15372 + }, + { + "epoch": 6.141829804234918, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15373 + }, + { + "epoch": 6.142229324810228, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15374 + }, + { + "epoch": 6.142628845385538, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15375 + }, + { + "epoch": 6.143028365960847, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15376 + }, + { + "epoch": 6.1434278865361565, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15377 + }, + { + "epoch": 6.143827407111466, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15378 + }, + { + "epoch": 6.144226927686776, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15379 + }, + { + "epoch": 6.144626448262086, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15380 + }, + { + "epoch": 6.145025968837395, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15381 + }, + { + "epoch": 6.1454254894127045, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 15382 + }, + { + "epoch": 6.145825009988014, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15383 + }, + { + "epoch": 6.146224530563324, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15384 + }, + { + "epoch": 6.146624051138634, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15385 + }, + { + "epoch": 6.147023571713944, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15386 + }, + { + "epoch": 6.1474230922892525, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15387 + }, + { + "epoch": 6.147822612864562, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15388 + }, + { + "epoch": 6.148222133439872, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 15389 + }, + { + "epoch": 6.148621654015182, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15390 + }, + { + "epoch": 6.149021174590492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15391 + }, + { + "epoch": 6.1494206951658015, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15392 + }, + { + "epoch": 6.14982021574111, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15393 + }, + { + "epoch": 6.15021973631642, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 15394 + }, + { + "epoch": 6.15061925689173, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15395 + }, + { + "epoch": 6.15101877746704, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15396 + }, + { + "epoch": 6.1514182980423495, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 15397 + }, + { + "epoch": 6.151817818617658, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 15398 + }, + { + "epoch": 6.152217339192968, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15399 + }, + { + "epoch": 6.152616859768278, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15400 + }, + { + "epoch": 6.153016380343588, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15401 + }, + { + "epoch": 6.153415900918898, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15402 + }, + { + "epoch": 6.153815421494207, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15403 + }, + { + "epoch": 6.154214942069516, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2963, + "step": 15404 + }, + { + "epoch": 6.154614462644826, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 15405 + }, + { + "epoch": 6.155013983220136, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15406 + }, + { + "epoch": 6.155413503795446, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15407 + }, + { + "epoch": 6.155813024370755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15408 + }, + { + "epoch": 6.156212544946065, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15409 + }, + { + "epoch": 6.156612065521374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15410 + }, + { + "epoch": 6.157011586096684, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15411 + }, + { + "epoch": 6.157411106671994, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15412 + }, + { + "epoch": 6.157810627247303, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15413 + }, + { + "epoch": 6.158210147822613, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15414 + }, + { + "epoch": 6.158609668397922, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15415 + }, + { + "epoch": 6.159009188973232, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15416 + }, + { + "epoch": 6.159408709548542, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15417 + }, + { + "epoch": 6.1598082301238515, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15418 + }, + { + "epoch": 6.160207750699161, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15419 + }, + { + "epoch": 6.160607271274471, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.2983, + "step": 15420 + }, + { + "epoch": 6.16100679184978, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 15421 + }, + { + "epoch": 6.16140631242509, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15422 + }, + { + "epoch": 6.1618058330003995, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15423 + }, + { + "epoch": 6.162205353575709, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15424 + }, + { + "epoch": 6.162604874151019, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15425 + }, + { + "epoch": 6.163004394726328, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15426 + }, + { + "epoch": 6.163403915301638, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15427 + }, + { + "epoch": 6.163803435876948, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 15428 + }, + { + "epoch": 6.164202956452257, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15429 + }, + { + "epoch": 6.164602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15430 + }, + { + "epoch": 6.165001997602877, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15431 + }, + { + "epoch": 6.165401518178186, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15432 + }, + { + "epoch": 6.165801038753496, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15433 + }, + { + "epoch": 6.166200559328805, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15434 + }, + { + "epoch": 6.166600079904115, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15435 + }, + { + "epoch": 6.166999600479425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15436 + }, + { + "epoch": 6.167399121054734, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 15437 + }, + { + "epoch": 6.167798641630044, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15438 + }, + { + "epoch": 6.168198162205353, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15439 + }, + { + "epoch": 6.168597682780663, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15440 + }, + { + "epoch": 6.168997203355973, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15441 + }, + { + "epoch": 6.169396723931283, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15442 + }, + { + "epoch": 6.169796244506592, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15443 + }, + { + "epoch": 6.1701957650819015, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 15444 + }, + { + "epoch": 6.170595285657211, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15445 + }, + { + "epoch": 6.170994806232521, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15446 + }, + { + "epoch": 6.171394326807831, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15447 + }, + { + "epoch": 6.171793847383141, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15448 + }, + { + "epoch": 6.1721933679584495, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 15449 + }, + { + "epoch": 6.172592888533759, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15450 + }, + { + "epoch": 6.172992409109069, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 15451 + }, + { + "epoch": 6.173391929684379, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15452 + }, + { + "epoch": 6.173791450259689, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15453 + }, + { + "epoch": 6.174190970834998, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15454 + }, + { + "epoch": 6.174590491410307, + "grad_norm": 0.111328125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15455 + }, + { + "epoch": 6.174990011985617, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15456 + }, + { + "epoch": 6.175389532560927, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15457 + }, + { + "epoch": 6.175789053136237, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 15458 + }, + { + "epoch": 6.1761885737115465, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15459 + }, + { + "epoch": 6.176588094286855, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15460 + }, + { + "epoch": 6.176987614862165, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15461 + }, + { + "epoch": 6.177387135437475, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 15462 + }, + { + "epoch": 6.177786656012785, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 15463 + }, + { + "epoch": 6.1781861765880945, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 15464 + }, + { + "epoch": 6.178585697163404, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15465 + }, + { + "epoch": 6.178985217738713, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15466 + }, + { + "epoch": 6.179384738314023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15467 + }, + { + "epoch": 6.179784258889333, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15468 + }, + { + "epoch": 6.180183779464643, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15469 + }, + { + "epoch": 6.180583300039952, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15470 + }, + { + "epoch": 6.180982820615261, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15471 + }, + { + "epoch": 6.181382341190571, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15472 + }, + { + "epoch": 6.181781861765881, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15473 + }, + { + "epoch": 6.182181382341191, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15474 + }, + { + "epoch": 6.1825809029165, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15475 + }, + { + "epoch": 6.18298042349181, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 15476 + }, + { + "epoch": 6.183379944067119, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15477 + }, + { + "epoch": 6.183779464642429, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 15478 + }, + { + "epoch": 6.184178985217739, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15479 + }, + { + "epoch": 6.1845785057930485, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15480 + }, + { + "epoch": 6.184978026368358, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15481 + }, + { + "epoch": 6.185377546943668, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15482 + }, + { + "epoch": 6.185777067518977, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 15483 + }, + { + "epoch": 6.186176588094287, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 15484 + }, + { + "epoch": 6.1865761086695965, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15485 + }, + { + "epoch": 6.186975629244906, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 15486 + }, + { + "epoch": 6.187375149820216, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15487 + }, + { + "epoch": 6.187774670395525, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15488 + }, + { + "epoch": 6.188174190970835, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15489 + }, + { + "epoch": 6.1885737115461446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15490 + }, + { + "epoch": 6.188973232121454, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15491 + }, + { + "epoch": 6.189372752696764, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15492 + }, + { + "epoch": 6.189772273272074, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15493 + }, + { + "epoch": 6.190171793847383, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15494 + }, + { + "epoch": 6.190571314422693, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15495 + }, + { + "epoch": 6.190970834998002, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15496 + }, + { + "epoch": 6.191370355573312, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 15497 + }, + { + "epoch": 6.191769876148622, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15498 + }, + { + "epoch": 6.192169396723932, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15499 + }, + { + "epoch": 6.192568917299241, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15500 + }, + { + "epoch": 6.19296843787455, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15501 + }, + { + "epoch": 6.19336795844986, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15502 + }, + { + "epoch": 6.19376747902517, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15503 + }, + { + "epoch": 6.19416699960048, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15504 + }, + { + "epoch": 6.194566520175789, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15505 + }, + { + "epoch": 6.1949660407510985, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15506 + }, + { + "epoch": 6.195365561326408, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15507 + }, + { + "epoch": 6.195765081901718, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15508 + }, + { + "epoch": 6.196164602477028, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15509 + }, + { + "epoch": 6.196564123052338, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15510 + }, + { + "epoch": 6.1969636436276465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 15511 + }, + { + "epoch": 6.197363164202956, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15512 + }, + { + "epoch": 6.197762684778266, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15513 + }, + { + "epoch": 6.198162205353576, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15514 + }, + { + "epoch": 6.198561725928886, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15515 + }, + { + "epoch": 6.1989612465041946, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15516 + }, + { + "epoch": 6.199360767079504, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15517 + }, + { + "epoch": 6.199760287654814, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 15518 + }, + { + "epoch": 6.200159808230124, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15519 + }, + { + "epoch": 6.200559328805434, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15520 + }, + { + "epoch": 6.2009588493807435, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 15521 + }, + { + "epoch": 6.201358369956052, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15522 + }, + { + "epoch": 6.201757890531362, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15523 + }, + { + "epoch": 6.202157411106672, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15524 + }, + { + "epoch": 6.202556931681982, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 15525 + }, + { + "epoch": 6.2029564522572915, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15526 + }, + { + "epoch": 6.2033559728326, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15527 + }, + { + "epoch": 6.20375549340791, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 15528 + }, + { + "epoch": 6.20415501398322, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15529 + }, + { + "epoch": 6.20455453455853, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15530 + }, + { + "epoch": 6.20495405513384, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 15531 + }, + { + "epoch": 6.205353575709149, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15532 + }, + { + "epoch": 6.205753096284458, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15533 + }, + { + "epoch": 6.206152616859768, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15534 + }, + { + "epoch": 6.206552137435078, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 15535 + }, + { + "epoch": 6.206951658010388, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15536 + }, + { + "epoch": 6.207351178585697, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15537 + }, + { + "epoch": 6.207750699161007, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15538 + }, + { + "epoch": 6.208150219736316, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 15539 + }, + { + "epoch": 6.208549740311626, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15540 + }, + { + "epoch": 6.208949260886936, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 15541 + }, + { + "epoch": 6.2093487814622454, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15542 + }, + { + "epoch": 6.209748302037555, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15543 + }, + { + "epoch": 6.210147822612864, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15544 + }, + { + "epoch": 6.210547343188174, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15545 + }, + { + "epoch": 6.210946863763484, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15546 + }, + { + "epoch": 6.2113463843387935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15547 + }, + { + "epoch": 6.211745904914103, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15548 + }, + { + "epoch": 6.212145425489413, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15549 + }, + { + "epoch": 6.212544946064722, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 15550 + }, + { + "epoch": 6.212944466640032, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 15551 + }, + { + "epoch": 6.2133439872153415, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 15552 + }, + { + "epoch": 6.213743507790651, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15553 + }, + { + "epoch": 6.214143028365961, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 15554 + }, + { + "epoch": 6.214542548941271, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15555 + }, + { + "epoch": 6.21494206951658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15556 + }, + { + "epoch": 6.21534159009189, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 15557 + }, + { + "epoch": 6.215741110667199, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 15558 + }, + { + "epoch": 6.216140631242509, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 15559 + }, + { + "epoch": 6.216540151817819, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15560 + }, + { + "epoch": 6.216939672393128, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15561 + }, + { + "epoch": 6.217339192968438, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15562 + }, + { + "epoch": 6.217738713543747, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15563 + }, + { + "epoch": 6.218138234119057, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15564 + }, + { + "epoch": 6.218537754694367, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15565 + }, + { + "epoch": 6.218937275269677, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15566 + }, + { + "epoch": 6.219336795844986, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15567 + }, + { + "epoch": 6.2197363164202955, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15568 + }, + { + "epoch": 6.220135836995605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15569 + }, + { + "epoch": 6.220535357570915, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15570 + }, + { + "epoch": 6.220934878146225, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15571 + }, + { + "epoch": 6.221334398721535, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15572 + }, + { + "epoch": 6.2217339192968435, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15573 + }, + { + "epoch": 6.222133439872153, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15574 + }, + { + "epoch": 6.222532960447463, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 15575 + }, + { + "epoch": 6.222932481022773, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15576 + }, + { + "epoch": 6.223332001598083, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15577 + }, + { + "epoch": 6.2237315221733915, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15578 + }, + { + "epoch": 6.224131042748701, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15579 + }, + { + "epoch": 6.224530563324011, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15580 + }, + { + "epoch": 6.224930083899321, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15581 + }, + { + "epoch": 6.225329604474631, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15582 + }, + { + "epoch": 6.2257291250499405, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 15583 + }, + { + "epoch": 6.226128645625249, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15584 + }, + { + "epoch": 6.226528166200559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15585 + }, + { + "epoch": 6.226927686775869, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15586 + }, + { + "epoch": 6.227327207351179, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15587 + }, + { + "epoch": 6.2277267279264885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15588 + }, + { + "epoch": 6.228126248501797, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15589 + }, + { + "epoch": 6.228525769077107, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15590 + }, + { + "epoch": 6.228925289652417, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 15591 + }, + { + "epoch": 6.229324810227727, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 15592 + }, + { + "epoch": 6.229724330803037, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15593 + }, + { + "epoch": 6.230123851378346, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 15594 + }, + { + "epoch": 6.230523371953655, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 15595 + }, + { + "epoch": 6.230922892528965, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15596 + }, + { + "epoch": 6.231322413104275, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15597 + }, + { + "epoch": 6.231721933679585, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15598 + }, + { + "epoch": 6.232121454254894, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15599 + }, + { + "epoch": 6.232520974830204, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15600 + }, + { + "epoch": 6.232920495405513, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15601 + }, + { + "epoch": 6.233320015980823, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15602 + }, + { + "epoch": 6.233719536556133, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15603 + }, + { + "epoch": 6.234119057131442, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 15604 + }, + { + "epoch": 6.234518577706752, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 15605 + }, + { + "epoch": 6.234918098282061, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 15606 + }, + { + "epoch": 6.235317618857371, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 15607 + }, + { + "epoch": 6.235717139432681, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15608 + }, + { + "epoch": 6.2361166600079905, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15609 + }, + { + "epoch": 6.2365161805833, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15610 + }, + { + "epoch": 6.23691570115861, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15611 + }, + { + "epoch": 6.237315221733919, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15612 + }, + { + "epoch": 6.237714742309229, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15613 + }, + { + "epoch": 6.2381142628845385, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 15614 + }, + { + "epoch": 6.238513783459848, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 15615 + }, + { + "epoch": 6.238913304035158, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15616 + }, + { + "epoch": 6.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15617 + }, + { + "epoch": 6.239712345185777, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15618 + }, + { + "epoch": 6.240111865761087, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15619 + }, + { + "epoch": 6.240511386336396, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15620 + }, + { + "epoch": 6.240910906911706, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15621 + }, + { + "epoch": 6.241310427487016, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15622 + }, + { + "epoch": 6.241709948062325, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15623 + }, + { + "epoch": 6.242109468637635, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15624 + }, + { + "epoch": 6.242508989212944, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15625 + }, + { + "epoch": 6.242908509788254, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15626 + }, + { + "epoch": 6.243308030363564, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15627 + }, + { + "epoch": 6.243707550938874, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15628 + }, + { + "epoch": 6.244107071514183, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 15629 + }, + { + "epoch": 6.244506592089492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15630 + }, + { + "epoch": 6.244906112664802, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15631 + }, + { + "epoch": 6.245305633240112, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15632 + }, + { + "epoch": 6.245705153815422, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 15633 + }, + { + "epoch": 6.246104674390731, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 15634 + }, + { + "epoch": 6.2465041949660405, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 15635 + }, + { + "epoch": 6.24690371554135, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15636 + }, + { + "epoch": 6.24730323611666, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15637 + }, + { + "epoch": 6.24770275669197, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15638 + }, + { + "epoch": 6.24810227726728, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 15639 + }, + { + "epoch": 6.2485017978425885, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15640 + }, + { + "epoch": 6.248901318417898, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15641 + }, + { + "epoch": 6.249300838993208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15642 + }, + { + "epoch": 6.249700359568518, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15643 + }, + { + "epoch": 6.250099880143828, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15644 + }, + { + "epoch": 6.2504994007191375, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15645 + }, + { + "epoch": 6.250898921294446, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15646 + }, + { + "epoch": 6.251298441869756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15647 + }, + { + "epoch": 6.251697962445066, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15648 + }, + { + "epoch": 6.252097483020376, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 15649 + }, + { + "epoch": 6.2524970035956855, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 15650 + }, + { + "epoch": 6.252896524170994, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15651 + }, + { + "epoch": 6.253296044746304, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15652 + }, + { + "epoch": 6.253695565321614, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15653 + }, + { + "epoch": 6.254095085896924, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 15654 + }, + { + "epoch": 6.2544946064722335, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15655 + }, + { + "epoch": 6.254894127047543, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15656 + }, + { + "epoch": 6.255293647622852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15657 + }, + { + "epoch": 6.255693168198162, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15658 + }, + { + "epoch": 6.256092688773472, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15659 + }, + { + "epoch": 6.256492209348782, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 15660 + }, + { + "epoch": 6.256891729924091, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 15661 + }, + { + "epoch": 6.257291250499401, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 15662 + }, + { + "epoch": 6.25769077107471, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15663 + }, + { + "epoch": 6.25809029165002, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15664 + }, + { + "epoch": 6.25848981222533, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15665 + }, + { + "epoch": 6.258889332800639, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15666 + }, + { + "epoch": 6.259288853375949, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 15667 + }, + { + "epoch": 6.259688373951258, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15668 + }, + { + "epoch": 6.260087894526568, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15669 + }, + { + "epoch": 6.260487415101878, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15670 + }, + { + "epoch": 6.2608869356771875, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 15671 + }, + { + "epoch": 6.261286456252497, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15672 + }, + { + "epoch": 6.261685976827807, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15673 + }, + { + "epoch": 6.262085497403116, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15674 + }, + { + "epoch": 6.262485017978426, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 15675 + }, + { + "epoch": 6.2628845385537355, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15676 + }, + { + "epoch": 6.263284059129045, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15677 + }, + { + "epoch": 6.263683579704355, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.329, + "step": 15678 + }, + { + "epoch": 6.264083100279665, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 15679 + }, + { + "epoch": 6.264482620854974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 15680 + }, + { + "epoch": 6.2648821414302835, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 15681 + }, + { + "epoch": 6.265281662005593, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15682 + }, + { + "epoch": 6.265681182580903, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15683 + }, + { + "epoch": 6.266080703156213, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15684 + }, + { + "epoch": 6.266480223731522, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15685 + }, + { + "epoch": 6.266879744306832, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15686 + }, + { + "epoch": 6.267279264882141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15687 + }, + { + "epoch": 6.267678785457451, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15688 + }, + { + "epoch": 6.268078306032761, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15689 + }, + { + "epoch": 6.26847782660807, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15690 + }, + { + "epoch": 6.26887734718338, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 15691 + }, + { + "epoch": 6.269276867758689, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15692 + }, + { + "epoch": 6.269676388333999, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15693 + }, + { + "epoch": 6.270075908909309, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15694 + }, + { + "epoch": 6.270475429484619, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15695 + }, + { + "epoch": 6.270874950059928, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15696 + }, + { + "epoch": 6.2712744706352375, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15697 + }, + { + "epoch": 6.271673991210547, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15698 + }, + { + "epoch": 6.272073511785857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15699 + }, + { + "epoch": 6.272473032361167, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 15700 + }, + { + "epoch": 6.272872552936477, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 15701 + }, + { + "epoch": 6.2732720735117855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 15702 + }, + { + "epoch": 6.273671594087095, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 15703 + }, + { + "epoch": 6.274071114662405, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 15704 + }, + { + "epoch": 6.274470635237715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15705 + }, + { + "epoch": 6.274870155813025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 15706 + }, + { + "epoch": 6.2752696763883336, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 15707 + }, + { + "epoch": 6.275669196963643, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15708 + }, + { + "epoch": 6.276068717538953, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15709 + }, + { + "epoch": 6.276468238114263, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 15710 + }, + { + "epoch": 6.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15711 + }, + { + "epoch": 6.2772672792648825, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15712 + }, + { + "epoch": 6.277666799840191, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 15713 + }, + { + "epoch": 6.278066320415501, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15714 + }, + { + "epoch": 6.278465840990811, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15715 + }, + { + "epoch": 6.278865361566121, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15716 + }, + { + "epoch": 6.2792648821414305, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15717 + }, + { + "epoch": 6.27966440271674, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15718 + }, + { + "epoch": 6.280063923292049, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15719 + }, + { + "epoch": 6.280463443867359, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15720 + }, + { + "epoch": 6.280862964442669, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 15721 + }, + { + "epoch": 6.281262485017979, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 15722 + }, + { + "epoch": 6.281662005593288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15723 + }, + { + "epoch": 6.282061526168597, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 15724 + }, + { + "epoch": 6.282461046743907, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15725 + }, + { + "epoch": 6.282860567319217, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 15726 + }, + { + "epoch": 6.283260087894527, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15727 + }, + { + "epoch": 6.283659608469836, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15728 + }, + { + "epoch": 6.284059129045146, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15729 + }, + { + "epoch": 6.284458649620455, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15730 + }, + { + "epoch": 6.284858170195765, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 15731 + }, + { + "epoch": 6.285257690771075, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 15732 + }, + { + "epoch": 6.285657211346384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15733 + }, + { + "epoch": 6.286056731921694, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15734 + }, + { + "epoch": 6.286456252497004, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15735 + }, + { + "epoch": 6.286855773072313, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 15736 + }, + { + "epoch": 6.287255293647623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15737 + }, + { + "epoch": 6.2876548142229325, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 15738 + }, + { + "epoch": 6.288054334798242, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 15739 + }, + { + "epoch": 6.288453855373552, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 15740 + }, + { + "epoch": 6.288853375948861, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 15741 + }, + { + "epoch": 6.289252896524171, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 15742 + }, + { + "epoch": 6.2896524170994805, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 15743 + }, + { + "epoch": 6.29005193767479, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 15744 + }, + { + "epoch": 6.2904514582501, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 15745 + }, + { + "epoch": 6.29085097882541, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 15746 + }, + { + "epoch": 6.291250499400719, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15747 + }, + { + "epoch": 6.291650019976029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15748 + }, + { + "epoch": 6.292049540551338, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 15749 + }, + { + "epoch": 6.292449061126648, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15750 + }, + { + "epoch": 6.292848581701958, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 15751 + }, + { + "epoch": 6.293248102277268, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 15752 + }, + { + "epoch": 6.293647622852577, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15753 + }, + { + "epoch": 6.294047143427886, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15754 + }, + { + "epoch": 6.294446664003196, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15755 + }, + { + "epoch": 6.294846184578506, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 15756 + }, + { + "epoch": 6.295245705153816, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 15757 + }, + { + "epoch": 6.295645225729125, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 15758 + }, + { + "epoch": 6.2960447463044344, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15759 + }, + { + "epoch": 6.296444266879744, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15760 + }, + { + "epoch": 6.296843787455054, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15761 + }, + { + "epoch": 6.297243308030364, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15762 + }, + { + "epoch": 6.297642828605674, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15763 + }, + { + "epoch": 6.2980423491809825, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 15764 + }, + { + "epoch": 6.298441869756292, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15765 + }, + { + "epoch": 6.298841390331602, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15766 + }, + { + "epoch": 6.299240910906912, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15767 + }, + { + "epoch": 6.299640431482222, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 15768 + }, + { + "epoch": 6.300039952057531, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15769 + }, + { + "epoch": 6.30043947263284, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15770 + }, + { + "epoch": 6.30083899320815, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 15771 + }, + { + "epoch": 6.30123851378346, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15772 + }, + { + "epoch": 6.30163803435877, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 15773 + }, + { + "epoch": 6.3020375549340795, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15774 + }, + { + "epoch": 6.302437075509388, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15775 + }, + { + "epoch": 6.302836596084698, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15776 + }, + { + "epoch": 6.303236116660008, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15777 + }, + { + "epoch": 6.303635637235318, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 15778 + }, + { + "epoch": 6.3040351578106275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15779 + }, + { + "epoch": 6.304434678385936, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 15780 + }, + { + "epoch": 6.304834198961246, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15781 + }, + { + "epoch": 6.305233719536556, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15782 + }, + { + "epoch": 6.305633240111866, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15783 + }, + { + "epoch": 6.3060327606871756, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15784 + }, + { + "epoch": 6.306432281262485, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 15785 + }, + { + "epoch": 6.306831801837794, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15786 + }, + { + "epoch": 6.307231322413104, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15787 + }, + { + "epoch": 6.307630842988414, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15788 + }, + { + "epoch": 6.308030363563724, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15789 + }, + { + "epoch": 6.308429884139033, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 15790 + }, + { + "epoch": 6.308829404714343, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15791 + }, + { + "epoch": 6.309228925289652, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15792 + }, + { + "epoch": 6.309628445864962, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15793 + }, + { + "epoch": 6.310027966440272, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15794 + }, + { + "epoch": 6.310427487015581, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15795 + }, + { + "epoch": 6.310827007590891, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15796 + }, + { + "epoch": 6.3112265281662, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15797 + }, + { + "epoch": 6.31162604874151, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 15798 + }, + { + "epoch": 6.31202556931682, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15799 + }, + { + "epoch": 6.3124250898921295, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 15800 + }, + { + "epoch": 6.312824610467439, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 15801 + }, + { + "epoch": 6.313224131042749, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 15802 + }, + { + "epoch": 6.313623651618058, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 15803 + }, + { + "epoch": 6.314023172193368, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15804 + }, + { + "epoch": 6.3144226927686775, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 15805 + }, + { + "epoch": 6.314822213343987, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15806 + }, + { + "epoch": 6.315221733919297, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 15807 + }, + { + "epoch": 6.315621254494607, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15808 + }, + { + "epoch": 6.316020775069916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15809 + }, + { + "epoch": 6.316420295645226, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 15810 + }, + { + "epoch": 6.316819816220535, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15811 + }, + { + "epoch": 6.317219336795845, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 15812 + }, + { + "epoch": 6.317618857371155, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 15813 + }, + { + "epoch": 6.318018377946464, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15814 + }, + { + "epoch": 6.318417898521774, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 15815 + }, + { + "epoch": 6.318817419097083, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 15816 + }, + { + "epoch": 6.319216939672393, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 15817 + }, + { + "epoch": 6.319616460247703, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15818 + }, + { + "epoch": 6.320015980823013, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 15819 + }, + { + "epoch": 6.320415501398322, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15820 + }, + { + "epoch": 6.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 15821 + }, + { + "epoch": 6.321214542548941, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15822 + }, + { + "epoch": 6.321614063124251, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 15823 + }, + { + "epoch": 6.322013583699561, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15824 + }, + { + "epoch": 6.322413104274871, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 15825 + }, + { + "epoch": 6.3228126248501795, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15826 + }, + { + "epoch": 6.323212145425489, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 15827 + }, + { + "epoch": 6.323611666000799, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 15828 + }, + { + "epoch": 6.324011186576109, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 15829 + }, + { + "epoch": 6.324410707151419, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 15830 + }, + { + "epoch": 6.3248102277267275, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 15831 + }, + { + "epoch": 6.325209748302037, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 15832 + }, + { + "epoch": 6.325609268877347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15833 + }, + { + "epoch": 6.326008789452657, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15834 + }, + { + "epoch": 6.326408310027967, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15835 + }, + { + "epoch": 6.3268078306032765, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 15836 + }, + { + "epoch": 6.327207351178585, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 15837 + }, + { + "epoch": 6.327606871753895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 15838 + }, + { + "epoch": 6.328006392329205, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 15839 + }, + { + "epoch": 6.328405912904515, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 15840 + }, + { + "epoch": 6.3288054334798245, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 15841 + }, + { + "epoch": 6.329204954055134, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15842 + }, + { + "epoch": 6.329604474630443, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 15843 + }, + { + "epoch": 6.330003995205753, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 15844 + }, + { + "epoch": 6.330403515781063, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 15845 + }, + { + "epoch": 6.3308030363563725, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 15846 + }, + { + "epoch": 6.331202556931682, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 15847 + }, + { + "epoch": 6.331602077506991, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15848 + }, + { + "epoch": 6.332001598082301, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 15849 + }, + { + "epoch": 6.332401118657611, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15850 + }, + { + "epoch": 6.332800639232921, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 15851 + }, + { + "epoch": 6.33320015980823, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 15852 + }, + { + "epoch": 6.33359968038354, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15853 + }, + { + "epoch": 6.333999200958849, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15854 + }, + { + "epoch": 6.334398721534159, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15855 + }, + { + "epoch": 6.334798242109469, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15856 + }, + { + "epoch": 6.335197762684778, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15857 + }, + { + "epoch": 6.335597283260088, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15858 + }, + { + "epoch": 6.335996803835398, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 15859 + }, + { + "epoch": 6.336396324410707, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 15860 + }, + { + "epoch": 6.336795844986017, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 15861 + }, + { + "epoch": 6.3371953655613265, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 15862 + }, + { + "epoch": 6.337594886136636, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 15863 + }, + { + "epoch": 6.337994406711946, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15864 + }, + { + "epoch": 6.338393927287255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15865 + }, + { + "epoch": 6.338793447862565, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15866 + }, + { + "epoch": 6.3391929684378745, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15867 + }, + { + "epoch": 6.339592489013184, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 15868 + }, + { + "epoch": 6.339992009588494, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 15869 + }, + { + "epoch": 6.340391530163803, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15870 + }, + { + "epoch": 6.340791050739113, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 15871 + }, + { + "epoch": 6.3411905713144225, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15872 + }, + { + "epoch": 6.341590091889732, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 15873 + }, + { + "epoch": 6.341989612465042, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15874 + }, + { + "epoch": 6.342389133040352, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15875 + }, + { + "epoch": 6.342788653615661, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 15876 + }, + { + "epoch": 6.343188174190971, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 15877 + }, + { + "epoch": 6.34358769476628, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 15878 + }, + { + "epoch": 6.34398721534159, + "grad_norm": 0.1123046875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15879 + }, + { + "epoch": 6.3443867359169, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 15880 + }, + { + "epoch": 6.34478625649221, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 15881 + }, + { + "epoch": 6.345185777067519, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15882 + }, + { + "epoch": 6.345585297642828, + "grad_norm": 0.1396484375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15883 + }, + { + "epoch": 6.345984818218138, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15884 + }, + { + "epoch": 6.346384338793448, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15885 + }, + { + "epoch": 6.346783859368758, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 15886 + }, + { + "epoch": 6.347183379944067, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 15887 + }, + { + "epoch": 6.3475829005193765, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15888 + }, + { + "epoch": 6.347982421094686, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15889 + }, + { + "epoch": 6.348381941669996, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 15890 + }, + { + "epoch": 6.348781462245306, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 15891 + }, + { + "epoch": 6.349180982820616, + "grad_norm": 0.10107421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 15892 + }, + { + "epoch": 6.3495805033959245, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15893 + }, + { + "epoch": 6.349980023971234, + "grad_norm": 0.173828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15894 + }, + { + "epoch": 6.350379544546544, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 15895 + }, + { + "epoch": 6.350779065121854, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 15896 + }, + { + "epoch": 6.351178585697164, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 15897 + }, + { + "epoch": 6.351578106272473, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15898 + }, + { + "epoch": 6.351977626847782, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15899 + }, + { + "epoch": 6.352377147423092, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 15900 + }, + { + "epoch": 6.352776667998402, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 15901 + }, + { + "epoch": 6.353176188573712, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 15902 + }, + { + "epoch": 6.3535757091490215, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 15903 + }, + { + "epoch": 6.35397522972433, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 15904 + }, + { + "epoch": 6.35437475029964, + "grad_norm": 0.1611328125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 15905 + }, + { + "epoch": 6.35477427087495, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 15906 + }, + { + "epoch": 6.35517379145026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 15907 + }, + { + "epoch": 6.3555733120255695, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 15908 + }, + { + "epoch": 6.355972832600879, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 15909 + }, + { + "epoch": 6.356372353176188, + "grad_norm": 0.1953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15910 + }, + { + "epoch": 6.356771873751498, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 15911 + }, + { + "epoch": 6.357171394326808, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 15912 + }, + { + "epoch": 6.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 15913 + }, + { + "epoch": 6.357970435477427, + "grad_norm": 0.162109375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 15914 + }, + { + "epoch": 6.358369956052737, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 15915 + }, + { + "epoch": 6.358769476628046, + "grad_norm": 0.11767578125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 15916 + }, + { + "epoch": 6.359168997203356, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15917 + }, + { + "epoch": 6.359568517778666, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 15918 + }, + { + "epoch": 6.359968038353975, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15919 + }, + { + "epoch": 6.360367558929285, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 15920 + }, + { + "epoch": 6.360767079504594, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 15921 + }, + { + "epoch": 6.361166600079904, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15922 + }, + { + "epoch": 6.361566120655214, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15923 + }, + { + "epoch": 6.361965641230523, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 15924 + }, + { + "epoch": 6.362365161805833, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 15925 + }, + { + "epoch": 6.362764682381143, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15926 + }, + { + "epoch": 6.363164202956452, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 15927 + }, + { + "epoch": 6.363563723531762, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 15928 + }, + { + "epoch": 6.3639632441070715, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 15929 + }, + { + "epoch": 6.364362764682381, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 15930 + }, + { + "epoch": 6.364762285257691, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 15931 + }, + { + "epoch": 6.365161805833001, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15932 + }, + { + "epoch": 6.36556132640831, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 15933 + }, + { + "epoch": 6.3659608469836195, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 15934 + }, + { + "epoch": 6.366360367558929, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 15935 + }, + { + "epoch": 6.366759888134239, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.297, + "step": 15936 + }, + { + "epoch": 6.367159408709549, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15937 + }, + { + "epoch": 6.367558929284858, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 15938 + }, + { + "epoch": 6.367958449860168, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 15939 + }, + { + "epoch": 6.368357970435477, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 15940 + }, + { + "epoch": 6.368757491010787, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15941 + }, + { + "epoch": 6.369157011586097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 15942 + }, + { + "epoch": 6.369556532161407, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 15943 + }, + { + "epoch": 6.369956052736716, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 15944 + }, + { + "epoch": 6.370355573312025, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 15945 + }, + { + "epoch": 6.370755093887335, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 15946 + }, + { + "epoch": 6.371154614462645, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15947 + }, + { + "epoch": 6.371554135037955, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 15948 + }, + { + "epoch": 6.3719536556132645, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15949 + }, + { + "epoch": 6.372353176188573, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 15950 + }, + { + "epoch": 6.372752696763883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15951 + }, + { + "epoch": 6.373152217339193, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15952 + }, + { + "epoch": 6.373551737914503, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3362, + "step": 15953 + }, + { + "epoch": 6.373951258489813, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 15954 + }, + { + "epoch": 6.3743507790651215, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 15955 + }, + { + "epoch": 6.374750299640431, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 15956 + }, + { + "epoch": 6.375149820215741, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 15957 + }, + { + "epoch": 6.375549340791051, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15958 + }, + { + "epoch": 6.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 15959 + }, + { + "epoch": 6.3763483819416695, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 15960 + }, + { + "epoch": 6.376747902516979, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 15961 + }, + { + "epoch": 6.377147423092289, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 15962 + }, + { + "epoch": 6.377546943667599, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 15963 + }, + { + "epoch": 6.377946464242909, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15964 + }, + { + "epoch": 6.3783459848182185, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15965 + }, + { + "epoch": 6.378745505393527, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 15966 + }, + { + "epoch": 6.379145025968837, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 15967 + }, + { + "epoch": 6.379544546544147, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 15968 + }, + { + "epoch": 6.379944067119457, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 15969 + }, + { + "epoch": 6.3803435876947665, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 15970 + }, + { + "epoch": 6.380743108270076, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 15971 + }, + { + "epoch": 6.381142628845385, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 15972 + }, + { + "epoch": 6.381542149420695, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 15973 + }, + { + "epoch": 6.381941669996005, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15974 + }, + { + "epoch": 6.3823411905713145, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 15975 + }, + { + "epoch": 6.382740711146624, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 15976 + }, + { + "epoch": 6.383140231721933, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 15977 + }, + { + "epoch": 6.383539752297243, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 15978 + }, + { + "epoch": 6.383939272872553, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 15979 + }, + { + "epoch": 6.384338793447863, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 15980 + }, + { + "epoch": 6.384738314023172, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 15981 + }, + { + "epoch": 6.385137834598482, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 15982 + }, + { + "epoch": 6.385537355173791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 15983 + }, + { + "epoch": 6.385936875749101, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 15984 + }, + { + "epoch": 6.386336396324411, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 15985 + }, + { + "epoch": 6.38673591689972, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 15986 + }, + { + "epoch": 6.38713543747503, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 15987 + }, + { + "epoch": 6.38753495805034, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 15988 + }, + { + "epoch": 6.387934478625649, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 15989 + }, + { + "epoch": 6.388333999200959, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 15990 + }, + { + "epoch": 6.3887335197762685, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 15991 + }, + { + "epoch": 6.389133040351578, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 15992 + }, + { + "epoch": 6.389532560926888, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 15993 + }, + { + "epoch": 6.389932081502197, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 15994 + }, + { + "epoch": 6.390331602077507, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 15995 + }, + { + "epoch": 6.3907311226528165, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 15996 + }, + { + "epoch": 6.391130643228126, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 15997 + }, + { + "epoch": 6.391530163803436, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 15998 + }, + { + "epoch": 6.391929684378746, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 15999 + }, + { + "epoch": 6.392329204954055, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16000 + }, + { + "epoch": 6.3927287255293646, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16001 + }, + { + "epoch": 6.393128246104674, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 16002 + }, + { + "epoch": 6.393527766679984, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16003 + }, + { + "epoch": 6.393927287255294, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16004 + }, + { + "epoch": 6.394326807830604, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16005 + }, + { + "epoch": 6.394726328405913, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16006 + }, + { + "epoch": 6.395125848981222, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16007 + }, + { + "epoch": 6.395525369556532, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16008 + }, + { + "epoch": 6.395924890131842, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16009 + }, + { + "epoch": 6.396324410707152, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 16010 + }, + { + "epoch": 6.396723931282461, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16011 + }, + { + "epoch": 6.39712345185777, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16012 + }, + { + "epoch": 6.39752297243308, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16013 + }, + { + "epoch": 6.39792249300839, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16014 + }, + { + "epoch": 6.3983220135837, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16015 + }, + { + "epoch": 6.39872153415901, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16016 + }, + { + "epoch": 6.3991210547343185, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16017 + }, + { + "epoch": 6.399520575309628, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16018 + }, + { + "epoch": 6.399920095884938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16019 + }, + { + "epoch": 6.400319616460248, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16020 + }, + { + "epoch": 6.400719137035558, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16021 + }, + { + "epoch": 6.401118657610867, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16022 + }, + { + "epoch": 6.401518178186176, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16023 + }, + { + "epoch": 6.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16024 + }, + { + "epoch": 6.402317219336796, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16025 + }, + { + "epoch": 6.402716739912106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16026 + }, + { + "epoch": 6.4031162604874154, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16027 + }, + { + "epoch": 6.403515781062724, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16028 + }, + { + "epoch": 6.403915301638034, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16029 + }, + { + "epoch": 6.404314822213344, + "grad_norm": 0.10986328125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16030 + }, + { + "epoch": 6.404714342788654, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16031 + }, + { + "epoch": 6.4051138633639635, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16032 + }, + { + "epoch": 6.405513383939273, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16033 + }, + { + "epoch": 6.405912904514582, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16034 + }, + { + "epoch": 6.406312425089892, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16035 + }, + { + "epoch": 6.406711945665202, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16036 + }, + { + "epoch": 6.4071114662405115, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16037 + }, + { + "epoch": 6.407510986815821, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16038 + }, + { + "epoch": 6.407910507391131, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16039 + }, + { + "epoch": 6.40831002796644, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16040 + }, + { + "epoch": 6.40870954854175, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16041 + }, + { + "epoch": 6.40910906911706, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16042 + }, + { + "epoch": 6.409508589692369, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16043 + }, + { + "epoch": 6.409908110267679, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16044 + }, + { + "epoch": 6.410307630842988, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16045 + }, + { + "epoch": 6.410707151418298, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16046 + }, + { + "epoch": 6.411106671993608, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16047 + }, + { + "epoch": 6.411506192568917, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16048 + }, + { + "epoch": 6.411905713144227, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16049 + }, + { + "epoch": 6.412305233719536, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16050 + }, + { + "epoch": 6.412704754294846, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16051 + }, + { + "epoch": 6.413104274870156, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16052 + }, + { + "epoch": 6.4135037954454654, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16053 + }, + { + "epoch": 6.413903316020775, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16054 + }, + { + "epoch": 6.414302836596085, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16055 + }, + { + "epoch": 6.414702357171394, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16056 + }, + { + "epoch": 6.415101877746704, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16057 + }, + { + "epoch": 6.4155013983220135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16058 + }, + { + "epoch": 6.415900918897323, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16059 + }, + { + "epoch": 6.416300439472633, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16060 + }, + { + "epoch": 6.416699960047943, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16061 + }, + { + "epoch": 6.417099480623252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 16062 + }, + { + "epoch": 6.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16063 + }, + { + "epoch": 6.417898521773871, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 16064 + }, + { + "epoch": 6.418298042349181, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16065 + }, + { + "epoch": 6.418697562924491, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16066 + }, + { + "epoch": 6.4190970834998, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16067 + }, + { + "epoch": 6.41949660407511, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16068 + }, + { + "epoch": 6.419896124650419, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16069 + }, + { + "epoch": 6.420295645225729, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16070 + }, + { + "epoch": 6.420695165801039, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16071 + }, + { + "epoch": 6.421094686376349, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16072 + }, + { + "epoch": 6.421494206951658, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16073 + }, + { + "epoch": 6.421893727526967, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16074 + }, + { + "epoch": 6.422293248102277, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16075 + }, + { + "epoch": 6.422692768677587, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16076 + }, + { + "epoch": 6.423092289252897, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16077 + }, + { + "epoch": 6.4234918098282066, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16078 + }, + { + "epoch": 6.4238913304035155, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16079 + }, + { + "epoch": 6.424290850978825, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16080 + }, + { + "epoch": 6.424690371554135, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16081 + }, + { + "epoch": 6.425089892129445, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16082 + }, + { + "epoch": 6.425489412704755, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16083 + }, + { + "epoch": 6.4258889332800635, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16084 + }, + { + "epoch": 6.426288453855373, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16085 + }, + { + "epoch": 6.426687974430683, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16086 + }, + { + "epoch": 6.427087495005993, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16087 + }, + { + "epoch": 6.427487015581303, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16088 + }, + { + "epoch": 6.427886536156612, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16089 + }, + { + "epoch": 6.428286056731921, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16090 + }, + { + "epoch": 6.428685577307231, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16091 + }, + { + "epoch": 6.429085097882541, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16092 + }, + { + "epoch": 6.429484618457851, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16093 + }, + { + "epoch": 6.4298841390331605, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16094 + }, + { + "epoch": 6.43028365960847, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16095 + }, + { + "epoch": 6.430683180183779, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16096 + }, + { + "epoch": 6.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16097 + }, + { + "epoch": 6.431482221334399, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16098 + }, + { + "epoch": 6.4318817419097085, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16099 + }, + { + "epoch": 6.432281262485018, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16100 + }, + { + "epoch": 6.432680783060327, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16101 + }, + { + "epoch": 6.433080303635637, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16102 + }, + { + "epoch": 6.433479824210947, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 16103 + }, + { + "epoch": 6.433879344786257, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16104 + }, + { + "epoch": 6.434278865361566, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16105 + }, + { + "epoch": 6.434678385936876, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16106 + }, + { + "epoch": 6.435077906512185, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16107 + }, + { + "epoch": 6.435477427087495, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16108 + }, + { + "epoch": 6.435876947662805, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16109 + }, + { + "epoch": 6.436276468238114, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16110 + }, + { + "epoch": 6.436675988813424, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16111 + }, + { + "epoch": 6.437075509388734, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16112 + }, + { + "epoch": 6.437475029964043, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16113 + }, + { + "epoch": 6.437874550539353, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16114 + }, + { + "epoch": 6.438274071114662, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3339, + "step": 16115 + }, + { + "epoch": 6.438673591689972, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16116 + }, + { + "epoch": 6.439073112265282, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 16117 + }, + { + "epoch": 6.439472632840591, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16118 + }, + { + "epoch": 6.439872153415901, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16119 + }, + { + "epoch": 6.4402716739912105, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16120 + }, + { + "epoch": 6.44067119456652, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16121 + }, + { + "epoch": 6.44107071514183, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16122 + }, + { + "epoch": 6.44147023571714, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16123 + }, + { + "epoch": 6.441869756292449, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16124 + }, + { + "epoch": 6.4422692768677585, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16125 + }, + { + "epoch": 6.442668797443068, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16126 + }, + { + "epoch": 6.443068318018378, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16127 + }, + { + "epoch": 6.443467838593688, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16128 + }, + { + "epoch": 6.443867359168998, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 16129 + }, + { + "epoch": 6.444266879744307, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16130 + }, + { + "epoch": 6.444666400319616, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16131 + }, + { + "epoch": 6.445065920894926, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16132 + }, + { + "epoch": 6.445465441470236, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16133 + }, + { + "epoch": 6.445864962045546, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16134 + }, + { + "epoch": 6.446264482620855, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 16135 + }, + { + "epoch": 6.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16136 + }, + { + "epoch": 6.447063523771474, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16137 + }, + { + "epoch": 6.447463044346784, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16138 + }, + { + "epoch": 6.447862564922094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16139 + }, + { + "epoch": 6.448262085497403, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16140 + }, + { + "epoch": 6.448661606072712, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16141 + }, + { + "epoch": 6.449061126648022, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16142 + }, + { + "epoch": 6.449460647223332, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16143 + }, + { + "epoch": 6.449860167798642, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16144 + }, + { + "epoch": 6.450259688373952, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16145 + }, + { + "epoch": 6.4506592089492605, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16146 + }, + { + "epoch": 6.45105872952457, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16147 + }, + { + "epoch": 6.45145825009988, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16148 + }, + { + "epoch": 6.45185777067519, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16149 + }, + { + "epoch": 6.4522572912505, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16150 + }, + { + "epoch": 6.452656811825809, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16151 + }, + { + "epoch": 6.453056332401118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16152 + }, + { + "epoch": 6.453455852976428, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16153 + }, + { + "epoch": 6.453855373551738, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16154 + }, + { + "epoch": 6.454254894127048, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16155 + }, + { + "epoch": 6.4546544147023575, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16156 + }, + { + "epoch": 6.455053935277666, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16157 + }, + { + "epoch": 6.455453455852976, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16158 + }, + { + "epoch": 6.455852976428286, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16159 + }, + { + "epoch": 6.456252497003596, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16160 + }, + { + "epoch": 6.4566520175789055, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16161 + }, + { + "epoch": 6.457051538154215, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16162 + }, + { + "epoch": 6.457451058729524, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16163 + }, + { + "epoch": 6.457850579304834, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16164 + }, + { + "epoch": 6.458250099880144, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16165 + }, + { + "epoch": 6.4586496204554535, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16166 + }, + { + "epoch": 6.459049141030763, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16167 + }, + { + "epoch": 6.459448661606073, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16168 + }, + { + "epoch": 6.459848182181382, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16169 + }, + { + "epoch": 6.460247702756692, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 16170 + }, + { + "epoch": 6.460647223332002, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16171 + }, + { + "epoch": 6.461046743907311, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16172 + }, + { + "epoch": 6.461446264482621, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16173 + }, + { + "epoch": 6.46184578505793, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16174 + }, + { + "epoch": 6.46224530563324, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16175 + }, + { + "epoch": 6.46264482620855, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16176 + }, + { + "epoch": 6.463044346783859, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16177 + }, + { + "epoch": 6.463443867359169, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16178 + }, + { + "epoch": 6.463843387934479, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16179 + }, + { + "epoch": 6.464242908509788, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16180 + }, + { + "epoch": 6.464642429085098, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16181 + }, + { + "epoch": 6.4650419496604075, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16182 + }, + { + "epoch": 6.465441470235717, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16183 + }, + { + "epoch": 6.465840990811027, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16184 + }, + { + "epoch": 6.466240511386337, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16185 + }, + { + "epoch": 6.466640031961646, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16186 + }, + { + "epoch": 6.4670395525369555, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16187 + }, + { + "epoch": 6.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 16188 + }, + { + "epoch": 6.467838593687575, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16189 + }, + { + "epoch": 6.468238114262885, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16190 + }, + { + "epoch": 6.468637634838194, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16191 + }, + { + "epoch": 6.4690371554135035, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16192 + }, + { + "epoch": 6.469436675988813, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16193 + }, + { + "epoch": 6.469836196564123, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16194 + }, + { + "epoch": 6.470235717139433, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16195 + }, + { + "epoch": 6.470635237714743, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16196 + }, + { + "epoch": 6.471034758290052, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16197 + }, + { + "epoch": 6.471434278865361, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16198 + }, + { + "epoch": 6.471833799440671, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16199 + }, + { + "epoch": 6.472233320015981, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16200 + }, + { + "epoch": 6.472632840591291, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.2986, + "step": 16201 + }, + { + "epoch": 6.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16202 + }, + { + "epoch": 6.473431881741909, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16203 + }, + { + "epoch": 6.473831402317219, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16204 + }, + { + "epoch": 6.474230922892529, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16205 + }, + { + "epoch": 6.474630443467839, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16206 + }, + { + "epoch": 6.475029964043149, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16207 + }, + { + "epoch": 6.4754294846184575, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16208 + }, + { + "epoch": 6.475829005193767, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16209 + }, + { + "epoch": 6.476228525769077, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16210 + }, + { + "epoch": 6.476628046344387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16211 + }, + { + "epoch": 6.477027566919697, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16212 + }, + { + "epoch": 6.477427087495006, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16213 + }, + { + "epoch": 6.477826608070315, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16214 + }, + { + "epoch": 6.478226128645625, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16215 + }, + { + "epoch": 6.478625649220935, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16216 + }, + { + "epoch": 6.479025169796245, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16217 + }, + { + "epoch": 6.479424690371554, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16218 + }, + { + "epoch": 6.479824210946864, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 16219 + }, + { + "epoch": 6.480223731522173, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16220 + }, + { + "epoch": 6.480623252097483, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16221 + }, + { + "epoch": 6.481022772672793, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16222 + }, + { + "epoch": 6.4814222932481025, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16223 + }, + { + "epoch": 6.481821813823412, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16224 + }, + { + "epoch": 6.482221334398721, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16225 + }, + { + "epoch": 6.482620854974031, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16226 + }, + { + "epoch": 6.483020375549341, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16227 + }, + { + "epoch": 6.4834198961246505, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16228 + }, + { + "epoch": 6.48381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16229 + }, + { + "epoch": 6.484218937275269, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16230 + }, + { + "epoch": 6.484618457850579, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16231 + }, + { + "epoch": 6.485017978425889, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16232 + }, + { + "epoch": 6.485417499001199, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16233 + }, + { + "epoch": 6.485817019576508, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16234 + }, + { + "epoch": 6.486216540151818, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16235 + }, + { + "epoch": 6.486616060727127, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16236 + }, + { + "epoch": 6.487015581302437, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16237 + }, + { + "epoch": 6.487415101877747, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16238 + }, + { + "epoch": 6.487814622453056, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16239 + }, + { + "epoch": 6.488214143028366, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16240 + }, + { + "epoch": 6.488613663603676, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16241 + }, + { + "epoch": 6.489013184178985, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16242 + }, + { + "epoch": 6.489412704754295, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16243 + }, + { + "epoch": 6.4898122253296044, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16244 + }, + { + "epoch": 6.490211745904914, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16245 + }, + { + "epoch": 6.490611266480224, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16246 + }, + { + "epoch": 6.491010787055533, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16247 + }, + { + "epoch": 6.491410307630843, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16248 + }, + { + "epoch": 6.4918098282061525, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16249 + }, + { + "epoch": 6.492209348781462, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16250 + }, + { + "epoch": 6.492608869356772, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.335, + "step": 16251 + }, + { + "epoch": 6.493008389932082, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16252 + }, + { + "epoch": 6.493407910507391, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16253 + }, + { + "epoch": 6.4938074310827005, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16254 + }, + { + "epoch": 6.49420695165801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16255 + }, + { + "epoch": 6.49460647223332, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16256 + }, + { + "epoch": 6.49500599280863, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16257 + }, + { + "epoch": 6.49540551338394, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16258 + }, + { + "epoch": 6.495805033959249, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16259 + }, + { + "epoch": 6.496204554534558, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16260 + }, + { + "epoch": 6.496604075109868, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16261 + }, + { + "epoch": 6.497003595685178, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16262 + }, + { + "epoch": 6.497403116260488, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16263 + }, + { + "epoch": 6.497802636835797, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16264 + }, + { + "epoch": 6.498202157411106, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16265 + }, + { + "epoch": 6.498601677986416, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16266 + }, + { + "epoch": 6.499001198561726, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 16267 + }, + { + "epoch": 6.499400719137036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 16268 + }, + { + "epoch": 6.4998002397123456, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16269 + }, + { + "epoch": 6.5001997602876544, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 16270 + }, + { + "epoch": 6.500599280862964, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 16271 + }, + { + "epoch": 6.500998801438274, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16272 + }, + { + "epoch": 6.501398322013584, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3266, + "step": 16273 + }, + { + "epoch": 6.501797842588894, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16274 + }, + { + "epoch": 6.502197363164203, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16275 + }, + { + "epoch": 6.502596883739512, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16276 + }, + { + "epoch": 6.502996404314822, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16277 + }, + { + "epoch": 6.503395924890132, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16278 + }, + { + "epoch": 6.503795445465442, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16279 + }, + { + "epoch": 6.504194966040751, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16280 + }, + { + "epoch": 6.50459448661606, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.2957, + "step": 16281 + }, + { + "epoch": 6.50499400719137, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 16282 + }, + { + "epoch": 6.50539352776668, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16283 + }, + { + "epoch": 6.50579304834199, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3334, + "step": 16284 + }, + { + "epoch": 6.5061925689172995, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16285 + }, + { + "epoch": 6.506592089492609, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16286 + }, + { + "epoch": 6.506991610067918, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16287 + }, + { + "epoch": 6.507391130643228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16288 + }, + { + "epoch": 6.507790651218538, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16289 + }, + { + "epoch": 6.5081901717938475, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 16290 + }, + { + "epoch": 6.508589692369157, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16291 + }, + { + "epoch": 6.508989212944467, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16292 + }, + { + "epoch": 6.509388733519776, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16293 + }, + { + "epoch": 6.509788254095086, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16294 + }, + { + "epoch": 6.5101877746703956, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16295 + }, + { + "epoch": 6.510587295245705, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 16296 + }, + { + "epoch": 6.510986815821015, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16297 + }, + { + "epoch": 6.511386336396324, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16298 + }, + { + "epoch": 6.511785856971634, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16299 + }, + { + "epoch": 6.512185377546944, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16300 + }, + { + "epoch": 6.512584898122253, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16301 + }, + { + "epoch": 6.512984418697563, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16302 + }, + { + "epoch": 6.513383939272872, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16303 + }, + { + "epoch": 6.513783459848182, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16304 + }, + { + "epoch": 6.514182980423492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 16305 + }, + { + "epoch": 6.514582500998801, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16306 + }, + { + "epoch": 6.514982021574111, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16307 + }, + { + "epoch": 6.515381542149421, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16308 + }, + { + "epoch": 6.515781062724731, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16309 + }, + { + "epoch": 6.51618058330004, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16310 + }, + { + "epoch": 6.5165801038753495, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16311 + }, + { + "epoch": 6.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16312 + }, + { + "epoch": 6.517379145025969, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16313 + }, + { + "epoch": 6.517778665601279, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16314 + }, + { + "epoch": 6.518178186176588, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16315 + }, + { + "epoch": 6.5185777067518975, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16316 + }, + { + "epoch": 6.518977227327207, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16317 + }, + { + "epoch": 6.519376747902517, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3328, + "step": 16318 + }, + { + "epoch": 6.519776268477827, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16319 + }, + { + "epoch": 6.520175789053136, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 16320 + }, + { + "epoch": 6.520575309628446, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16321 + }, + { + "epoch": 6.520974830203755, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16322 + }, + { + "epoch": 6.521374350779065, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16323 + }, + { + "epoch": 6.521773871354375, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16324 + }, + { + "epoch": 6.522173391929685, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16325 + }, + { + "epoch": 6.5225729125049945, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16326 + }, + { + "epoch": 6.522972433080303, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16327 + }, + { + "epoch": 6.523371953655613, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16328 + }, + { + "epoch": 6.523771474230923, + "grad_norm": 0.12451171875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 16329 + }, + { + "epoch": 6.524170994806233, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16330 + }, + { + "epoch": 6.5245705153815425, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 16331 + }, + { + "epoch": 6.524970035956851, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16332 + }, + { + "epoch": 6.525369556532161, + "grad_norm": 0.205078125, + "learning_rate": 0.0002, + "loss": 1.3347, + "step": 16333 + }, + { + "epoch": 6.525769077107471, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 16334 + }, + { + "epoch": 6.526168597682781, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16335 + }, + { + "epoch": 6.526568118258091, + "grad_norm": 0.1650390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16336 + }, + { + "epoch": 6.5269676388333995, + "grad_norm": 0.1923828125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16337 + }, + { + "epoch": 6.527367159408709, + "grad_norm": 0.1533203125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16338 + }, + { + "epoch": 6.527766679984019, + "grad_norm": 0.1435546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16339 + }, + { + "epoch": 6.528166200559329, + "grad_norm": 0.1796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16340 + }, + { + "epoch": 6.528565721134639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16341 + }, + { + "epoch": 6.528965241709948, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16342 + }, + { + "epoch": 6.529364762285257, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16343 + }, + { + "epoch": 6.529764282860567, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16344 + }, + { + "epoch": 6.530163803435877, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16345 + }, + { + "epoch": 6.530563324011187, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3342, + "step": 16346 + }, + { + "epoch": 6.5309628445864965, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 16347 + }, + { + "epoch": 6.531362365161806, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 16348 + }, + { + "epoch": 6.531761885737115, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16349 + }, + { + "epoch": 6.532161406312425, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16350 + }, + { + "epoch": 6.532560926887735, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16351 + }, + { + "epoch": 6.5329604474630445, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16352 + }, + { + "epoch": 6.533359968038354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16353 + }, + { + "epoch": 6.533759488613663, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 16354 + }, + { + "epoch": 6.534159009188973, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16355 + }, + { + "epoch": 6.534558529764283, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16356 + }, + { + "epoch": 6.5349580503395925, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16357 + }, + { + "epoch": 6.535357570914902, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 16358 + }, + { + "epoch": 6.535757091490212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16359 + }, + { + "epoch": 6.536156612065521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16360 + }, + { + "epoch": 6.536556132640831, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16361 + }, + { + "epoch": 6.536955653216141, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 16362 + }, + { + "epoch": 6.53735517379145, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16363 + }, + { + "epoch": 6.53775469436676, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16364 + }, + { + "epoch": 6.53815421494207, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 16365 + }, + { + "epoch": 6.538553735517379, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16366 + }, + { + "epoch": 6.538953256092689, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16367 + }, + { + "epoch": 6.539352776667998, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 16368 + }, + { + "epoch": 6.539752297243308, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16369 + }, + { + "epoch": 6.540151817818618, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 16370 + }, + { + "epoch": 6.540551338393927, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16371 + }, + { + "epoch": 6.540950858969237, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16372 + }, + { + "epoch": 6.5413503795445465, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16373 + }, + { + "epoch": 6.541749900119856, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16374 + }, + { + "epoch": 6.542149420695166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16375 + }, + { + "epoch": 6.542548941270476, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16376 + }, + { + "epoch": 6.542948461845785, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16377 + }, + { + "epoch": 6.5433479824210945, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16378 + }, + { + "epoch": 6.543747502996404, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16379 + }, + { + "epoch": 6.544147023571714, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16380 + }, + { + "epoch": 6.544546544147024, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16381 + }, + { + "epoch": 6.544946064722334, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16382 + }, + { + "epoch": 6.5453455852976425, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16383 + }, + { + "epoch": 6.545745105872952, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16384 + }, + { + "epoch": 6.546144626448262, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16385 + }, + { + "epoch": 6.546544147023572, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16386 + }, + { + "epoch": 6.546943667598882, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16387 + }, + { + "epoch": 6.547343188174191, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 16388 + }, + { + "epoch": 6.5477427087495, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16389 + }, + { + "epoch": 6.54814222932481, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16390 + }, + { + "epoch": 6.54854174990012, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16391 + }, + { + "epoch": 6.54894127047543, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16392 + }, + { + "epoch": 6.549340791050739, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16393 + }, + { + "epoch": 6.549740311626048, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16394 + }, + { + "epoch": 6.550139832201358, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16395 + }, + { + "epoch": 6.550539352776668, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16396 + }, + { + "epoch": 6.550938873351978, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3309, + "step": 16397 + }, + { + "epoch": 6.551338393927288, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 16398 + }, + { + "epoch": 6.551737914502597, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16399 + }, + { + "epoch": 6.552137435077906, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16400 + }, + { + "epoch": 6.552536955653216, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16401 + }, + { + "epoch": 6.552936476228526, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16402 + }, + { + "epoch": 6.553335996803836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16403 + }, + { + "epoch": 6.553735517379145, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3338, + "step": 16404 + }, + { + "epoch": 6.554135037954454, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16405 + }, + { + "epoch": 6.554534558529764, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16406 + }, + { + "epoch": 6.554934079105074, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16407 + }, + { + "epoch": 6.555333599680384, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 16408 + }, + { + "epoch": 6.555733120255693, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16409 + }, + { + "epoch": 6.556132640831002, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16410 + }, + { + "epoch": 6.556532161406312, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 16411 + }, + { + "epoch": 6.556931681981622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16412 + }, + { + "epoch": 6.557331202556932, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16413 + }, + { + "epoch": 6.5577307231322415, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16414 + }, + { + "epoch": 6.558130243707551, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 16415 + }, + { + "epoch": 6.558529764282861, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16416 + }, + { + "epoch": 6.55892928485817, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16417 + }, + { + "epoch": 6.55932880543348, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16418 + }, + { + "epoch": 6.5597283260087895, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16419 + }, + { + "epoch": 6.560127846584099, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16420 + }, + { + "epoch": 6.560527367159409, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16421 + }, + { + "epoch": 6.560926887734718, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16422 + }, + { + "epoch": 6.561326408310028, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16423 + }, + { + "epoch": 6.561725928885338, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16424 + }, + { + "epoch": 6.562125449460647, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16425 + }, + { + "epoch": 6.562524970035957, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16426 + }, + { + "epoch": 6.562924490611266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16427 + }, + { + "epoch": 6.563324011186576, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16428 + }, + { + "epoch": 6.563723531761886, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16429 + }, + { + "epoch": 6.564123052337195, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16430 + }, + { + "epoch": 6.564522572912505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16431 + }, + { + "epoch": 6.564922093487815, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16432 + }, + { + "epoch": 6.565321614063124, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16433 + }, + { + "epoch": 6.565721134638434, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16434 + }, + { + "epoch": 6.566120655213743, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16435 + }, + { + "epoch": 6.566520175789053, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16436 + }, + { + "epoch": 6.566919696364363, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16437 + }, + { + "epoch": 6.567319216939673, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16438 + }, + { + "epoch": 6.567718737514982, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 16439 + }, + { + "epoch": 6.5681182580902915, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 16440 + }, + { + "epoch": 6.568517778665601, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16441 + }, + { + "epoch": 6.568917299240911, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16442 + }, + { + "epoch": 6.569316819816221, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16443 + }, + { + "epoch": 6.56971634039153, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16444 + }, + { + "epoch": 6.5701158609668395, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16445 + }, + { + "epoch": 6.570515381542149, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16446 + }, + { + "epoch": 6.570914902117459, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 16447 + }, + { + "epoch": 6.571314422692769, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16448 + }, + { + "epoch": 6.571713943268079, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16449 + }, + { + "epoch": 6.572113463843388, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16450 + }, + { + "epoch": 6.572512984418697, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16451 + }, + { + "epoch": 6.572912504994007, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16452 + }, + { + "epoch": 6.573312025569317, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16453 + }, + { + "epoch": 6.573711546144627, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16454 + }, + { + "epoch": 6.5741110667199365, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16455 + }, + { + "epoch": 6.574510587295245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16456 + }, + { + "epoch": 6.574910107870555, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16457 + }, + { + "epoch": 6.575309628445865, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16458 + }, + { + "epoch": 6.575709149021175, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 16459 + }, + { + "epoch": 6.5761086695964845, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16460 + }, + { + "epoch": 6.5765081901717934, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16461 + }, + { + "epoch": 6.576907710747103, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16462 + }, + { + "epoch": 6.577307231322413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16463 + }, + { + "epoch": 6.577706751897723, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16464 + }, + { + "epoch": 6.578106272473033, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16465 + }, + { + "epoch": 6.578505793048342, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16466 + }, + { + "epoch": 6.578905313623651, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16467 + }, + { + "epoch": 6.579304834198961, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16468 + }, + { + "epoch": 6.579704354774271, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16469 + }, + { + "epoch": 6.580103875349581, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16470 + }, + { + "epoch": 6.58050339592489, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16471 + }, + { + "epoch": 6.5809029165002, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16472 + }, + { + "epoch": 6.581302437075509, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16473 + }, + { + "epoch": 6.581701957650819, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16474 + }, + { + "epoch": 6.582101478226129, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 16475 + }, + { + "epoch": 6.5825009988014385, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16476 + }, + { + "epoch": 6.582900519376748, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16477 + }, + { + "epoch": 6.583300039952057, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16478 + }, + { + "epoch": 6.583699560527367, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16479 + }, + { + "epoch": 6.584099081102677, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16480 + }, + { + "epoch": 6.5844986016779865, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16481 + }, + { + "epoch": 6.584898122253296, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16482 + }, + { + "epoch": 6.585297642828605, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16483 + }, + { + "epoch": 6.585697163403915, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16484 + }, + { + "epoch": 6.586096683979225, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16485 + }, + { + "epoch": 6.5864962045545346, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16486 + }, + { + "epoch": 6.586895725129844, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16487 + }, + { + "epoch": 6.587295245705154, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16488 + }, + { + "epoch": 6.587694766280464, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16489 + }, + { + "epoch": 6.588094286855773, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16490 + }, + { + "epoch": 6.588493807431083, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16491 + }, + { + "epoch": 6.588893328006392, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16492 + }, + { + "epoch": 6.589292848581702, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16493 + }, + { + "epoch": 6.589692369157012, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 16494 + }, + { + "epoch": 6.590091889732321, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16495 + }, + { + "epoch": 6.590491410307631, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16496 + }, + { + "epoch": 6.59089093088294, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16497 + }, + { + "epoch": 6.59129045145825, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16498 + }, + { + "epoch": 6.59168997203356, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16499 + }, + { + "epoch": 6.592089492608869, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16500 + }, + { + "epoch": 6.592489013184179, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16501 + }, + { + "epoch": 6.5928885337594885, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16502 + }, + { + "epoch": 6.593288054334798, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16503 + }, + { + "epoch": 6.593687574910108, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16504 + }, + { + "epoch": 6.594087095485418, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16505 + }, + { + "epoch": 6.594486616060728, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16506 + }, + { + "epoch": 6.5948861366360365, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16507 + }, + { + "epoch": 6.595285657211346, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16508 + }, + { + "epoch": 6.595685177786656, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16509 + }, + { + "epoch": 6.596084698361966, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16510 + }, + { + "epoch": 6.596484218937276, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16511 + }, + { + "epoch": 6.5968837395125846, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16512 + }, + { + "epoch": 6.597283260087894, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16513 + }, + { + "epoch": 6.597682780663204, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 16514 + }, + { + "epoch": 6.598082301238514, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16515 + }, + { + "epoch": 6.598481821813824, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 16516 + }, + { + "epoch": 6.598881342389133, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 16517 + }, + { + "epoch": 6.599280862964442, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16518 + }, + { + "epoch": 6.599680383539752, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16519 + }, + { + "epoch": 6.600079904115062, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16520 + }, + { + "epoch": 6.600479424690372, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16521 + }, + { + "epoch": 6.6008789452656815, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16522 + }, + { + "epoch": 6.60127846584099, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16523 + }, + { + "epoch": 6.6016779864163, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16524 + }, + { + "epoch": 6.60207750699161, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 16525 + }, + { + "epoch": 6.60247702756692, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16526 + }, + { + "epoch": 6.60287654814223, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16527 + }, + { + "epoch": 6.603276068717539, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16528 + }, + { + "epoch": 6.603675589292848, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16529 + }, + { + "epoch": 6.604075109868158, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16530 + }, + { + "epoch": 6.604474630443468, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16531 + }, + { + "epoch": 6.604874151018778, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16532 + }, + { + "epoch": 6.605273671594087, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 16533 + }, + { + "epoch": 6.605673192169396, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16534 + }, + { + "epoch": 6.606072712744706, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16535 + }, + { + "epoch": 6.606472233320016, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16536 + }, + { + "epoch": 6.606871753895326, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16537 + }, + { + "epoch": 6.6072712744706354, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16538 + }, + { + "epoch": 6.607670795045945, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16539 + }, + { + "epoch": 6.608070315621254, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16540 + }, + { + "epoch": 6.608469836196564, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16541 + }, + { + "epoch": 6.608869356771874, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 16542 + }, + { + "epoch": 6.6092688773471835, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16543 + }, + { + "epoch": 6.609668397922493, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16544 + }, + { + "epoch": 6.610067918497803, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16545 + }, + { + "epoch": 6.610467439073112, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16546 + }, + { + "epoch": 6.610866959648422, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16547 + }, + { + "epoch": 6.6112664802237315, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16548 + }, + { + "epoch": 6.611666000799041, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16549 + }, + { + "epoch": 6.612065521374351, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16550 + }, + { + "epoch": 6.61246504194966, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16551 + }, + { + "epoch": 6.61286456252497, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16552 + }, + { + "epoch": 6.61326408310028, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.338, + "step": 16553 + }, + { + "epoch": 6.613663603675589, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16554 + }, + { + "epoch": 6.614063124250899, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 16555 + }, + { + "epoch": 6.614462644826209, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16556 + }, + { + "epoch": 6.614862165401518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 16557 + }, + { + "epoch": 6.615261685976828, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16558 + }, + { + "epoch": 6.615661206552137, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16559 + }, + { + "epoch": 6.616060727127447, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16560 + }, + { + "epoch": 6.616460247702757, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16561 + }, + { + "epoch": 6.616859768278067, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16562 + }, + { + "epoch": 6.617259288853376, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16563 + }, + { + "epoch": 6.6176588094286855, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 16564 + }, + { + "epoch": 6.618058330003995, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16565 + }, + { + "epoch": 6.618457850579305, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16566 + }, + { + "epoch": 6.618857371154615, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16567 + }, + { + "epoch": 6.619256891729924, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16568 + }, + { + "epoch": 6.6196564123052335, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 16569 + }, + { + "epoch": 6.620055932880543, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16570 + }, + { + "epoch": 6.620455453455853, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16571 + }, + { + "epoch": 6.620854974031163, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16572 + }, + { + "epoch": 6.621254494606472, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 16573 + }, + { + "epoch": 6.6216540151817815, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16574 + }, + { + "epoch": 6.622053535757091, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 16575 + }, + { + "epoch": 6.622453056332401, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16576 + }, + { + "epoch": 6.622852576907711, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 16577 + }, + { + "epoch": 6.623252097483021, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16578 + }, + { + "epoch": 6.6236516180583305, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16579 + }, + { + "epoch": 6.624051138633639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16580 + }, + { + "epoch": 6.624450659208949, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16581 + }, + { + "epoch": 6.624850179784259, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16582 + }, + { + "epoch": 6.625249700359569, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16583 + }, + { + "epoch": 6.6256492209348785, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16584 + }, + { + "epoch": 6.626048741510187, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16585 + }, + { + "epoch": 6.626448262085497, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16586 + }, + { + "epoch": 6.626847782660807, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 16587 + }, + { + "epoch": 6.627247303236117, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16588 + }, + { + "epoch": 6.627646823811427, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16589 + }, + { + "epoch": 6.6280463443867355, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 16590 + }, + { + "epoch": 6.628445864962045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 16591 + }, + { + "epoch": 6.628845385537355, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16592 + }, + { + "epoch": 6.629244906112665, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 16593 + }, + { + "epoch": 6.629644426687975, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 16594 + }, + { + "epoch": 6.630043947263284, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16595 + }, + { + "epoch": 6.630443467838594, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16596 + }, + { + "epoch": 6.630842988413903, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16597 + }, + { + "epoch": 6.631242508989213, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16598 + }, + { + "epoch": 6.631642029564523, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 16599 + }, + { + "epoch": 6.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16600 + }, + { + "epoch": 6.632441070715142, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16601 + }, + { + "epoch": 6.632840591290451, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16602 + }, + { + "epoch": 6.633240111865761, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16603 + }, + { + "epoch": 6.633639632441071, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16604 + }, + { + "epoch": 6.6340391530163805, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16605 + }, + { + "epoch": 6.63443867359169, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16606 + }, + { + "epoch": 6.634838194166999, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16607 + }, + { + "epoch": 6.635237714742309, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16608 + }, + { + "epoch": 6.635637235317619, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 16609 + }, + { + "epoch": 6.6360367558929285, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16610 + }, + { + "epoch": 6.636436276468238, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16611 + }, + { + "epoch": 6.636835797043548, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16612 + }, + { + "epoch": 6.637235317618857, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16613 + }, + { + "epoch": 6.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16614 + }, + { + "epoch": 6.638034358769477, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 16615 + }, + { + "epoch": 6.638433879344786, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 16616 + }, + { + "epoch": 6.638833399920096, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 16617 + }, + { + "epoch": 6.639232920495406, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16618 + }, + { + "epoch": 6.639632441070715, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16619 + }, + { + "epoch": 6.640031961646025, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16620 + }, + { + "epoch": 6.640431482221334, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16621 + }, + { + "epoch": 6.640831002796644, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16622 + }, + { + "epoch": 6.641230523371954, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16623 + }, + { + "epoch": 6.641630043947263, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 16624 + }, + { + "epoch": 6.642029564522573, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16625 + }, + { + "epoch": 6.642429085097882, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16626 + }, + { + "epoch": 6.642828605673192, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16627 + }, + { + "epoch": 6.643228126248502, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16628 + }, + { + "epoch": 6.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16629 + }, + { + "epoch": 6.644027167399121, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 16630 + }, + { + "epoch": 6.6444266879744305, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16631 + }, + { + "epoch": 6.64482620854974, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16632 + }, + { + "epoch": 6.64522572912505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 16633 + }, + { + "epoch": 6.64562524970036, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16634 + }, + { + "epoch": 6.64602477027567, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 16635 + }, + { + "epoch": 6.6464242908509785, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16636 + }, + { + "epoch": 6.646823811426288, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 16637 + }, + { + "epoch": 6.647223332001598, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16638 + }, + { + "epoch": 6.647622852576908, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16639 + }, + { + "epoch": 6.648022373152218, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16640 + }, + { + "epoch": 6.648421893727527, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16641 + }, + { + "epoch": 6.648821414302836, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16642 + }, + { + "epoch": 6.649220934878146, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 16643 + }, + { + "epoch": 6.649620455453456, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 16644 + }, + { + "epoch": 6.650019976028766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16645 + }, + { + "epoch": 6.6504194966040755, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16646 + }, + { + "epoch": 6.650819017179384, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16647 + }, + { + "epoch": 6.651218537754694, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16648 + }, + { + "epoch": 6.651618058330004, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16649 + }, + { + "epoch": 6.652017578905314, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16650 + }, + { + "epoch": 6.6524170994806235, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16651 + }, + { + "epoch": 6.652816620055933, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16652 + }, + { + "epoch": 6.653216140631242, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 16653 + }, + { + "epoch": 6.653615661206552, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16654 + }, + { + "epoch": 6.654015181781862, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16655 + }, + { + "epoch": 6.654414702357172, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16656 + }, + { + "epoch": 6.654814222932481, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 16657 + }, + { + "epoch": 6.65521374350779, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 16658 + }, + { + "epoch": 6.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16659 + }, + { + "epoch": 6.65601278465841, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 16660 + }, + { + "epoch": 6.65641230523372, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16661 + }, + { + "epoch": 6.656811825809029, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 16662 + }, + { + "epoch": 6.657211346384338, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 16663 + }, + { + "epoch": 6.657610866959648, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16664 + }, + { + "epoch": 6.658010387534958, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16665 + }, + { + "epoch": 6.658409908110268, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 16666 + }, + { + "epoch": 6.6588094286855775, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16667 + }, + { + "epoch": 6.659208949260887, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16668 + }, + { + "epoch": 6.659608469836197, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 16669 + }, + { + "epoch": 6.660007990411506, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 16670 + }, + { + "epoch": 6.660407510986816, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16671 + }, + { + "epoch": 6.6608070315621255, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16672 + }, + { + "epoch": 6.661206552137435, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 16673 + }, + { + "epoch": 6.661606072712745, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 16674 + }, + { + "epoch": 6.662005593288054, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 16675 + }, + { + "epoch": 6.662405113863364, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16676 + }, + { + "epoch": 6.6628046344386735, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16677 + }, + { + "epoch": 6.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16678 + }, + { + "epoch": 6.663603675589293, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16679 + }, + { + "epoch": 6.664003196164602, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16680 + }, + { + "epoch": 6.664402716739912, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16681 + }, + { + "epoch": 6.664802237315222, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16682 + }, + { + "epoch": 6.665201757890531, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16683 + }, + { + "epoch": 6.665601278465841, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16684 + }, + { + "epoch": 6.666000799041151, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 16685 + }, + { + "epoch": 6.666400319616461, + "grad_norm": 0.14453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16686 + }, + { + "epoch": 6.66679984019177, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16687 + }, + { + "epoch": 6.667199360767079, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16688 + }, + { + "epoch": 6.667598881342389, + "grad_norm": 0.1328125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16689 + }, + { + "epoch": 6.667998401917699, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16690 + }, + { + "epoch": 6.668397922493009, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16691 + }, + { + "epoch": 6.668797443068318, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16692 + }, + { + "epoch": 6.6691969636436275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16693 + }, + { + "epoch": 6.669596484218937, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16694 + }, + { + "epoch": 6.669996004794247, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16695 + }, + { + "epoch": 6.670395525369557, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16696 + }, + { + "epoch": 6.670795045944866, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 16697 + }, + { + "epoch": 6.6711945665201755, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 16698 + }, + { + "epoch": 6.671594087095485, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16699 + }, + { + "epoch": 6.671993607670795, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16700 + }, + { + "epoch": 6.672393128246105, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 16701 + }, + { + "epoch": 6.672792648821415, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 16702 + }, + { + "epoch": 6.6731921693967235, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16703 + }, + { + "epoch": 6.673591689972033, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16704 + }, + { + "epoch": 6.673991210547343, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16705 + }, + { + "epoch": 6.674390731122653, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.299, + "step": 16706 + }, + { + "epoch": 6.674790251697963, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 16707 + }, + { + "epoch": 6.6751897722732725, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 16708 + }, + { + "epoch": 6.675589292848581, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16709 + }, + { + "epoch": 6.675988813423891, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16710 + }, + { + "epoch": 6.676388333999201, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16711 + }, + { + "epoch": 6.676787854574511, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16712 + }, + { + "epoch": 6.6771873751498205, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16713 + }, + { + "epoch": 6.677586895725129, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 16714 + }, + { + "epoch": 6.677986416300439, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16715 + }, + { + "epoch": 6.678385936875749, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16716 + }, + { + "epoch": 6.678785457451059, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 16717 + }, + { + "epoch": 6.679184978026369, + "grad_norm": 0.123046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16718 + }, + { + "epoch": 6.679584498601678, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16719 + }, + { + "epoch": 6.679984019176987, + "grad_norm": 0.1640625, + "learning_rate": 0.0002, + "loss": 1.3301, + "step": 16720 + }, + { + "epoch": 6.680383539752297, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16721 + }, + { + "epoch": 6.680783060327607, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16722 + }, + { + "epoch": 6.681182580902917, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16723 + }, + { + "epoch": 6.681582101478226, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16724 + }, + { + "epoch": 6.681981622053536, + "grad_norm": 0.10498046875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 16725 + }, + { + "epoch": 6.682381142628845, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16726 + }, + { + "epoch": 6.682780663204155, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16727 + }, + { + "epoch": 6.683180183779465, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16728 + }, + { + "epoch": 6.683579704354774, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16729 + }, + { + "epoch": 6.683979224930084, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16730 + }, + { + "epoch": 6.684378745505393, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16731 + }, + { + "epoch": 6.684778266080703, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 16732 + }, + { + "epoch": 6.685177786656013, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 16733 + }, + { + "epoch": 6.6855773072313225, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 16734 + }, + { + "epoch": 6.685976827806632, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16735 + }, + { + "epoch": 6.686376348381942, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16736 + }, + { + "epoch": 6.686775868957251, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16737 + }, + { + "epoch": 6.687175389532561, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 16738 + }, + { + "epoch": 6.6875749101078705, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 16739 + }, + { + "epoch": 6.68797443068318, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16740 + }, + { + "epoch": 6.68837395125849, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16741 + }, + { + "epoch": 6.6887734718338, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 16742 + }, + { + "epoch": 6.689172992409109, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16743 + }, + { + "epoch": 6.689572512984419, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 16744 + }, + { + "epoch": 6.689972033559728, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16745 + }, + { + "epoch": 6.690371554135038, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 16746 + }, + { + "epoch": 6.690771074710348, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16747 + }, + { + "epoch": 6.691170595285657, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 16748 + }, + { + "epoch": 6.691570115860967, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16749 + }, + { + "epoch": 6.691969636436276, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 16750 + }, + { + "epoch": 6.692369157011586, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 16751 + }, + { + "epoch": 6.692768677586896, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 16752 + }, + { + "epoch": 6.693168198162205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16753 + }, + { + "epoch": 6.693567718737515, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16754 + }, + { + "epoch": 6.6939672393128244, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16755 + }, + { + "epoch": 6.694366759888134, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16756 + }, + { + "epoch": 6.694766280463444, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16757 + }, + { + "epoch": 6.695165801038754, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16758 + }, + { + "epoch": 6.695565321614064, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16759 + }, + { + "epoch": 6.6959648421893725, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16760 + }, + { + "epoch": 6.696364362764682, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 16761 + }, + { + "epoch": 6.696763883339992, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 16762 + }, + { + "epoch": 6.697163403915302, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16763 + }, + { + "epoch": 6.697562924490612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 16764 + }, + { + "epoch": 6.6979624450659205, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16765 + }, + { + "epoch": 6.69836196564123, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 16766 + }, + { + "epoch": 6.69876148621654, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 16767 + }, + { + "epoch": 6.69916100679185, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 16768 + }, + { + "epoch": 6.69956052736716, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16769 + }, + { + "epoch": 6.699960047942469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16770 + }, + { + "epoch": 6.700359568517778, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16771 + }, + { + "epoch": 6.700759089093088, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 16772 + }, + { + "epoch": 6.701158609668398, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16773 + }, + { + "epoch": 6.701558130243708, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 16774 + }, + { + "epoch": 6.7019576508190175, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16775 + }, + { + "epoch": 6.702357171394327, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 16776 + }, + { + "epoch": 6.702756691969636, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16777 + }, + { + "epoch": 6.703156212544946, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 16778 + }, + { + "epoch": 6.703555733120256, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16779 + }, + { + "epoch": 6.7039552536955656, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 16780 + }, + { + "epoch": 6.704354774270875, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16781 + }, + { + "epoch": 6.704754294846184, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 16782 + }, + { + "epoch": 6.705153815421494, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16783 + }, + { + "epoch": 6.705553335996804, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 16784 + }, + { + "epoch": 6.705952856572114, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16785 + }, + { + "epoch": 6.706352377147423, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16786 + }, + { + "epoch": 6.706751897722732, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16787 + }, + { + "epoch": 6.707151418298042, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16788 + }, + { + "epoch": 6.707550938873352, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16789 + }, + { + "epoch": 6.707950459448662, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 16790 + }, + { + "epoch": 6.708349980023971, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16791 + }, + { + "epoch": 6.708749500599281, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 16792 + }, + { + "epoch": 6.70914902117459, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16793 + }, + { + "epoch": 6.7095485417499, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16794 + }, + { + "epoch": 6.70994806232521, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16795 + }, + { + "epoch": 6.7103475829005195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16796 + }, + { + "epoch": 6.710747103475829, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16797 + }, + { + "epoch": 6.711146624051139, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16798 + }, + { + "epoch": 6.711546144626448, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 16799 + }, + { + "epoch": 6.711945665201758, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 16800 + }, + { + "epoch": 6.7123451857770675, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16801 + }, + { + "epoch": 6.712744706352377, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.2968, + "step": 16802 + }, + { + "epoch": 6.713144226927687, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 16803 + }, + { + "epoch": 6.713543747502996, + "grad_norm": 0.1572265625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16804 + }, + { + "epoch": 6.713943268078306, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16805 + }, + { + "epoch": 6.714342788653616, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16806 + }, + { + "epoch": 6.714742309228925, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16807 + }, + { + "epoch": 6.715141829804235, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 16808 + }, + { + "epoch": 6.715541350379545, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16809 + }, + { + "epoch": 6.715940870954854, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 16810 + }, + { + "epoch": 6.716340391530164, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16811 + }, + { + "epoch": 6.716739912105473, + "grad_norm": 0.134765625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16812 + }, + { + "epoch": 6.717139432680783, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 16813 + }, + { + "epoch": 6.717538953256093, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 16814 + }, + { + "epoch": 6.717938473831403, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 16815 + }, + { + "epoch": 6.718337994406712, + "grad_norm": 0.21484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16816 + }, + { + "epoch": 6.718737514982021, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 16817 + }, + { + "epoch": 6.719137035557331, + "grad_norm": 0.15625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 16818 + }, + { + "epoch": 6.719536556132641, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16819 + }, + { + "epoch": 6.719936076707951, + "grad_norm": 0.16796875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 16820 + }, + { + "epoch": 6.72033559728326, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 16821 + }, + { + "epoch": 6.7207351178585695, + "grad_norm": 0.1513671875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 16822 + }, + { + "epoch": 6.721134638433879, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 16823 + }, + { + "epoch": 6.721534159009189, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 16824 + }, + { + "epoch": 6.721933679584499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16825 + }, + { + "epoch": 6.722333200159808, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 16826 + }, + { + "epoch": 6.7227327207351175, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 16827 + }, + { + "epoch": 6.723132241310427, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 16828 + }, + { + "epoch": 6.723531761885737, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16829 + }, + { + "epoch": 6.723931282461047, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16830 + }, + { + "epoch": 6.724330803036357, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16831 + }, + { + "epoch": 6.7247303236116664, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16832 + }, + { + "epoch": 6.725129844186975, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 16833 + }, + { + "epoch": 6.725529364762285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16834 + }, + { + "epoch": 6.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16835 + }, + { + "epoch": 6.726328405912905, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 16836 + }, + { + "epoch": 6.7267279264882145, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 16837 + }, + { + "epoch": 6.727127447063523, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16838 + }, + { + "epoch": 6.727526967638833, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16839 + }, + { + "epoch": 6.727926488214143, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 16840 + }, + { + "epoch": 6.728326008789453, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16841 + }, + { + "epoch": 6.7287255293647625, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 16842 + }, + { + "epoch": 6.729125049940071, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 16843 + }, + { + "epoch": 6.729524570515381, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16844 + }, + { + "epoch": 6.729924091090691, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 16845 + }, + { + "epoch": 6.730323611666001, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16846 + }, + { + "epoch": 6.730723132241311, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 16847 + }, + { + "epoch": 6.73112265281662, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16848 + }, + { + "epoch": 6.73152217339193, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 16849 + }, + { + "epoch": 6.731921693967239, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 16850 + }, + { + "epoch": 6.732321214542549, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16851 + }, + { + "epoch": 6.732720735117859, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 16852 + }, + { + "epoch": 6.733120255693168, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16853 + }, + { + "epoch": 6.733519776268478, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 16854 + }, + { + "epoch": 6.733919296843787, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 16855 + }, + { + "epoch": 6.734318817419097, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16856 + }, + { + "epoch": 6.734718337994407, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16857 + }, + { + "epoch": 6.7351178585697165, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 16858 + }, + { + "epoch": 6.735517379145026, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16859 + }, + { + "epoch": 6.735916899720335, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 16860 + }, + { + "epoch": 6.736316420295645, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16861 + }, + { + "epoch": 6.736715940870955, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 16862 + }, + { + "epoch": 6.7371154614462645, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 16863 + }, + { + "epoch": 6.737514982021574, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 16864 + }, + { + "epoch": 6.737914502596884, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16865 + }, + { + "epoch": 6.738314023172194, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 16866 + }, + { + "epoch": 6.738713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 16867 + }, + { + "epoch": 6.7391130643228125, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 16868 + }, + { + "epoch": 6.739512584898122, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16869 + }, + { + "epoch": 6.739912105473432, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 16870 + }, + { + "epoch": 6.740311626048742, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 16871 + }, + { + "epoch": 6.740711146624051, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 16872 + }, + { + "epoch": 6.741110667199361, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16873 + }, + { + "epoch": 6.74151018777467, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 16874 + }, + { + "epoch": 6.74190970834998, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16875 + }, + { + "epoch": 6.74230922892529, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16876 + }, + { + "epoch": 6.742708749500599, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 16877 + }, + { + "epoch": 6.743108270075909, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 16878 + }, + { + "epoch": 6.743507790651218, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16879 + }, + { + "epoch": 6.743907311226528, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 16880 + }, + { + "epoch": 6.744306831801838, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16881 + }, + { + "epoch": 6.744706352377148, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 16882 + }, + { + "epoch": 6.745105872952457, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16883 + }, + { + "epoch": 6.7455053935277665, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16884 + }, + { + "epoch": 6.745904914103076, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 16885 + }, + { + "epoch": 6.746304434678386, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 16886 + }, + { + "epoch": 6.746703955253696, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 16887 + }, + { + "epoch": 6.747103475829006, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16888 + }, + { + "epoch": 6.7475029964043145, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 16889 + }, + { + "epoch": 6.747902516979624, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16890 + }, + { + "epoch": 6.748302037554934, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 16891 + }, + { + "epoch": 6.748701558130244, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 16892 + }, + { + "epoch": 6.749101078705554, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 16893 + }, + { + "epoch": 6.7495005992808625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 16894 + }, + { + "epoch": 6.749900119856172, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 16895 + }, + { + "epoch": 6.750299640431482, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16896 + }, + { + "epoch": 6.750699161006792, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 16897 + }, + { + "epoch": 6.751098681582102, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 16898 + }, + { + "epoch": 6.7514982021574115, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.296, + "step": 16899 + }, + { + "epoch": 6.75189772273272, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 16900 + }, + { + "epoch": 6.75229724330803, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16901 + }, + { + "epoch": 6.75269676388334, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16902 + }, + { + "epoch": 6.75309628445865, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16903 + }, + { + "epoch": 6.7534958050339595, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 16904 + }, + { + "epoch": 6.753895325609269, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 16905 + }, + { + "epoch": 6.754294846184578, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 16906 + }, + { + "epoch": 6.754694366759888, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.2997, + "step": 16907 + }, + { + "epoch": 6.755093887335198, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 16908 + }, + { + "epoch": 6.755493407910508, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 16909 + }, + { + "epoch": 6.755892928485817, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 16910 + }, + { + "epoch": 6.756292449061126, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 16911 + }, + { + "epoch": 6.756691969636436, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 16912 + }, + { + "epoch": 6.757091490211746, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 16913 + }, + { + "epoch": 6.757491010787056, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 16914 + }, + { + "epoch": 6.757890531362365, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 16915 + }, + { + "epoch": 6.758290051937674, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 16916 + }, + { + "epoch": 6.758689572512984, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16917 + }, + { + "epoch": 6.759089093088294, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 16918 + }, + { + "epoch": 6.759488613663604, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 16919 + }, + { + "epoch": 6.759888134238913, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 16920 + }, + { + "epoch": 6.760287654814223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 16921 + }, + { + "epoch": 6.760687175389533, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 16922 + }, + { + "epoch": 6.761086695964842, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 16923 + }, + { + "epoch": 6.761486216540152, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 16924 + }, + { + "epoch": 6.7618857371154615, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16925 + }, + { + "epoch": 6.762285257690771, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 16926 + }, + { + "epoch": 6.762684778266081, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 16927 + }, + { + "epoch": 6.76308429884139, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16928 + }, + { + "epoch": 6.7634838194167, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 16929 + }, + { + "epoch": 6.7638833399920095, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 16930 + }, + { + "epoch": 6.764282860567319, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 16931 + }, + { + "epoch": 6.764682381142629, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 16932 + }, + { + "epoch": 6.765081901717938, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16933 + }, + { + "epoch": 6.765481422293248, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 16934 + }, + { + "epoch": 6.765880942868558, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 16935 + }, + { + "epoch": 6.766280463443867, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16936 + }, + { + "epoch": 6.766679984019177, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 16937 + }, + { + "epoch": 6.767079504594487, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 16938 + }, + { + "epoch": 6.767479025169797, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 16939 + }, + { + "epoch": 6.767878545745106, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 16940 + }, + { + "epoch": 6.768278066320415, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 16941 + }, + { + "epoch": 6.768677586895725, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 16942 + }, + { + "epoch": 6.769077107471035, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 16943 + }, + { + "epoch": 6.769476628046345, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16944 + }, + { + "epoch": 6.769876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 16945 + }, + { + "epoch": 6.770275669196963, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 16946 + }, + { + "epoch": 6.770675189772273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16947 + }, + { + "epoch": 6.771074710347583, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 16948 + }, + { + "epoch": 6.771474230922893, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16949 + }, + { + "epoch": 6.771873751498202, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 16950 + }, + { + "epoch": 6.7722732720735115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 16951 + }, + { + "epoch": 6.772672792648821, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 16952 + }, + { + "epoch": 6.773072313224131, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 16953 + }, + { + "epoch": 6.773471833799441, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16954 + }, + { + "epoch": 6.773871354374751, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 16955 + }, + { + "epoch": 6.77427087495006, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 16956 + }, + { + "epoch": 6.774670395525369, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 16957 + }, + { + "epoch": 6.775069916100679, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 16958 + }, + { + "epoch": 6.775469436675989, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 16959 + }, + { + "epoch": 6.775868957251299, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 16960 + }, + { + "epoch": 6.7762684778266085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16961 + }, + { + "epoch": 6.776667998401917, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16962 + }, + { + "epoch": 6.777067518977227, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 16963 + }, + { + "epoch": 6.777467039552537, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 16964 + }, + { + "epoch": 6.777866560127847, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 16965 + }, + { + "epoch": 6.7782660807031565, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 16966 + }, + { + "epoch": 6.778665601278465, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 16967 + }, + { + "epoch": 6.779065121853775, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 16968 + }, + { + "epoch": 6.779464642429085, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 16969 + }, + { + "epoch": 6.779864163004395, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 16970 + }, + { + "epoch": 6.7802636835797045, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 16971 + }, + { + "epoch": 6.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 16972 + }, + { + "epoch": 6.781062724730323, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 16973 + }, + { + "epoch": 6.781462245305633, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 16974 + }, + { + "epoch": 6.781861765880943, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 16975 + }, + { + "epoch": 6.782261286456253, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 16976 + }, + { + "epoch": 6.782660807031562, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 16977 + }, + { + "epoch": 6.783060327606872, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16978 + }, + { + "epoch": 6.783459848182181, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 16979 + }, + { + "epoch": 6.783859368757491, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 16980 + }, + { + "epoch": 6.784258889332801, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 16981 + }, + { + "epoch": 6.78465840990811, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 16982 + }, + { + "epoch": 6.78505793048342, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 16983 + }, + { + "epoch": 6.785457451058729, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 16984 + }, + { + "epoch": 6.785856971634039, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16985 + }, + { + "epoch": 6.786256492209349, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3003, + "step": 16986 + }, + { + "epoch": 6.7866560127846585, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 16987 + }, + { + "epoch": 6.787055533359968, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 16988 + }, + { + "epoch": 6.787455053935278, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 16989 + }, + { + "epoch": 6.787854574510587, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 16990 + }, + { + "epoch": 6.788254095085897, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 16991 + }, + { + "epoch": 6.7886536156612065, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 16992 + }, + { + "epoch": 6.789053136236516, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 16993 + }, + { + "epoch": 6.789452656811826, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 16994 + }, + { + "epoch": 6.789852177387136, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 16995 + }, + { + "epoch": 6.790251697962445, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 16996 + }, + { + "epoch": 6.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 16997 + }, + { + "epoch": 6.791050739113064, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 16998 + }, + { + "epoch": 6.791450259688374, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 16999 + }, + { + "epoch": 6.791849780263684, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17000 + }, + { + "epoch": 6.792249300838993, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17001 + }, + { + "epoch": 6.792648821414303, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17002 + }, + { + "epoch": 6.793048341989612, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17003 + }, + { + "epoch": 6.793447862564922, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17004 + }, + { + "epoch": 6.793847383140232, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17005 + }, + { + "epoch": 6.794246903715541, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17006 + }, + { + "epoch": 6.794646424290851, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17007 + }, + { + "epoch": 6.79504594486616, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17008 + }, + { + "epoch": 6.79544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17009 + }, + { + "epoch": 6.79584498601678, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17010 + }, + { + "epoch": 6.79624450659209, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17011 + }, + { + "epoch": 6.7966440271674, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17012 + }, + { + "epoch": 6.7970435477427085, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17013 + }, + { + "epoch": 6.797443068318018, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17014 + }, + { + "epoch": 6.797842588893328, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17015 + }, + { + "epoch": 6.798242109468638, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17016 + }, + { + "epoch": 6.798641630043948, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17017 + }, + { + "epoch": 6.7990411506192565, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17018 + }, + { + "epoch": 6.799440671194566, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17019 + }, + { + "epoch": 6.799840191769876, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17020 + }, + { + "epoch": 6.800239712345186, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17021 + }, + { + "epoch": 6.800639232920496, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17022 + }, + { + "epoch": 6.801038753495805, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17023 + }, + { + "epoch": 6.801438274071114, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17024 + }, + { + "epoch": 6.801837794646424, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17025 + }, + { + "epoch": 6.802237315221734, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17026 + }, + { + "epoch": 6.802636835797044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17027 + }, + { + "epoch": 6.8030363563723535, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17028 + }, + { + "epoch": 6.803435876947663, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17029 + }, + { + "epoch": 6.803835397522972, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17030 + }, + { + "epoch": 6.804234918098282, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17031 + }, + { + "epoch": 6.804634438673592, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17032 + }, + { + "epoch": 6.8050339592489015, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17033 + }, + { + "epoch": 6.805433479824211, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17034 + }, + { + "epoch": 6.80583300039952, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17035 + }, + { + "epoch": 6.80623252097483, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17036 + }, + { + "epoch": 6.80663204155014, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17037 + }, + { + "epoch": 6.80703156212545, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17038 + }, + { + "epoch": 6.807431082700759, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17039 + }, + { + "epoch": 6.807830603276068, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17040 + }, + { + "epoch": 6.808230123851378, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17041 + }, + { + "epoch": 6.808629644426688, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17042 + }, + { + "epoch": 6.809029165001998, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17043 + }, + { + "epoch": 6.809428685577307, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17044 + }, + { + "epoch": 6.809828206152617, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17045 + }, + { + "epoch": 6.810227726727927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17046 + }, + { + "epoch": 6.810627247303236, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17047 + }, + { + "epoch": 6.811026767878546, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17048 + }, + { + "epoch": 6.8114262884538554, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17049 + }, + { + "epoch": 6.811825809029165, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17050 + }, + { + "epoch": 6.812225329604475, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17051 + }, + { + "epoch": 6.812624850179784, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17052 + }, + { + "epoch": 6.813024370755094, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17053 + }, + { + "epoch": 6.8134238913304035, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17054 + }, + { + "epoch": 6.813823411905713, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 17055 + }, + { + "epoch": 6.814222932481023, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17056 + }, + { + "epoch": 6.814622453056332, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 17057 + }, + { + "epoch": 6.815021973631642, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17058 + }, + { + "epoch": 6.8154214942069515, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17059 + }, + { + "epoch": 6.815821014782261, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17060 + }, + { + "epoch": 6.816220535357571, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 17061 + }, + { + "epoch": 6.816620055932881, + "grad_norm": 0.10693359375, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 17062 + }, + { + "epoch": 6.81701957650819, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17063 + }, + { + "epoch": 6.8174190970835, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17064 + }, + { + "epoch": 6.817818617658809, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17065 + }, + { + "epoch": 6.818218138234119, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17066 + }, + { + "epoch": 6.818617658809429, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17067 + }, + { + "epoch": 6.819017179384739, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 17068 + }, + { + "epoch": 6.819416699960048, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17069 + }, + { + "epoch": 6.819816220535357, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17070 + }, + { + "epoch": 6.820215741110667, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17071 + }, + { + "epoch": 6.820615261685977, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17072 + }, + { + "epoch": 6.821014782261287, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17073 + }, + { + "epoch": 6.821414302836596, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17074 + }, + { + "epoch": 6.8218138234119055, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3305, + "step": 17075 + }, + { + "epoch": 6.822213343987215, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17076 + }, + { + "epoch": 6.822612864562525, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 17077 + }, + { + "epoch": 6.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17078 + }, + { + "epoch": 6.823411905713145, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17079 + }, + { + "epoch": 6.8238114262884535, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17080 + }, + { + "epoch": 6.824210946863763, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17081 + }, + { + "epoch": 6.824610467439073, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17082 + }, + { + "epoch": 6.825009988014383, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17083 + }, + { + "epoch": 6.825409508589693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17084 + }, + { + "epoch": 6.825809029165002, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17085 + }, + { + "epoch": 6.826208549740311, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17086 + }, + { + "epoch": 6.826608070315621, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17087 + }, + { + "epoch": 6.827007590890931, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17088 + }, + { + "epoch": 6.827407111466241, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17089 + }, + { + "epoch": 6.8278066320415505, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17090 + }, + { + "epoch": 6.828206152616859, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17091 + }, + { + "epoch": 6.828605673192169, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 17092 + }, + { + "epoch": 6.829005193767479, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17093 + }, + { + "epoch": 6.829404714342789, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17094 + }, + { + "epoch": 6.8298042349180985, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17095 + }, + { + "epoch": 6.830203755493407, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17096 + }, + { + "epoch": 6.830603276068717, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17097 + }, + { + "epoch": 6.831002796644027, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17098 + }, + { + "epoch": 6.831402317219337, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17099 + }, + { + "epoch": 6.831801837794647, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17100 + }, + { + "epoch": 6.832201358369956, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17101 + }, + { + "epoch": 6.832600878945266, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17102 + }, + { + "epoch": 6.833000399520575, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17103 + }, + { + "epoch": 6.833399920095885, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17104 + }, + { + "epoch": 6.833799440671195, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17105 + }, + { + "epoch": 6.834198961246504, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17106 + }, + { + "epoch": 6.834598481821814, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17107 + }, + { + "epoch": 6.834998002397123, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17108 + }, + { + "epoch": 6.835397522972433, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17109 + }, + { + "epoch": 6.835797043547743, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17110 + }, + { + "epoch": 6.836196564123052, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17111 + }, + { + "epoch": 6.836596084698362, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17112 + }, + { + "epoch": 6.836995605273671, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17113 + }, + { + "epoch": 6.837395125848981, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17114 + }, + { + "epoch": 6.837794646424291, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17115 + }, + { + "epoch": 6.8381941669996005, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17116 + }, + { + "epoch": 6.83859368757491, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17117 + }, + { + "epoch": 6.83899320815022, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17118 + }, + { + "epoch": 6.83939272872553, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17119 + }, + { + "epoch": 6.839792249300839, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17120 + }, + { + "epoch": 6.8401917698761485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17121 + }, + { + "epoch": 6.840591290451458, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17122 + }, + { + "epoch": 6.840990811026768, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17123 + }, + { + "epoch": 6.841390331602078, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17124 + }, + { + "epoch": 6.841789852177387, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17125 + }, + { + "epoch": 6.842189372752697, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17126 + }, + { + "epoch": 6.842588893328006, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17127 + }, + { + "epoch": 6.842988413903316, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17128 + }, + { + "epoch": 6.843387934478626, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17129 + }, + { + "epoch": 6.843787455053935, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 17130 + }, + { + "epoch": 6.844186975629245, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17131 + }, + { + "epoch": 6.844586496204554, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17132 + }, + { + "epoch": 6.844986016779864, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17133 + }, + { + "epoch": 6.845385537355174, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17134 + }, + { + "epoch": 6.845785057930484, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17135 + }, + { + "epoch": 6.8461845785057935, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17136 + }, + { + "epoch": 6.846584099081102, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17137 + }, + { + "epoch": 6.846983619656412, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17138 + }, + { + "epoch": 6.847383140231722, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17139 + }, + { + "epoch": 6.847782660807032, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17140 + }, + { + "epoch": 6.848182181382342, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17141 + }, + { + "epoch": 6.8485817019576505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17142 + }, + { + "epoch": 6.84898122253296, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17143 + }, + { + "epoch": 6.84938074310827, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17144 + }, + { + "epoch": 6.84978026368358, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17145 + }, + { + "epoch": 6.85017978425889, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17146 + }, + { + "epoch": 6.8505793048341985, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17147 + }, + { + "epoch": 6.850978825409508, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17148 + }, + { + "epoch": 6.851378345984818, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17149 + }, + { + "epoch": 6.851777866560128, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17150 + }, + { + "epoch": 6.852177387135438, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3303, + "step": 17151 + }, + { + "epoch": 6.8525769077107475, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.2944, + "step": 17152 + }, + { + "epoch": 6.852976428286056, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17153 + }, + { + "epoch": 6.853375948861366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17154 + }, + { + "epoch": 6.853775469436676, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17155 + }, + { + "epoch": 6.854174990011986, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17156 + }, + { + "epoch": 6.8545745105872955, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17157 + }, + { + "epoch": 6.854974031162605, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17158 + }, + { + "epoch": 6.855373551737914, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17159 + }, + { + "epoch": 6.855773072313224, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17160 + }, + { + "epoch": 6.856172592888534, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17161 + }, + { + "epoch": 6.8565721134638435, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3351, + "step": 17162 + }, + { + "epoch": 6.856971634039153, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17163 + }, + { + "epoch": 6.857371154614462, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17164 + }, + { + "epoch": 6.857770675189772, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 17165 + }, + { + "epoch": 6.858170195765082, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17166 + }, + { + "epoch": 6.858569716340392, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17167 + }, + { + "epoch": 6.858969236915701, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17168 + }, + { + "epoch": 6.859368757491011, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17169 + }, + { + "epoch": 6.85976827806632, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17170 + }, + { + "epoch": 6.86016779864163, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17171 + }, + { + "epoch": 6.86056731921694, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17172 + }, + { + "epoch": 6.860966839792249, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17173 + }, + { + "epoch": 6.861366360367559, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17174 + }, + { + "epoch": 6.861765880942869, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17175 + }, + { + "epoch": 6.862165401518178, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17176 + }, + { + "epoch": 6.862564922093488, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17177 + }, + { + "epoch": 6.8629644426687975, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17178 + }, + { + "epoch": 6.863363963244107, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17179 + }, + { + "epoch": 6.863763483819417, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17180 + }, + { + "epoch": 6.864163004394726, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17181 + }, + { + "epoch": 6.864562524970036, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17182 + }, + { + "epoch": 6.8649620455453455, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17183 + }, + { + "epoch": 6.865361566120655, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17184 + }, + { + "epoch": 6.865761086695965, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17185 + }, + { + "epoch": 6.866160607271274, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17186 + }, + { + "epoch": 6.866560127846584, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17187 + }, + { + "epoch": 6.8669596484218935, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17188 + }, + { + "epoch": 6.867359168997203, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17189 + }, + { + "epoch": 6.867758689572513, + "grad_norm": 0.1005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17190 + }, + { + "epoch": 6.868158210147823, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17191 + }, + { + "epoch": 6.868557730723133, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17192 + }, + { + "epoch": 6.868957251298442, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17193 + }, + { + "epoch": 6.869356771873751, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17194 + }, + { + "epoch": 6.869756292449061, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17195 + }, + { + "epoch": 6.870155813024371, + "grad_norm": 0.1484375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17196 + }, + { + "epoch": 6.870555333599681, + "grad_norm": 0.197265625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17197 + }, + { + "epoch": 6.87095485417499, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17198 + }, + { + "epoch": 6.871354374750299, + "grad_norm": 0.240234375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17199 + }, + { + "epoch": 6.871753895325609, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17200 + }, + { + "epoch": 6.872153415900919, + "grad_norm": 0.234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17201 + }, + { + "epoch": 6.872552936476229, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3354, + "step": 17202 + }, + { + "epoch": 6.872952457051538, + "grad_norm": 0.2080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17203 + }, + { + "epoch": 6.8733519776268475, + "grad_norm": 0.1494140625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17204 + }, + { + "epoch": 6.873751498202157, + "grad_norm": 0.1943359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17205 + }, + { + "epoch": 6.874151018777467, + "grad_norm": 0.1728515625, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17206 + }, + { + "epoch": 6.874550539352777, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 17207 + }, + { + "epoch": 6.874950059928087, + "grad_norm": 0.1416015625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17208 + }, + { + "epoch": 6.875349580503396, + "grad_norm": 0.11083984375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17209 + }, + { + "epoch": 6.875749101078705, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17210 + }, + { + "epoch": 6.876148621654015, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17211 + }, + { + "epoch": 6.876548142229325, + "grad_norm": 0.115234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17212 + }, + { + "epoch": 6.876947662804635, + "grad_norm": 0.1708984375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17213 + }, + { + "epoch": 6.877347183379944, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17214 + }, + { + "epoch": 6.877746703955253, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17215 + }, + { + "epoch": 6.878146224530563, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17216 + }, + { + "epoch": 6.878545745105873, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17217 + }, + { + "epoch": 6.878945265681183, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17218 + }, + { + "epoch": 6.8793447862564925, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17219 + }, + { + "epoch": 6.879744306831801, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17220 + }, + { + "epoch": 6.880143827407111, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17221 + }, + { + "epoch": 6.880543347982421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17222 + }, + { + "epoch": 6.880942868557731, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17223 + }, + { + "epoch": 6.8813423891330405, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 17224 + }, + { + "epoch": 6.88174190970835, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17225 + }, + { + "epoch": 6.88214143028366, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17226 + }, + { + "epoch": 6.882540950858969, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17227 + }, + { + "epoch": 6.882940471434279, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17228 + }, + { + "epoch": 6.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17229 + }, + { + "epoch": 6.883739512584898, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17230 + }, + { + "epoch": 6.884139033160208, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17231 + }, + { + "epoch": 6.884538553735517, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17232 + }, + { + "epoch": 6.884938074310827, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17233 + }, + { + "epoch": 6.885337594886137, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17234 + }, + { + "epoch": 6.885737115461446, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17235 + }, + { + "epoch": 6.886136636036756, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17236 + }, + { + "epoch": 6.886536156612065, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17237 + }, + { + "epoch": 6.886935677187375, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17238 + }, + { + "epoch": 6.887335197762685, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17239 + }, + { + "epoch": 6.887734718337994, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 17240 + }, + { + "epoch": 6.888134238913304, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17241 + }, + { + "epoch": 6.888533759488614, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17242 + }, + { + "epoch": 6.888933280063923, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17243 + }, + { + "epoch": 6.889332800639233, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17244 + }, + { + "epoch": 6.8897323212145425, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17245 + }, + { + "epoch": 6.890131841789852, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17246 + }, + { + "epoch": 6.890531362365162, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17247 + }, + { + "epoch": 6.890930882940472, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17248 + }, + { + "epoch": 6.891330403515781, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17249 + }, + { + "epoch": 6.8917299240910905, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17250 + }, + { + "epoch": 6.8921294446664, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17251 + }, + { + "epoch": 6.89252896524171, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 17252 + }, + { + "epoch": 6.89292848581702, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17253 + }, + { + "epoch": 6.893328006392329, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17254 + }, + { + "epoch": 6.893727526967639, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17255 + }, + { + "epoch": 6.894127047542948, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17256 + }, + { + "epoch": 6.894526568118258, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17257 + }, + { + "epoch": 6.894926088693568, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17258 + }, + { + "epoch": 6.895325609268878, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17259 + }, + { + "epoch": 6.895725129844187, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17260 + }, + { + "epoch": 6.896124650419496, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17261 + }, + { + "epoch": 6.896524170994806, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17262 + }, + { + "epoch": 6.896923691570116, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17263 + }, + { + "epoch": 6.897323212145426, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17264 + }, + { + "epoch": 6.8977227327207356, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17265 + }, + { + "epoch": 6.8981222532960444, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17266 + }, + { + "epoch": 6.898521773871354, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17267 + }, + { + "epoch": 6.898921294446664, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17268 + }, + { + "epoch": 6.899320815021974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 17269 + }, + { + "epoch": 6.899720335597284, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17270 + }, + { + "epoch": 6.9001198561725925, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17271 + }, + { + "epoch": 6.900519376747902, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17272 + }, + { + "epoch": 6.900918897323212, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17273 + }, + { + "epoch": 6.901318417898522, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17274 + }, + { + "epoch": 6.901717938473832, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17275 + }, + { + "epoch": 6.9021174590491405, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.2994, + "step": 17276 + }, + { + "epoch": 6.90251697962445, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17277 + }, + { + "epoch": 6.90291650019976, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17278 + }, + { + "epoch": 6.90331602077507, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17279 + }, + { + "epoch": 6.90371554135038, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17280 + }, + { + "epoch": 6.9041150619256895, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17281 + }, + { + "epoch": 6.904514582500999, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17282 + }, + { + "epoch": 6.904914103076308, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17283 + }, + { + "epoch": 6.905313623651618, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17284 + }, + { + "epoch": 6.905713144226928, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17285 + }, + { + "epoch": 6.9061126648022375, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17286 + }, + { + "epoch": 6.906512185377547, + "grad_norm": 0.048583984375, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17287 + }, + { + "epoch": 6.906911705952856, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17288 + }, + { + "epoch": 6.907311226528166, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17289 + }, + { + "epoch": 6.907710747103476, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17290 + }, + { + "epoch": 6.9081102676787856, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17291 + }, + { + "epoch": 6.908509788254095, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17292 + }, + { + "epoch": 6.908909308829404, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17293 + }, + { + "epoch": 6.909308829404714, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17294 + }, + { + "epoch": 6.909708349980024, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17295 + }, + { + "epoch": 6.910107870555334, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17296 + }, + { + "epoch": 6.910507391130643, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17297 + }, + { + "epoch": 6.910906911705953, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17298 + }, + { + "epoch": 6.911306432281263, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17299 + }, + { + "epoch": 6.911705952856572, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17300 + }, + { + "epoch": 6.912105473431882, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17301 + }, + { + "epoch": 6.912504994007191, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17302 + }, + { + "epoch": 6.912904514582501, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17303 + }, + { + "epoch": 6.913304035157811, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17304 + }, + { + "epoch": 6.91370355573312, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17305 + }, + { + "epoch": 6.91410307630843, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17306 + }, + { + "epoch": 6.9145025968837395, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17307 + }, + { + "epoch": 6.914902117459049, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 17308 + }, + { + "epoch": 6.915301638034359, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17309 + }, + { + "epoch": 6.915701158609668, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17310 + }, + { + "epoch": 6.916100679184978, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17311 + }, + { + "epoch": 6.9165001997602875, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17312 + }, + { + "epoch": 6.916899720335597, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17313 + }, + { + "epoch": 6.917299240910907, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17314 + }, + { + "epoch": 6.917698761486217, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 17315 + }, + { + "epoch": 6.918098282061527, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17316 + }, + { + "epoch": 6.918497802636836, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17317 + }, + { + "epoch": 6.918897323212145, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17318 + }, + { + "epoch": 6.919296843787455, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17319 + }, + { + "epoch": 6.919696364362765, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 17320 + }, + { + "epoch": 6.920095884938075, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17321 + }, + { + "epoch": 6.920495405513384, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17322 + }, + { + "epoch": 6.920894926088693, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17323 + }, + { + "epoch": 6.921294446664003, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17324 + }, + { + "epoch": 6.921693967239313, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3306, + "step": 17325 + }, + { + "epoch": 6.922093487814623, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17326 + }, + { + "epoch": 6.922493008389932, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17327 + }, + { + "epoch": 6.922892528965241, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17328 + }, + { + "epoch": 6.923292049540551, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17329 + }, + { + "epoch": 6.923691570115861, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.2958, + "step": 17330 + }, + { + "epoch": 6.924091090691171, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17331 + }, + { + "epoch": 6.924490611266481, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17332 + }, + { + "epoch": 6.9248901318417895, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17333 + }, + { + "epoch": 6.925289652417099, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17334 + }, + { + "epoch": 6.925689172992409, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17335 + }, + { + "epoch": 6.926088693567719, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17336 + }, + { + "epoch": 6.926488214143029, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 17337 + }, + { + "epoch": 6.926887734718338, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17338 + }, + { + "epoch": 6.927287255293647, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17339 + }, + { + "epoch": 6.927686775868957, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3336, + "step": 17340 + }, + { + "epoch": 6.928086296444267, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 17341 + }, + { + "epoch": 6.928485817019577, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17342 + }, + { + "epoch": 6.9288853375948865, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 17343 + }, + { + "epoch": 6.929284858170195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17344 + }, + { + "epoch": 6.929684378745505, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17345 + }, + { + "epoch": 6.930083899320815, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 17346 + }, + { + "epoch": 6.930483419896125, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17347 + }, + { + "epoch": 6.9308829404714345, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 17348 + }, + { + "epoch": 6.931282461046744, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17349 + }, + { + "epoch": 6.931681981622053, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17350 + }, + { + "epoch": 6.932081502197363, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 17351 + }, + { + "epoch": 6.932481022772673, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17352 + }, + { + "epoch": 6.9328805433479825, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17353 + }, + { + "epoch": 6.933280063923292, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17354 + }, + { + "epoch": 6.933679584498602, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17355 + }, + { + "epoch": 6.934079105073911, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17356 + }, + { + "epoch": 6.934478625649221, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17357 + }, + { + "epoch": 6.934878146224531, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17358 + }, + { + "epoch": 6.93527766679984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17359 + }, + { + "epoch": 6.93567718737515, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17360 + }, + { + "epoch": 6.936076707950459, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17361 + }, + { + "epoch": 6.936476228525769, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3372, + "step": 17362 + }, + { + "epoch": 6.936875749101079, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17363 + }, + { + "epoch": 6.937275269676388, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17364 + }, + { + "epoch": 6.937674790251698, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17365 + }, + { + "epoch": 6.938074310827007, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17366 + }, + { + "epoch": 6.938473831402317, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17367 + }, + { + "epoch": 6.938873351977627, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3308, + "step": 17368 + }, + { + "epoch": 6.9392728725529365, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17369 + }, + { + "epoch": 6.939672393128246, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17370 + }, + { + "epoch": 6.940071913703556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17371 + }, + { + "epoch": 6.940471434278866, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17372 + }, + { + "epoch": 6.940870954854175, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17373 + }, + { + "epoch": 6.9412704754294845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17374 + }, + { + "epoch": 6.941669996004794, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17375 + }, + { + "epoch": 6.942069516580104, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17376 + }, + { + "epoch": 6.942469037155414, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17377 + }, + { + "epoch": 6.942868557730723, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17378 + }, + { + "epoch": 6.9432680783060325, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17379 + }, + { + "epoch": 6.943667598881342, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 17380 + }, + { + "epoch": 6.944067119456652, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17381 + }, + { + "epoch": 6.944466640031962, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 17382 + }, + { + "epoch": 6.944866160607271, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17383 + }, + { + "epoch": 6.945265681182581, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 17384 + }, + { + "epoch": 6.94566520175789, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17385 + }, + { + "epoch": 6.9460647223332, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17386 + }, + { + "epoch": 6.94646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17387 + }, + { + "epoch": 6.94686376348382, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 17388 + }, + { + "epoch": 6.9472632840591295, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17389 + }, + { + "epoch": 6.947662804634438, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17390 + }, + { + "epoch": 6.948062325209748, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 17391 + }, + { + "epoch": 6.948461845785058, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17392 + }, + { + "epoch": 6.948861366360368, + "grad_norm": 0.0458984375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17393 + }, + { + "epoch": 6.949260886935678, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17394 + }, + { + "epoch": 6.9496604075109865, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17395 + }, + { + "epoch": 6.950059928086296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 17396 + }, + { + "epoch": 6.950459448661606, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 17397 + }, + { + "epoch": 6.950858969236916, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17398 + }, + { + "epoch": 6.951258489812226, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17399 + }, + { + "epoch": 6.9516580103875345, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17400 + }, + { + "epoch": 6.952057530962844, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17401 + }, + { + "epoch": 6.952457051538154, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17402 + }, + { + "epoch": 6.952856572113464, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17403 + }, + { + "epoch": 6.953256092688774, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17404 + }, + { + "epoch": 6.953655613264083, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17405 + }, + { + "epoch": 6.954055133839393, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17406 + }, + { + "epoch": 6.954454654414702, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17407 + }, + { + "epoch": 6.954854174990012, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17408 + }, + { + "epoch": 6.955253695565322, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17409 + }, + { + "epoch": 6.9556532161406315, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17410 + }, + { + "epoch": 6.956052736715941, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 17411 + }, + { + "epoch": 6.95645225729125, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17412 + }, + { + "epoch": 6.95685177786656, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17413 + }, + { + "epoch": 6.95725129844187, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17414 + }, + { + "epoch": 6.9576508190171795, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17415 + }, + { + "epoch": 6.958050339592489, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17416 + }, + { + "epoch": 6.958449860167798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17417 + }, + { + "epoch": 6.958849380743108, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17418 + }, + { + "epoch": 6.959248901318418, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17419 + }, + { + "epoch": 6.959648421893728, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 17420 + }, + { + "epoch": 6.960047942469037, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17421 + }, + { + "epoch": 6.960447463044347, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 17422 + }, + { + "epoch": 6.960846983619656, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17423 + }, + { + "epoch": 6.961246504194966, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17424 + }, + { + "epoch": 6.961646024770276, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17425 + }, + { + "epoch": 6.962045545345585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17426 + }, + { + "epoch": 6.962445065920895, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17427 + }, + { + "epoch": 6.962844586496205, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17428 + }, + { + "epoch": 6.963244107071514, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17429 + }, + { + "epoch": 6.963643627646824, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17430 + }, + { + "epoch": 6.964043148222133, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17431 + }, + { + "epoch": 6.964442668797443, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17432 + }, + { + "epoch": 6.964842189372753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17433 + }, + { + "epoch": 6.965241709948062, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17434 + }, + { + "epoch": 6.965641230523372, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17435 + }, + { + "epoch": 6.9660407510986815, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17436 + }, + { + "epoch": 6.966440271673991, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17437 + }, + { + "epoch": 6.966839792249301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17438 + }, + { + "epoch": 6.967239312824611, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 17439 + }, + { + "epoch": 6.96763883339992, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17440 + }, + { + "epoch": 6.9680383539752295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17441 + }, + { + "epoch": 6.968437874550539, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17442 + }, + { + "epoch": 6.968837395125849, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17443 + }, + { + "epoch": 6.969236915701159, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17444 + }, + { + "epoch": 6.969636436276469, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17445 + }, + { + "epoch": 6.970035956851778, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17446 + }, + { + "epoch": 6.970435477427087, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 17447 + }, + { + "epoch": 6.970834998002397, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17448 + }, + { + "epoch": 6.971234518577707, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17449 + }, + { + "epoch": 6.971634039153017, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17450 + }, + { + "epoch": 6.972033559728326, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17451 + }, + { + "epoch": 6.972433080303635, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17452 + }, + { + "epoch": 6.972832600878945, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 17453 + }, + { + "epoch": 6.973232121454255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17454 + }, + { + "epoch": 6.973631642029565, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17455 + }, + { + "epoch": 6.974031162604874, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17456 + }, + { + "epoch": 6.974430683180183, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 17457 + }, + { + "epoch": 6.974830203755493, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17458 + }, + { + "epoch": 6.975229724330803, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17459 + }, + { + "epoch": 6.975629244906113, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 17460 + }, + { + "epoch": 6.976028765481423, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17461 + }, + { + "epoch": 6.976428286056732, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 17462 + }, + { + "epoch": 6.976827806632041, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17463 + }, + { + "epoch": 6.977227327207351, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17464 + }, + { + "epoch": 6.977626847782661, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17465 + }, + { + "epoch": 6.978026368357971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17466 + }, + { + "epoch": 6.97842588893328, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17467 + }, + { + "epoch": 6.978825409508589, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17468 + }, + { + "epoch": 6.979224930083899, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17469 + }, + { + "epoch": 6.979624450659209, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17470 + }, + { + "epoch": 6.980023971234519, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17471 + }, + { + "epoch": 6.9804234918098285, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17472 + }, + { + "epoch": 6.980823012385137, + "grad_norm": 0.10546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17473 + }, + { + "epoch": 6.981222532960447, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17474 + }, + { + "epoch": 6.981622053535757, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 17475 + }, + { + "epoch": 6.982021574111067, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17476 + }, + { + "epoch": 6.9824210946863765, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17477 + }, + { + "epoch": 6.982820615261686, + "grad_norm": 0.1318359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17478 + }, + { + "epoch": 6.983220135836996, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 17479 + }, + { + "epoch": 6.983619656412305, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17480 + }, + { + "epoch": 6.984019176987615, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17481 + }, + { + "epoch": 6.9844186975629245, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17482 + }, + { + "epoch": 6.984818218138234, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17483 + }, + { + "epoch": 6.985217738713544, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17484 + }, + { + "epoch": 6.985617259288853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3322, + "step": 17485 + }, + { + "epoch": 6.986016779864163, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17486 + }, + { + "epoch": 6.986416300439473, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17487 + }, + { + "epoch": 6.986815821014782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17488 + }, + { + "epoch": 6.987215341590092, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17489 + }, + { + "epoch": 6.987614862165401, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17490 + }, + { + "epoch": 6.988014382740711, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17491 + }, + { + "epoch": 6.988413903316021, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17492 + }, + { + "epoch": 6.98881342389133, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17493 + }, + { + "epoch": 6.98921294446664, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17494 + }, + { + "epoch": 6.98961246504195, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17495 + }, + { + "epoch": 6.99001198561726, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17496 + }, + { + "epoch": 6.990411506192569, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17497 + }, + { + "epoch": 6.9908110267678785, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17498 + }, + { + "epoch": 6.991210547343188, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17499 + }, + { + "epoch": 6.991610067918498, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17500 + }, + { + "epoch": 6.992009588493808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17501 + }, + { + "epoch": 6.992409109069117, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 17502 + }, + { + "epoch": 6.9928086296444265, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17503 + }, + { + "epoch": 6.993208150219736, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17504 + }, + { + "epoch": 6.993607670795046, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3006, + "step": 17505 + }, + { + "epoch": 6.994007191370356, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17506 + }, + { + "epoch": 6.994406711945665, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17507 + }, + { + "epoch": 6.9948062325209746, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17508 + }, + { + "epoch": 6.995205753096284, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17509 + }, + { + "epoch": 6.995605273671594, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 17510 + }, + { + "epoch": 6.996004794246904, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17511 + }, + { + "epoch": 6.996404314822214, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17512 + }, + { + "epoch": 6.996803835397523, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3299, + "step": 17513 + }, + { + "epoch": 6.997203355972832, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17514 + }, + { + "epoch": 6.997602876548142, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17515 + }, + { + "epoch": 6.998002397123452, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17516 + }, + { + "epoch": 6.998401917698762, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3327, + "step": 17517 + }, + { + "epoch": 6.9988014382740715, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17518 + }, + { + "epoch": 6.99920095884938, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 17519 + }, + { + "epoch": 6.99960047942469, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17520 + }, + { + "epoch": 7.0, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17521 + }, + { + "epoch": 7.00039952057531, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17522 + }, + { + "epoch": 7.00079904115062, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17523 + }, + { + "epoch": 7.0011985617259285, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17524 + }, + { + "epoch": 7.001598082301238, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17525 + }, + { + "epoch": 7.001997602876548, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3304, + "step": 17526 + }, + { + "epoch": 7.002397123451858, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17527 + }, + { + "epoch": 7.002796644027168, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17528 + }, + { + "epoch": 7.003196164602477, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 17529 + }, + { + "epoch": 7.003595685177786, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 17530 + }, + { + "epoch": 7.003995205753096, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17531 + }, + { + "epoch": 7.004394726328406, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17532 + }, + { + "epoch": 7.004794246903716, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17533 + }, + { + "epoch": 7.0051937674790254, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 17534 + }, + { + "epoch": 7.005593288054335, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 17535 + }, + { + "epoch": 7.005992808629644, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17536 + }, + { + "epoch": 7.006392329204954, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17537 + }, + { + "epoch": 7.006791849780264, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17538 + }, + { + "epoch": 7.0071913703555735, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17539 + }, + { + "epoch": 7.007590890930883, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17540 + }, + { + "epoch": 7.007990411506192, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17541 + }, + { + "epoch": 7.008389932081502, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17542 + }, + { + "epoch": 7.008789452656812, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 17543 + }, + { + "epoch": 7.0091889732321215, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17544 + }, + { + "epoch": 7.009588493807431, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17545 + }, + { + "epoch": 7.009988014382741, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17546 + }, + { + "epoch": 7.01038753495805, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17547 + }, + { + "epoch": 7.01078705553336, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17548 + }, + { + "epoch": 7.01118657610867, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 17549 + }, + { + "epoch": 7.011586096683979, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17550 + }, + { + "epoch": 7.011985617259289, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17551 + }, + { + "epoch": 7.012385137834598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17552 + }, + { + "epoch": 7.012784658409908, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17553 + }, + { + "epoch": 7.013184178985218, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17554 + }, + { + "epoch": 7.013583699560527, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17555 + }, + { + "epoch": 7.013983220135837, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17556 + }, + { + "epoch": 7.014382740711147, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17557 + }, + { + "epoch": 7.014782261286456, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17558 + }, + { + "epoch": 7.015181781861766, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17559 + }, + { + "epoch": 7.0155813024370755, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 17560 + }, + { + "epoch": 7.015980823012385, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17561 + }, + { + "epoch": 7.016380343587695, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17562 + }, + { + "epoch": 7.016779864163005, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17563 + }, + { + "epoch": 7.017179384738314, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 17564 + }, + { + "epoch": 7.0175789053136235, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 17565 + }, + { + "epoch": 7.017978425888933, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17566 + }, + { + "epoch": 7.018377946464243, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3311, + "step": 17567 + }, + { + "epoch": 7.018777467039553, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17568 + }, + { + "epoch": 7.019176987614862, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 17569 + }, + { + "epoch": 7.0195765081901715, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17570 + }, + { + "epoch": 7.019976028765481, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17571 + }, + { + "epoch": 7.020375549340791, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17572 + }, + { + "epoch": 7.020775069916101, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17573 + }, + { + "epoch": 7.021174590491411, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17574 + }, + { + "epoch": 7.02157411106672, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 17575 + }, + { + "epoch": 7.021973631642029, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17576 + }, + { + "epoch": 7.022373152217339, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17577 + }, + { + "epoch": 7.022772672792649, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17578 + }, + { + "epoch": 7.023172193367959, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17579 + }, + { + "epoch": 7.0235717139432685, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 17580 + }, + { + "epoch": 7.023971234518577, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17581 + }, + { + "epoch": 7.024370755093887, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17582 + }, + { + "epoch": 7.024770275669197, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 17583 + }, + { + "epoch": 7.025169796244507, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17584 + }, + { + "epoch": 7.025569316819817, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17585 + }, + { + "epoch": 7.0259688373951255, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17586 + }, + { + "epoch": 7.026368357970435, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17587 + }, + { + "epoch": 7.026767878545745, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17588 + }, + { + "epoch": 7.027167399121055, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17589 + }, + { + "epoch": 7.027566919696365, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 17590 + }, + { + "epoch": 7.027966440271674, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 17591 + }, + { + "epoch": 7.028365960846983, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 17592 + }, + { + "epoch": 7.028765481422293, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17593 + }, + { + "epoch": 7.029165001997603, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17594 + }, + { + "epoch": 7.029564522572913, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17595 + }, + { + "epoch": 7.029964043148222, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 17596 + }, + { + "epoch": 7.030363563723531, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 17597 + }, + { + "epoch": 7.030763084298841, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 17598 + }, + { + "epoch": 7.031162604874151, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17599 + }, + { + "epoch": 7.031562125449461, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 17600 + }, + { + "epoch": 7.0319616460247705, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 17601 + }, + { + "epoch": 7.03236116660008, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17602 + }, + { + "epoch": 7.032760687175389, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17603 + }, + { + "epoch": 7.033160207750699, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17604 + }, + { + "epoch": 7.033559728326009, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17605 + }, + { + "epoch": 7.0339592489013185, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 17606 + }, + { + "epoch": 7.034358769476628, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17607 + }, + { + "epoch": 7.034758290051938, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 17608 + }, + { + "epoch": 7.035157810627247, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17609 + }, + { + "epoch": 7.035557331202557, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17610 + }, + { + "epoch": 7.035956851777867, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17611 + }, + { + "epoch": 7.036356372353176, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17612 + }, + { + "epoch": 7.036755892928486, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17613 + }, + { + "epoch": 7.037155413503795, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17614 + }, + { + "epoch": 7.037554934079105, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 17615 + }, + { + "epoch": 7.037954454654415, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 17616 + }, + { + "epoch": 7.038353975229724, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 17617 + }, + { + "epoch": 7.038753495805034, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17618 + }, + { + "epoch": 7.039153016380344, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17619 + }, + { + "epoch": 7.039552536955653, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17620 + }, + { + "epoch": 7.039952057530963, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17621 + }, + { + "epoch": 7.040351578106272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 17622 + }, + { + "epoch": 7.040751098681582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17623 + }, + { + "epoch": 7.041150619256892, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17624 + }, + { + "epoch": 7.041550139832202, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 17625 + }, + { + "epoch": 7.041949660407511, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17626 + }, + { + "epoch": 7.0423491809828205, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17627 + }, + { + "epoch": 7.04274870155813, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17628 + }, + { + "epoch": 7.04314822213344, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17629 + }, + { + "epoch": 7.04354774270875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17630 + }, + { + "epoch": 7.043947263284059, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17631 + }, + { + "epoch": 7.0443467838593685, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17632 + }, + { + "epoch": 7.044746304434678, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17633 + }, + { + "epoch": 7.045145825009988, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17634 + }, + { + "epoch": 7.045545345585298, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17635 + }, + { + "epoch": 7.045944866160608, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17636 + }, + { + "epoch": 7.046344386735917, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.2991, + "step": 17637 + }, + { + "epoch": 7.046743907311226, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17638 + }, + { + "epoch": 7.047143427886536, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17639 + }, + { + "epoch": 7.047542948461846, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17640 + }, + { + "epoch": 7.047942469037156, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17641 + }, + { + "epoch": 7.048341989612465, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 17642 + }, + { + "epoch": 7.048741510187774, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17643 + }, + { + "epoch": 7.049141030763084, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17644 + }, + { + "epoch": 7.049540551338394, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17645 + }, + { + "epoch": 7.049940071913704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17646 + }, + { + "epoch": 7.0503395924890135, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17647 + }, + { + "epoch": 7.050739113064322, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17648 + }, + { + "epoch": 7.051138633639632, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17649 + }, + { + "epoch": 7.051538154214942, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 17650 + }, + { + "epoch": 7.051937674790252, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17651 + }, + { + "epoch": 7.052337195365562, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17652 + }, + { + "epoch": 7.052736715940871, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17653 + }, + { + "epoch": 7.05313623651618, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17654 + }, + { + "epoch": 7.05353575709149, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17655 + }, + { + "epoch": 7.0539352776668, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17656 + }, + { + "epoch": 7.05433479824211, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 17657 + }, + { + "epoch": 7.054734318817419, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17658 + }, + { + "epoch": 7.055133839392728, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17659 + }, + { + "epoch": 7.055533359968038, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 17660 + }, + { + "epoch": 7.055932880543348, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 17661 + }, + { + "epoch": 7.056332401118658, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17662 + }, + { + "epoch": 7.0567319216939675, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17663 + }, + { + "epoch": 7.057131442269277, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17664 + }, + { + "epoch": 7.057530962844586, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 17665 + }, + { + "epoch": 7.057930483419896, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17666 + }, + { + "epoch": 7.058330003995206, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17667 + }, + { + "epoch": 7.0587295245705155, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 17668 + }, + { + "epoch": 7.059129045145825, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17669 + }, + { + "epoch": 7.059528565721135, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17670 + }, + { + "epoch": 7.059928086296444, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17671 + }, + { + "epoch": 7.060327606871754, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17672 + }, + { + "epoch": 7.0607271274470635, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 17673 + }, + { + "epoch": 7.061126648022373, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17674 + }, + { + "epoch": 7.061526168597683, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17675 + }, + { + "epoch": 7.061925689172992, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17676 + }, + { + "epoch": 7.062325209748302, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17677 + }, + { + "epoch": 7.062724730323612, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17678 + }, + { + "epoch": 7.063124250898921, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17679 + }, + { + "epoch": 7.063523771474231, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17680 + }, + { + "epoch": 7.063923292049541, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17681 + }, + { + "epoch": 7.06432281262485, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 17682 + }, + { + "epoch": 7.06472233320016, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 17683 + }, + { + "epoch": 7.065121853775469, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17684 + }, + { + "epoch": 7.065521374350779, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17685 + }, + { + "epoch": 7.065920894926089, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17686 + }, + { + "epoch": 7.066320415501398, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17687 + }, + { + "epoch": 7.066719936076708, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17688 + }, + { + "epoch": 7.0671194566520175, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 17689 + }, + { + "epoch": 7.067518977227327, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17690 + }, + { + "epoch": 7.067918497802637, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 17691 + }, + { + "epoch": 7.068318018377947, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17692 + }, + { + "epoch": 7.068717538953256, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17693 + }, + { + "epoch": 7.0691170595285655, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17694 + }, + { + "epoch": 7.069516580103875, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17695 + }, + { + "epoch": 7.069916100679185, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17696 + }, + { + "epoch": 7.070315621254495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 17697 + }, + { + "epoch": 7.070715141829805, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17698 + }, + { + "epoch": 7.0711146624051135, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3286, + "step": 17699 + }, + { + "epoch": 7.071514182980423, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17700 + }, + { + "epoch": 7.071913703555733, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17701 + }, + { + "epoch": 7.072313224131043, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 17702 + }, + { + "epoch": 7.072712744706353, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 17703 + }, + { + "epoch": 7.073112265281662, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17704 + }, + { + "epoch": 7.073511785856971, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17705 + }, + { + "epoch": 7.073911306432281, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 17706 + }, + { + "epoch": 7.074310827007591, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17707 + }, + { + "epoch": 7.074710347582901, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17708 + }, + { + "epoch": 7.0751098681582105, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 17709 + }, + { + "epoch": 7.075509388733519, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17710 + }, + { + "epoch": 7.075908909308829, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 17711 + }, + { + "epoch": 7.076308429884139, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17712 + }, + { + "epoch": 7.076707950459449, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17713 + }, + { + "epoch": 7.077107471034759, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17714 + }, + { + "epoch": 7.077506991610068, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 17715 + }, + { + "epoch": 7.077906512185377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17716 + }, + { + "epoch": 7.078306032760687, + "grad_norm": 0.11865234375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17717 + }, + { + "epoch": 7.078705553335997, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17718 + }, + { + "epoch": 7.079105073911307, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 17719 + }, + { + "epoch": 7.079504594486616, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 17720 + }, + { + "epoch": 7.079904115061925, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17721 + }, + { + "epoch": 7.080303635637235, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17722 + }, + { + "epoch": 7.080703156212545, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17723 + }, + { + "epoch": 7.081102676787855, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 17724 + }, + { + "epoch": 7.081502197363164, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17725 + }, + { + "epoch": 7.081901717938474, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17726 + }, + { + "epoch": 7.082301238513783, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17727 + }, + { + "epoch": 7.082700759089093, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17728 + }, + { + "epoch": 7.083100279664403, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 17729 + }, + { + "epoch": 7.0834998002397125, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 17730 + }, + { + "epoch": 7.083899320815022, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 17731 + }, + { + "epoch": 7.084298841390331, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17732 + }, + { + "epoch": 7.084698361965641, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17733 + }, + { + "epoch": 7.085097882540951, + "grad_norm": 0.126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17734 + }, + { + "epoch": 7.0854974031162605, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17735 + }, + { + "epoch": 7.08589692369157, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17736 + }, + { + "epoch": 7.08629644426688, + "grad_norm": 0.1279296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17737 + }, + { + "epoch": 7.086695964842189, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17738 + }, + { + "epoch": 7.087095485417499, + "grad_norm": 0.13671875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17739 + }, + { + "epoch": 7.087495005992809, + "grad_norm": 0.12158203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 17740 + }, + { + "epoch": 7.087894526568118, + "grad_norm": 0.138671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 17741 + }, + { + "epoch": 7.088294047143428, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 17742 + }, + { + "epoch": 7.088693567718738, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17743 + }, + { + "epoch": 7.089093088294047, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17744 + }, + { + "epoch": 7.089492608869357, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17745 + }, + { + "epoch": 7.089892129444666, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17746 + }, + { + "epoch": 7.090291650019976, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17747 + }, + { + "epoch": 7.090691170595286, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 17748 + }, + { + "epoch": 7.091090691170595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3364, + "step": 17749 + }, + { + "epoch": 7.091490211745905, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17750 + }, + { + "epoch": 7.0918897323212144, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17751 + }, + { + "epoch": 7.092289252896524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 17752 + }, + { + "epoch": 7.092688773471834, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17753 + }, + { + "epoch": 7.093088294047144, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 17754 + }, + { + "epoch": 7.093487814622453, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17755 + }, + { + "epoch": 7.0938873351977625, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 17756 + }, + { + "epoch": 7.094286855773072, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17757 + }, + { + "epoch": 7.094686376348382, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17758 + }, + { + "epoch": 7.095085896923692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17759 + }, + { + "epoch": 7.095485417499002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17760 + }, + { + "epoch": 7.0958849380743105, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17761 + }, + { + "epoch": 7.09628445864962, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17762 + }, + { + "epoch": 7.09668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 17763 + }, + { + "epoch": 7.09708349980024, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 17764 + }, + { + "epoch": 7.09748302037555, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 17765 + }, + { + "epoch": 7.097882540950859, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17766 + }, + { + "epoch": 7.098282061526168, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17767 + }, + { + "epoch": 7.098681582101478, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17768 + }, + { + "epoch": 7.099081102676788, + "grad_norm": 0.12353515625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17769 + }, + { + "epoch": 7.099480623252098, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17770 + }, + { + "epoch": 7.0998801438274075, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 17771 + }, + { + "epoch": 7.100279664402716, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17772 + }, + { + "epoch": 7.100679184978026, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17773 + }, + { + "epoch": 7.101078705553336, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17774 + }, + { + "epoch": 7.101478226128646, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 17775 + }, + { + "epoch": 7.1018777467039556, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 17776 + }, + { + "epoch": 7.1022772672792644, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 17777 + }, + { + "epoch": 7.102676787854574, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17778 + }, + { + "epoch": 7.103076308429884, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17779 + }, + { + "epoch": 7.103475829005194, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17780 + }, + { + "epoch": 7.103875349580504, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17781 + }, + { + "epoch": 7.104274870155813, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17782 + }, + { + "epoch": 7.104674390731122, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17783 + }, + { + "epoch": 7.105073911306432, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 17784 + }, + { + "epoch": 7.105473431881742, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 17785 + }, + { + "epoch": 7.105872952457052, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 17786 + }, + { + "epoch": 7.106272473032361, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17787 + }, + { + "epoch": 7.106671993607671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17788 + }, + { + "epoch": 7.10707151418298, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17789 + }, + { + "epoch": 7.10747103475829, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17790 + }, + { + "epoch": 7.1078705553336, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17791 + }, + { + "epoch": 7.1082700759089095, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17792 + }, + { + "epoch": 7.108669596484219, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17793 + }, + { + "epoch": 7.109069117059528, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17794 + }, + { + "epoch": 7.109468637634838, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 17795 + }, + { + "epoch": 7.109868158210148, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17796 + }, + { + "epoch": 7.1102676787854575, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17797 + }, + { + "epoch": 7.110667199360767, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17798 + }, + { + "epoch": 7.111066719936077, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17799 + }, + { + "epoch": 7.111466240511386, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17800 + }, + { + "epoch": 7.111865761086696, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17801 + }, + { + "epoch": 7.112265281662006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 17802 + }, + { + "epoch": 7.112664802237315, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17803 + }, + { + "epoch": 7.113064322812625, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17804 + }, + { + "epoch": 7.113463843387935, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 17805 + }, + { + "epoch": 7.113863363963244, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 17806 + }, + { + "epoch": 7.114262884538554, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 17807 + }, + { + "epoch": 7.114662405113863, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 17808 + }, + { + "epoch": 7.115061925689173, + "grad_norm": 0.1357421875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 17809 + }, + { + "epoch": 7.115461446264483, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17810 + }, + { + "epoch": 7.115860966839792, + "grad_norm": 0.12890625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17811 + }, + { + "epoch": 7.116260487415102, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 17812 + }, + { + "epoch": 7.116660007990411, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17813 + }, + { + "epoch": 7.117059528565721, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 17814 + }, + { + "epoch": 7.117459049141031, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 17815 + }, + { + "epoch": 7.117858569716341, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17816 + }, + { + "epoch": 7.11825809029165, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 17817 + }, + { + "epoch": 7.1186576108669595, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3321, + "step": 17818 + }, + { + "epoch": 7.119057131442269, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17819 + }, + { + "epoch": 7.119456652017579, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17820 + }, + { + "epoch": 7.119856172592889, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17821 + }, + { + "epoch": 7.120255693168198, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17822 + }, + { + "epoch": 7.1206552137435075, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17823 + }, + { + "epoch": 7.121054734318817, + "grad_norm": 0.10009765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 17824 + }, + { + "epoch": 7.121454254894127, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 17825 + }, + { + "epoch": 7.121853775469437, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 17826 + }, + { + "epoch": 7.122253296044747, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 17827 + }, + { + "epoch": 7.122652816620056, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 17828 + }, + { + "epoch": 7.123052337195365, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 17829 + }, + { + "epoch": 7.123451857770675, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 17830 + }, + { + "epoch": 7.123851378345985, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 17831 + }, + { + "epoch": 7.124250898921295, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 17832 + }, + { + "epoch": 7.1246504194966045, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 17833 + }, + { + "epoch": 7.125049940071913, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 17834 + }, + { + "epoch": 7.125449460647223, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 17835 + }, + { + "epoch": 7.125848981222533, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17836 + }, + { + "epoch": 7.126248501797843, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17837 + }, + { + "epoch": 7.1266480223731525, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 17838 + }, + { + "epoch": 7.127047542948461, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17839 + }, + { + "epoch": 7.127447063523771, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17840 + }, + { + "epoch": 7.127846584099081, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17841 + }, + { + "epoch": 7.128246104674391, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 17842 + }, + { + "epoch": 7.128645625249701, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 17843 + }, + { + "epoch": 7.12904514582501, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17844 + }, + { + "epoch": 7.129444666400319, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 17845 + }, + { + "epoch": 7.129844186975629, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17846 + }, + { + "epoch": 7.130243707550939, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 17847 + }, + { + "epoch": 7.130643228126249, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17848 + }, + { + "epoch": 7.131042748701558, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 17849 + }, + { + "epoch": 7.131442269276867, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17850 + }, + { + "epoch": 7.131841789852177, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 17851 + }, + { + "epoch": 7.132241310427487, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 17852 + }, + { + "epoch": 7.132640831002797, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17853 + }, + { + "epoch": 7.1330403515781065, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 17854 + }, + { + "epoch": 7.133439872153416, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17855 + }, + { + "epoch": 7.133839392728725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 17856 + }, + { + "epoch": 7.134238913304035, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17857 + }, + { + "epoch": 7.134638433879345, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 17858 + }, + { + "epoch": 7.1350379544546545, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17859 + }, + { + "epoch": 7.135437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 17860 + }, + { + "epoch": 7.135836995605274, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17861 + }, + { + "epoch": 7.136236516180583, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17862 + }, + { + "epoch": 7.136636036755893, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17863 + }, + { + "epoch": 7.1370355573312025, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 17864 + }, + { + "epoch": 7.137435077906512, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17865 + }, + { + "epoch": 7.137834598481822, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 17866 + }, + { + "epoch": 7.138234119057131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3002, + "step": 17867 + }, + { + "epoch": 7.138633639632441, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 17868 + }, + { + "epoch": 7.139033160207751, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17869 + }, + { + "epoch": 7.13943268078306, + "grad_norm": 0.0986328125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 17870 + }, + { + "epoch": 7.13983220135837, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17871 + }, + { + "epoch": 7.14023172193368, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17872 + }, + { + "epoch": 7.140631242508989, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 17873 + }, + { + "epoch": 7.141030763084299, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 17874 + }, + { + "epoch": 7.141430283659608, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17875 + }, + { + "epoch": 7.141829804234918, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 17876 + }, + { + "epoch": 7.142229324810228, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 17877 + }, + { + "epoch": 7.142628845385538, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 17878 + }, + { + "epoch": 7.143028365960847, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 17879 + }, + { + "epoch": 7.1434278865361565, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 17880 + }, + { + "epoch": 7.143827407111466, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 17881 + }, + { + "epoch": 7.144226927686776, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17882 + }, + { + "epoch": 7.144626448262086, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 17883 + }, + { + "epoch": 7.145025968837395, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 17884 + }, + { + "epoch": 7.1454254894127045, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 17885 + }, + { + "epoch": 7.145825009988014, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 17886 + }, + { + "epoch": 7.146224530563324, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 17887 + }, + { + "epoch": 7.146624051138634, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17888 + }, + { + "epoch": 7.147023571713944, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 17889 + }, + { + "epoch": 7.1474230922892525, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 17890 + }, + { + "epoch": 7.147822612864562, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17891 + }, + { + "epoch": 7.148222133439872, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17892 + }, + { + "epoch": 7.148621654015182, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 17893 + }, + { + "epoch": 7.149021174590492, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17894 + }, + { + "epoch": 7.1494206951658015, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17895 + }, + { + "epoch": 7.14982021574111, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17896 + }, + { + "epoch": 7.15021973631642, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17897 + }, + { + "epoch": 7.15061925689173, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 17898 + }, + { + "epoch": 7.15101877746704, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17899 + }, + { + "epoch": 7.1514182980423495, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 17900 + }, + { + "epoch": 7.151817818617658, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 17901 + }, + { + "epoch": 7.152217339192968, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 17902 + }, + { + "epoch": 7.152616859768278, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 17903 + }, + { + "epoch": 7.153016380343588, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 17904 + }, + { + "epoch": 7.153415900918898, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 17905 + }, + { + "epoch": 7.153815421494207, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17906 + }, + { + "epoch": 7.154214942069516, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 17907 + }, + { + "epoch": 7.154614462644826, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17908 + }, + { + "epoch": 7.155013983220136, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3054, + "step": 17909 + }, + { + "epoch": 7.155413503795446, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17910 + }, + { + "epoch": 7.155813024370755, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17911 + }, + { + "epoch": 7.156212544946065, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 17912 + }, + { + "epoch": 7.156612065521374, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 17913 + }, + { + "epoch": 7.157011586096684, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 17914 + }, + { + "epoch": 7.157411106671994, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17915 + }, + { + "epoch": 7.157810627247303, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 17916 + }, + { + "epoch": 7.158210147822613, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 17917 + }, + { + "epoch": 7.158609668397922, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 17918 + }, + { + "epoch": 7.159009188973232, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 17919 + }, + { + "epoch": 7.159408709548542, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 17920 + }, + { + "epoch": 7.1598082301238515, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 17921 + }, + { + "epoch": 7.160207750699161, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 17922 + }, + { + "epoch": 7.160607271274471, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 17923 + }, + { + "epoch": 7.16100679184978, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 17924 + }, + { + "epoch": 7.16140631242509, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 17925 + }, + { + "epoch": 7.1618058330003995, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17926 + }, + { + "epoch": 7.162205353575709, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 17927 + }, + { + "epoch": 7.162604874151019, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 17928 + }, + { + "epoch": 7.163004394726328, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 17929 + }, + { + "epoch": 7.163403915301638, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 17930 + }, + { + "epoch": 7.163803435876948, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17931 + }, + { + "epoch": 7.164202956452257, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 17932 + }, + { + "epoch": 7.164602477027567, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17933 + }, + { + "epoch": 7.165001997602877, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 17934 + }, + { + "epoch": 7.165401518178186, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 17935 + }, + { + "epoch": 7.165801038753496, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 17936 + }, + { + "epoch": 7.166200559328805, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 17937 + }, + { + "epoch": 7.166600079904115, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17938 + }, + { + "epoch": 7.166999600479425, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 17939 + }, + { + "epoch": 7.167399121054734, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 17940 + }, + { + "epoch": 7.167798641630044, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17941 + }, + { + "epoch": 7.168198162205353, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 17942 + }, + { + "epoch": 7.168597682780663, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 17943 + }, + { + "epoch": 7.168997203355973, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 17944 + }, + { + "epoch": 7.169396723931283, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3346, + "step": 17945 + }, + { + "epoch": 7.169796244506592, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 17946 + }, + { + "epoch": 7.1701957650819015, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 17947 + }, + { + "epoch": 7.170595285657211, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17948 + }, + { + "epoch": 7.170994806232521, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 17949 + }, + { + "epoch": 7.171394326807831, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 17950 + }, + { + "epoch": 7.171793847383141, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17951 + }, + { + "epoch": 7.1721933679584495, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 17952 + }, + { + "epoch": 7.172592888533759, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 17953 + }, + { + "epoch": 7.172992409109069, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 17954 + }, + { + "epoch": 7.173391929684379, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 17955 + }, + { + "epoch": 7.173791450259689, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 17956 + }, + { + "epoch": 7.174190970834998, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17957 + }, + { + "epoch": 7.174590491410307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17958 + }, + { + "epoch": 7.174990011985617, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17959 + }, + { + "epoch": 7.175389532560927, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 17960 + }, + { + "epoch": 7.175789053136237, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 17961 + }, + { + "epoch": 7.1761885737115465, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 17962 + }, + { + "epoch": 7.176588094286855, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 17963 + }, + { + "epoch": 7.176987614862165, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17964 + }, + { + "epoch": 7.177387135437475, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 17965 + }, + { + "epoch": 7.177786656012785, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 17966 + }, + { + "epoch": 7.1781861765880945, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 17967 + }, + { + "epoch": 7.178585697163404, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 17968 + }, + { + "epoch": 7.178985217738713, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 17969 + }, + { + "epoch": 7.179384738314023, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 17970 + }, + { + "epoch": 7.179784258889333, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 17971 + }, + { + "epoch": 7.180183779464643, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 17972 + }, + { + "epoch": 7.180583300039952, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 17973 + }, + { + "epoch": 7.180982820615261, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 17974 + }, + { + "epoch": 7.181382341190571, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 17975 + }, + { + "epoch": 7.181781861765881, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 17976 + }, + { + "epoch": 7.182181382341191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 17977 + }, + { + "epoch": 7.1825809029165, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 17978 + }, + { + "epoch": 7.18298042349181, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 17979 + }, + { + "epoch": 7.183379944067119, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 17980 + }, + { + "epoch": 7.183779464642429, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 17981 + }, + { + "epoch": 7.184178985217739, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 17982 + }, + { + "epoch": 7.1845785057930485, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 17983 + }, + { + "epoch": 7.184978026368358, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 17984 + }, + { + "epoch": 7.185377546943668, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 17985 + }, + { + "epoch": 7.185777067518977, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 17986 + }, + { + "epoch": 7.186176588094287, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 17987 + }, + { + "epoch": 7.1865761086695965, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 17988 + }, + { + "epoch": 7.186975629244906, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 17989 + }, + { + "epoch": 7.187375149820216, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 17990 + }, + { + "epoch": 7.187774670395525, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 17991 + }, + { + "epoch": 7.188174190970835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 17992 + }, + { + "epoch": 7.1885737115461446, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 17993 + }, + { + "epoch": 7.188973232121454, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 17994 + }, + { + "epoch": 7.189372752696764, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 17995 + }, + { + "epoch": 7.189772273272074, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 17996 + }, + { + "epoch": 7.190171793847383, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 17997 + }, + { + "epoch": 7.190571314422693, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 17998 + }, + { + "epoch": 7.190970834998002, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 17999 + }, + { + "epoch": 7.191370355573312, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.303, + "step": 18000 + }, + { + "epoch": 7.191769876148622, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18001 + }, + { + "epoch": 7.192169396723932, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18002 + }, + { + "epoch": 7.192568917299241, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18003 + }, + { + "epoch": 7.19296843787455, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18004 + }, + { + "epoch": 7.19336795844986, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18005 + }, + { + "epoch": 7.19376747902517, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18006 + }, + { + "epoch": 7.19416699960048, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18007 + }, + { + "epoch": 7.194566520175789, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18008 + }, + { + "epoch": 7.1949660407510985, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18009 + }, + { + "epoch": 7.195365561326408, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18010 + }, + { + "epoch": 7.195765081901718, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18011 + }, + { + "epoch": 7.196164602477028, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18012 + }, + { + "epoch": 7.196564123052338, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18013 + }, + { + "epoch": 7.1969636436276465, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18014 + }, + { + "epoch": 7.197363164202956, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18015 + }, + { + "epoch": 7.197762684778266, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18016 + }, + { + "epoch": 7.198162205353576, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18017 + }, + { + "epoch": 7.198561725928886, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18018 + }, + { + "epoch": 7.1989612465041946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18019 + }, + { + "epoch": 7.199360767079504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18020 + }, + { + "epoch": 7.199760287654814, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18021 + }, + { + "epoch": 7.200159808230124, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18022 + }, + { + "epoch": 7.200559328805434, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18023 + }, + { + "epoch": 7.2009588493807435, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18024 + }, + { + "epoch": 7.201358369956052, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18025 + }, + { + "epoch": 7.201757890531362, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18026 + }, + { + "epoch": 7.202157411106672, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18027 + }, + { + "epoch": 7.202556931681982, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18028 + }, + { + "epoch": 7.2029564522572915, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18029 + }, + { + "epoch": 7.2033559728326, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18030 + }, + { + "epoch": 7.20375549340791, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18031 + }, + { + "epoch": 7.20415501398322, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18032 + }, + { + "epoch": 7.20455453455853, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18033 + }, + { + "epoch": 7.20495405513384, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 18034 + }, + { + "epoch": 7.205353575709149, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18035 + }, + { + "epoch": 7.205753096284458, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18036 + }, + { + "epoch": 7.206152616859768, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18037 + }, + { + "epoch": 7.206552137435078, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18038 + }, + { + "epoch": 7.206951658010388, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18039 + }, + { + "epoch": 7.207351178585697, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18040 + }, + { + "epoch": 7.207750699161007, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18041 + }, + { + "epoch": 7.208150219736316, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18042 + }, + { + "epoch": 7.208549740311626, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18043 + }, + { + "epoch": 7.208949260886936, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18044 + }, + { + "epoch": 7.2093487814622454, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18045 + }, + { + "epoch": 7.209748302037555, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18046 + }, + { + "epoch": 7.210147822612864, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18047 + }, + { + "epoch": 7.210547343188174, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18048 + }, + { + "epoch": 7.210946863763484, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18049 + }, + { + "epoch": 7.2113463843387935, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18050 + }, + { + "epoch": 7.211745904914103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18051 + }, + { + "epoch": 7.212145425489413, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18052 + }, + { + "epoch": 7.212544946064722, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18053 + }, + { + "epoch": 7.212944466640032, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18054 + }, + { + "epoch": 7.2133439872153415, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18055 + }, + { + "epoch": 7.213743507790651, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18056 + }, + { + "epoch": 7.214143028365961, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 18057 + }, + { + "epoch": 7.214542548941271, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18058 + }, + { + "epoch": 7.21494206951658, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18059 + }, + { + "epoch": 7.21534159009189, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18060 + }, + { + "epoch": 7.215741110667199, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18061 + }, + { + "epoch": 7.216140631242509, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18062 + }, + { + "epoch": 7.216540151817819, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18063 + }, + { + "epoch": 7.216939672393128, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18064 + }, + { + "epoch": 7.217339192968438, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18065 + }, + { + "epoch": 7.217738713543747, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18066 + }, + { + "epoch": 7.218138234119057, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18067 + }, + { + "epoch": 7.218537754694367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18068 + }, + { + "epoch": 7.218937275269677, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18069 + }, + { + "epoch": 7.219336795844986, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18070 + }, + { + "epoch": 7.2197363164202955, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18071 + }, + { + "epoch": 7.220135836995605, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18072 + }, + { + "epoch": 7.220535357570915, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18073 + }, + { + "epoch": 7.220934878146225, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18074 + }, + { + "epoch": 7.221334398721535, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18075 + }, + { + "epoch": 7.2217339192968435, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18076 + }, + { + "epoch": 7.222133439872153, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18077 + }, + { + "epoch": 7.222532960447463, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 18078 + }, + { + "epoch": 7.222932481022773, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18079 + }, + { + "epoch": 7.223332001598083, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18080 + }, + { + "epoch": 7.2237315221733915, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18081 + }, + { + "epoch": 7.224131042748701, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 18082 + }, + { + "epoch": 7.224530563324011, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18083 + }, + { + "epoch": 7.224930083899321, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18084 + }, + { + "epoch": 7.225329604474631, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18085 + }, + { + "epoch": 7.2257291250499405, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18086 + }, + { + "epoch": 7.226128645625249, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18087 + }, + { + "epoch": 7.226528166200559, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18088 + }, + { + "epoch": 7.226927686775869, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18089 + }, + { + "epoch": 7.227327207351179, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18090 + }, + { + "epoch": 7.2277267279264885, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18091 + }, + { + "epoch": 7.228126248501797, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18092 + }, + { + "epoch": 7.228525769077107, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18093 + }, + { + "epoch": 7.228925289652417, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18094 + }, + { + "epoch": 7.229324810227727, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18095 + }, + { + "epoch": 7.229724330803037, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18096 + }, + { + "epoch": 7.230123851378346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18097 + }, + { + "epoch": 7.230523371953655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18098 + }, + { + "epoch": 7.230922892528965, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18099 + }, + { + "epoch": 7.231322413104275, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18100 + }, + { + "epoch": 7.231721933679585, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18101 + }, + { + "epoch": 7.232121454254894, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18102 + }, + { + "epoch": 7.232520974830204, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18103 + }, + { + "epoch": 7.232920495405513, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18104 + }, + { + "epoch": 7.233320015980823, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18105 + }, + { + "epoch": 7.233719536556133, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18106 + }, + { + "epoch": 7.234119057131442, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18107 + }, + { + "epoch": 7.234518577706752, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18108 + }, + { + "epoch": 7.234918098282061, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18109 + }, + { + "epoch": 7.235317618857371, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18110 + }, + { + "epoch": 7.235717139432681, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18111 + }, + { + "epoch": 7.2361166600079905, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18112 + }, + { + "epoch": 7.2365161805833, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18113 + }, + { + "epoch": 7.23691570115861, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18114 + }, + { + "epoch": 7.237315221733919, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18115 + }, + { + "epoch": 7.237714742309229, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 18116 + }, + { + "epoch": 7.2381142628845385, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18117 + }, + { + "epoch": 7.238513783459848, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18118 + }, + { + "epoch": 7.238913304035158, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 18119 + }, + { + "epoch": 7.239312824610467, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18120 + }, + { + "epoch": 7.239712345185777, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18121 + }, + { + "epoch": 7.240111865761087, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18122 + }, + { + "epoch": 7.240511386336396, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18123 + }, + { + "epoch": 7.240910906911706, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18124 + }, + { + "epoch": 7.241310427487016, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18125 + }, + { + "epoch": 7.241709948062325, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18126 + }, + { + "epoch": 7.242109468637635, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18127 + }, + { + "epoch": 7.242508989212944, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18128 + }, + { + "epoch": 7.242908509788254, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18129 + }, + { + "epoch": 7.243308030363564, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3293, + "step": 18130 + }, + { + "epoch": 7.243707550938874, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18131 + }, + { + "epoch": 7.244107071514183, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18132 + }, + { + "epoch": 7.244506592089492, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18133 + }, + { + "epoch": 7.244906112664802, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18134 + }, + { + "epoch": 7.245305633240112, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 18135 + }, + { + "epoch": 7.245705153815422, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 18136 + }, + { + "epoch": 7.246104674390731, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18137 + }, + { + "epoch": 7.2465041949660405, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18138 + }, + { + "epoch": 7.24690371554135, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18139 + }, + { + "epoch": 7.24730323611666, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18140 + }, + { + "epoch": 7.24770275669197, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18141 + }, + { + "epoch": 7.24810227726728, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18142 + }, + { + "epoch": 7.2485017978425885, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18143 + }, + { + "epoch": 7.248901318417898, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18144 + }, + { + "epoch": 7.249300838993208, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18145 + }, + { + "epoch": 7.249700359568518, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18146 + }, + { + "epoch": 7.250099880143828, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18147 + }, + { + "epoch": 7.2504994007191375, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18148 + }, + { + "epoch": 7.250898921294446, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18149 + }, + { + "epoch": 7.251298441869756, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18150 + }, + { + "epoch": 7.251697962445066, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18151 + }, + { + "epoch": 7.252097483020376, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18152 + }, + { + "epoch": 7.2524970035956855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18153 + }, + { + "epoch": 7.252896524170994, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18154 + }, + { + "epoch": 7.253296044746304, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18155 + }, + { + "epoch": 7.253695565321614, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18156 + }, + { + "epoch": 7.254095085896924, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18157 + }, + { + "epoch": 7.2544946064722335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3049, + "step": 18158 + }, + { + "epoch": 7.254894127047543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18159 + }, + { + "epoch": 7.255293647622852, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18160 + }, + { + "epoch": 7.255693168198162, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18161 + }, + { + "epoch": 7.256092688773472, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18162 + }, + { + "epoch": 7.256492209348782, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18163 + }, + { + "epoch": 7.256891729924091, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18164 + }, + { + "epoch": 7.257291250499401, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 18165 + }, + { + "epoch": 7.25769077107471, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18166 + }, + { + "epoch": 7.25809029165002, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18167 + }, + { + "epoch": 7.25848981222533, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18168 + }, + { + "epoch": 7.258889332800639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18169 + }, + { + "epoch": 7.259288853375949, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 18170 + }, + { + "epoch": 7.259688373951258, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18171 + }, + { + "epoch": 7.260087894526568, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18172 + }, + { + "epoch": 7.260487415101878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18173 + }, + { + "epoch": 7.2608869356771875, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 18174 + }, + { + "epoch": 7.261286456252497, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18175 + }, + { + "epoch": 7.261685976827807, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18176 + }, + { + "epoch": 7.262085497403116, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18177 + }, + { + "epoch": 7.262485017978426, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18178 + }, + { + "epoch": 7.2628845385537355, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18179 + }, + { + "epoch": 7.263284059129045, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18180 + }, + { + "epoch": 7.263683579704355, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18181 + }, + { + "epoch": 7.264083100279665, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18182 + }, + { + "epoch": 7.264482620854974, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18183 + }, + { + "epoch": 7.2648821414302835, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18184 + }, + { + "epoch": 7.265281662005593, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18185 + }, + { + "epoch": 7.265681182580903, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18186 + }, + { + "epoch": 7.266080703156213, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18187 + }, + { + "epoch": 7.266480223731522, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3269, + "step": 18188 + }, + { + "epoch": 7.266879744306832, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18189 + }, + { + "epoch": 7.267279264882141, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18190 + }, + { + "epoch": 7.267678785457451, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18191 + }, + { + "epoch": 7.268078306032761, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18192 + }, + { + "epoch": 7.26847782660807, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 18193 + }, + { + "epoch": 7.26887734718338, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18194 + }, + { + "epoch": 7.269276867758689, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18195 + }, + { + "epoch": 7.269676388333999, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18196 + }, + { + "epoch": 7.270075908909309, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18197 + }, + { + "epoch": 7.270475429484619, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18198 + }, + { + "epoch": 7.270874950059928, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18199 + }, + { + "epoch": 7.2712744706352375, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18200 + }, + { + "epoch": 7.271673991210547, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18201 + }, + { + "epoch": 7.272073511785857, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 18202 + }, + { + "epoch": 7.272473032361167, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18203 + }, + { + "epoch": 7.272872552936477, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18204 + }, + { + "epoch": 7.2732720735117855, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18205 + }, + { + "epoch": 7.273671594087095, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18206 + }, + { + "epoch": 7.274071114662405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18207 + }, + { + "epoch": 7.274470635237715, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18208 + }, + { + "epoch": 7.274870155813025, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18209 + }, + { + "epoch": 7.2752696763883336, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18210 + }, + { + "epoch": 7.275669196963643, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18211 + }, + { + "epoch": 7.276068717538953, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2981, + "step": 18212 + }, + { + "epoch": 7.276468238114263, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18213 + }, + { + "epoch": 7.276867758689573, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18214 + }, + { + "epoch": 7.2772672792648825, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18215 + }, + { + "epoch": 7.277666799840191, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18216 + }, + { + "epoch": 7.278066320415501, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18217 + }, + { + "epoch": 7.278465840990811, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18218 + }, + { + "epoch": 7.278865361566121, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18219 + }, + { + "epoch": 7.2792648821414305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18220 + }, + { + "epoch": 7.27966440271674, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18221 + }, + { + "epoch": 7.280063923292049, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3312, + "step": 18222 + }, + { + "epoch": 7.280463443867359, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18223 + }, + { + "epoch": 7.280862964442669, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18224 + }, + { + "epoch": 7.281262485017979, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18225 + }, + { + "epoch": 7.281662005593288, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18226 + }, + { + "epoch": 7.282061526168597, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18227 + }, + { + "epoch": 7.282461046743907, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18228 + }, + { + "epoch": 7.282860567319217, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 18229 + }, + { + "epoch": 7.283260087894527, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 18230 + }, + { + "epoch": 7.283659608469836, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 18231 + }, + { + "epoch": 7.284059129045146, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18232 + }, + { + "epoch": 7.284458649620455, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18233 + }, + { + "epoch": 7.284858170195765, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18234 + }, + { + "epoch": 7.285257690771075, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18235 + }, + { + "epoch": 7.285657211346384, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18236 + }, + { + "epoch": 7.286056731921694, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18237 + }, + { + "epoch": 7.286456252497004, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18238 + }, + { + "epoch": 7.286855773072313, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18239 + }, + { + "epoch": 7.287255293647623, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18240 + }, + { + "epoch": 7.2876548142229325, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18241 + }, + { + "epoch": 7.288054334798242, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18242 + }, + { + "epoch": 7.288453855373552, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18243 + }, + { + "epoch": 7.288853375948861, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18244 + }, + { + "epoch": 7.289252896524171, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18245 + }, + { + "epoch": 7.2896524170994805, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18246 + }, + { + "epoch": 7.29005193767479, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18247 + }, + { + "epoch": 7.2904514582501, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18248 + }, + { + "epoch": 7.29085097882541, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18249 + }, + { + "epoch": 7.291250499400719, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18250 + }, + { + "epoch": 7.291650019976029, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18251 + }, + { + "epoch": 7.292049540551338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18252 + }, + { + "epoch": 7.292449061126648, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18253 + }, + { + "epoch": 7.292848581701958, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18254 + }, + { + "epoch": 7.293248102277268, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18255 + }, + { + "epoch": 7.293647622852577, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18256 + }, + { + "epoch": 7.294047143427886, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18257 + }, + { + "epoch": 7.294446664003196, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18258 + }, + { + "epoch": 7.294846184578506, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18259 + }, + { + "epoch": 7.295245705153816, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18260 + }, + { + "epoch": 7.295645225729125, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18261 + }, + { + "epoch": 7.2960447463044344, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18262 + }, + { + "epoch": 7.296444266879744, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18263 + }, + { + "epoch": 7.296843787455054, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18264 + }, + { + "epoch": 7.297243308030364, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18265 + }, + { + "epoch": 7.297642828605674, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 18266 + }, + { + "epoch": 7.2980423491809825, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18267 + }, + { + "epoch": 7.298441869756292, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 18268 + }, + { + "epoch": 7.298841390331602, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18269 + }, + { + "epoch": 7.299240910906912, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18270 + }, + { + "epoch": 7.299640431482222, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18271 + }, + { + "epoch": 7.300039952057531, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18272 + }, + { + "epoch": 7.30043947263284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18273 + }, + { + "epoch": 7.30083899320815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18274 + }, + { + "epoch": 7.30123851378346, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18275 + }, + { + "epoch": 7.30163803435877, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18276 + }, + { + "epoch": 7.3020375549340795, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18277 + }, + { + "epoch": 7.302437075509388, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18278 + }, + { + "epoch": 7.302836596084698, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18279 + }, + { + "epoch": 7.303236116660008, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18280 + }, + { + "epoch": 7.303635637235318, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18281 + }, + { + "epoch": 7.3040351578106275, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18282 + }, + { + "epoch": 7.304434678385936, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18283 + }, + { + "epoch": 7.304834198961246, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 18284 + }, + { + "epoch": 7.305233719536556, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18285 + }, + { + "epoch": 7.305633240111866, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18286 + }, + { + "epoch": 7.3060327606871756, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18287 + }, + { + "epoch": 7.306432281262485, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18288 + }, + { + "epoch": 7.306831801837794, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18289 + }, + { + "epoch": 7.307231322413104, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18290 + }, + { + "epoch": 7.307630842988414, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18291 + }, + { + "epoch": 7.308030363563724, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18292 + }, + { + "epoch": 7.308429884139033, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18293 + }, + { + "epoch": 7.308829404714343, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18294 + }, + { + "epoch": 7.309228925289652, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18295 + }, + { + "epoch": 7.309628445864962, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18296 + }, + { + "epoch": 7.310027966440272, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18297 + }, + { + "epoch": 7.310427487015581, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18298 + }, + { + "epoch": 7.310827007590891, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18299 + }, + { + "epoch": 7.3112265281662, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18300 + }, + { + "epoch": 7.31162604874151, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18301 + }, + { + "epoch": 7.31202556931682, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18302 + }, + { + "epoch": 7.3124250898921295, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18303 + }, + { + "epoch": 7.312824610467439, + "grad_norm": 0.125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18304 + }, + { + "epoch": 7.313224131042749, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18305 + }, + { + "epoch": 7.313623651618058, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18306 + }, + { + "epoch": 7.314023172193368, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 18307 + }, + { + "epoch": 7.3144226927686775, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 18308 + }, + { + "epoch": 7.314822213343987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18309 + }, + { + "epoch": 7.315221733919297, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18310 + }, + { + "epoch": 7.315621254494607, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18311 + }, + { + "epoch": 7.316020775069916, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18312 + }, + { + "epoch": 7.316420295645226, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 18313 + }, + { + "epoch": 7.316819816220535, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18314 + }, + { + "epoch": 7.317219336795845, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18315 + }, + { + "epoch": 7.317618857371155, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18316 + }, + { + "epoch": 7.318018377946464, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 18317 + }, + { + "epoch": 7.318417898521774, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18318 + }, + { + "epoch": 7.318817419097083, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18319 + }, + { + "epoch": 7.319216939672393, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18320 + }, + { + "epoch": 7.319616460247703, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18321 + }, + { + "epoch": 7.320015980823013, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18322 + }, + { + "epoch": 7.320415501398322, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18323 + }, + { + "epoch": 7.320815021973631, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18324 + }, + { + "epoch": 7.321214542548941, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18325 + }, + { + "epoch": 7.321614063124251, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18326 + }, + { + "epoch": 7.322013583699561, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18327 + }, + { + "epoch": 7.322413104274871, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18328 + }, + { + "epoch": 7.3228126248501795, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18329 + }, + { + "epoch": 7.323212145425489, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18330 + }, + { + "epoch": 7.323611666000799, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18331 + }, + { + "epoch": 7.324011186576109, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18332 + }, + { + "epoch": 7.324410707151419, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18333 + }, + { + "epoch": 7.3248102277267275, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18334 + }, + { + "epoch": 7.325209748302037, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 18335 + }, + { + "epoch": 7.325609268877347, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18336 + }, + { + "epoch": 7.326008789452657, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18337 + }, + { + "epoch": 7.326408310027967, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 18338 + }, + { + "epoch": 7.3268078306032765, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 18339 + }, + { + "epoch": 7.327207351178585, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18340 + }, + { + "epoch": 7.327606871753895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18341 + }, + { + "epoch": 7.328006392329205, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18342 + }, + { + "epoch": 7.328405912904515, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18343 + }, + { + "epoch": 7.3288054334798245, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18344 + }, + { + "epoch": 7.329204954055134, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18345 + }, + { + "epoch": 7.329604474630443, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18346 + }, + { + "epoch": 7.330003995205753, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18347 + }, + { + "epoch": 7.330403515781063, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 18348 + }, + { + "epoch": 7.3308030363563725, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18349 + }, + { + "epoch": 7.331202556931682, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18350 + }, + { + "epoch": 7.331602077506991, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18351 + }, + { + "epoch": 7.332001598082301, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18352 + }, + { + "epoch": 7.332401118657611, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18353 + }, + { + "epoch": 7.332800639232921, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18354 + }, + { + "epoch": 7.33320015980823, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 18355 + }, + { + "epoch": 7.33359968038354, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18356 + }, + { + "epoch": 7.333999200958849, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18357 + }, + { + "epoch": 7.334398721534159, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18358 + }, + { + "epoch": 7.334798242109469, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18359 + }, + { + "epoch": 7.335197762684778, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18360 + }, + { + "epoch": 7.335597283260088, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18361 + }, + { + "epoch": 7.335996803835398, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18362 + }, + { + "epoch": 7.336396324410707, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18363 + }, + { + "epoch": 7.336795844986017, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18364 + }, + { + "epoch": 7.3371953655613265, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18365 + }, + { + "epoch": 7.337594886136636, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18366 + }, + { + "epoch": 7.337994406711946, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18367 + }, + { + "epoch": 7.338393927287255, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18368 + }, + { + "epoch": 7.338793447862565, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18369 + }, + { + "epoch": 7.3391929684378745, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18370 + }, + { + "epoch": 7.339592489013184, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18371 + }, + { + "epoch": 7.339992009588494, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18372 + }, + { + "epoch": 7.340391530163803, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18373 + }, + { + "epoch": 7.340791050739113, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18374 + }, + { + "epoch": 7.3411905713144225, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18375 + }, + { + "epoch": 7.341590091889732, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18376 + }, + { + "epoch": 7.341989612465042, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18377 + }, + { + "epoch": 7.342389133040352, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18378 + }, + { + "epoch": 7.342788653615661, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18379 + }, + { + "epoch": 7.343188174190971, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18380 + }, + { + "epoch": 7.34358769476628, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18381 + }, + { + "epoch": 7.34398721534159, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18382 + }, + { + "epoch": 7.3443867359169, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3316, + "step": 18383 + }, + { + "epoch": 7.34478625649221, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18384 + }, + { + "epoch": 7.345185777067519, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18385 + }, + { + "epoch": 7.345585297642828, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18386 + }, + { + "epoch": 7.345984818218138, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18387 + }, + { + "epoch": 7.346384338793448, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18388 + }, + { + "epoch": 7.346783859368758, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18389 + }, + { + "epoch": 7.347183379944067, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18390 + }, + { + "epoch": 7.3475829005193765, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18391 + }, + { + "epoch": 7.347982421094686, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18392 + }, + { + "epoch": 7.348381941669996, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18393 + }, + { + "epoch": 7.348781462245306, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18394 + }, + { + "epoch": 7.349180982820616, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18395 + }, + { + "epoch": 7.3495805033959245, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18396 + }, + { + "epoch": 7.349980023971234, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18397 + }, + { + "epoch": 7.350379544546544, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18398 + }, + { + "epoch": 7.350779065121854, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18399 + }, + { + "epoch": 7.351178585697164, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18400 + }, + { + "epoch": 7.351578106272473, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18401 + }, + { + "epoch": 7.351977626847782, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18402 + }, + { + "epoch": 7.352377147423092, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 18403 + }, + { + "epoch": 7.352776667998402, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18404 + }, + { + "epoch": 7.353176188573712, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18405 + }, + { + "epoch": 7.3535757091490215, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18406 + }, + { + "epoch": 7.35397522972433, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18407 + }, + { + "epoch": 7.35437475029964, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18408 + }, + { + "epoch": 7.35477427087495, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18409 + }, + { + "epoch": 7.35517379145026, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18410 + }, + { + "epoch": 7.3555733120255695, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 18411 + }, + { + "epoch": 7.355972832600879, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18412 + }, + { + "epoch": 7.356372353176188, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18413 + }, + { + "epoch": 7.356771873751498, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3289, + "step": 18414 + }, + { + "epoch": 7.357171394326808, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 18415 + }, + { + "epoch": 7.357570914902118, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18416 + }, + { + "epoch": 7.357970435477427, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18417 + }, + { + "epoch": 7.358369956052737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18418 + }, + { + "epoch": 7.358769476628046, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18419 + }, + { + "epoch": 7.359168997203356, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18420 + }, + { + "epoch": 7.359568517778666, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 18421 + }, + { + "epoch": 7.359968038353975, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18422 + }, + { + "epoch": 7.360367558929285, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3016, + "step": 18423 + }, + { + "epoch": 7.360767079504594, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18424 + }, + { + "epoch": 7.361166600079904, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18425 + }, + { + "epoch": 7.361566120655214, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18426 + }, + { + "epoch": 7.361965641230523, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18427 + }, + { + "epoch": 7.362365161805833, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18428 + }, + { + "epoch": 7.362764682381143, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18429 + }, + { + "epoch": 7.363164202956452, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18430 + }, + { + "epoch": 7.363563723531762, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18431 + }, + { + "epoch": 7.3639632441070715, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18432 + }, + { + "epoch": 7.364362764682381, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18433 + }, + { + "epoch": 7.364762285257691, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18434 + }, + { + "epoch": 7.365161805833001, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18435 + }, + { + "epoch": 7.36556132640831, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18436 + }, + { + "epoch": 7.3659608469836195, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18437 + }, + { + "epoch": 7.366360367558929, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18438 + }, + { + "epoch": 7.366759888134239, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18439 + }, + { + "epoch": 7.367159408709549, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18440 + }, + { + "epoch": 7.367558929284858, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18441 + }, + { + "epoch": 7.367958449860168, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18442 + }, + { + "epoch": 7.368357970435477, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18443 + }, + { + "epoch": 7.368757491010787, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18444 + }, + { + "epoch": 7.369157011586097, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18445 + }, + { + "epoch": 7.369556532161407, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18446 + }, + { + "epoch": 7.369956052736716, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3051, + "step": 18447 + }, + { + "epoch": 7.370355573312025, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18448 + }, + { + "epoch": 7.370755093887335, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18449 + }, + { + "epoch": 7.371154614462645, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18450 + }, + { + "epoch": 7.371554135037955, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18451 + }, + { + "epoch": 7.3719536556132645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18452 + }, + { + "epoch": 7.372353176188573, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18453 + }, + { + "epoch": 7.372752696763883, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18454 + }, + { + "epoch": 7.373152217339193, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18455 + }, + { + "epoch": 7.373551737914503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 18456 + }, + { + "epoch": 7.373951258489813, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3264, + "step": 18457 + }, + { + "epoch": 7.3743507790651215, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18458 + }, + { + "epoch": 7.374750299640431, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18459 + }, + { + "epoch": 7.375149820215741, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18460 + }, + { + "epoch": 7.375549340791051, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18461 + }, + { + "epoch": 7.375948861366361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18462 + }, + { + "epoch": 7.3763483819416695, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18463 + }, + { + "epoch": 7.376747902516979, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18464 + }, + { + "epoch": 7.377147423092289, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 18465 + }, + { + "epoch": 7.377546943667599, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3337, + "step": 18466 + }, + { + "epoch": 7.377946464242909, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18467 + }, + { + "epoch": 7.3783459848182185, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18468 + }, + { + "epoch": 7.378745505393527, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 18469 + }, + { + "epoch": 7.379145025968837, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18470 + }, + { + "epoch": 7.379544546544147, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18471 + }, + { + "epoch": 7.379944067119457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18472 + }, + { + "epoch": 7.3803435876947665, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18473 + }, + { + "epoch": 7.380743108270076, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18474 + }, + { + "epoch": 7.381142628845385, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18475 + }, + { + "epoch": 7.381542149420695, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18476 + }, + { + "epoch": 7.381941669996005, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18477 + }, + { + "epoch": 7.3823411905713145, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18478 + }, + { + "epoch": 7.382740711146624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18479 + }, + { + "epoch": 7.383140231721933, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18480 + }, + { + "epoch": 7.383539752297243, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18481 + }, + { + "epoch": 7.383939272872553, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18482 + }, + { + "epoch": 7.384338793447863, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18483 + }, + { + "epoch": 7.384738314023172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18484 + }, + { + "epoch": 7.385137834598482, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18485 + }, + { + "epoch": 7.385537355173791, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18486 + }, + { + "epoch": 7.385936875749101, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3331, + "step": 18487 + }, + { + "epoch": 7.386336396324411, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18488 + }, + { + "epoch": 7.38673591689972, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18489 + }, + { + "epoch": 7.38713543747503, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18490 + }, + { + "epoch": 7.38753495805034, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18491 + }, + { + "epoch": 7.387934478625649, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18492 + }, + { + "epoch": 7.388333999200959, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18493 + }, + { + "epoch": 7.3887335197762685, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 18494 + }, + { + "epoch": 7.389133040351578, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18495 + }, + { + "epoch": 7.389532560926888, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 18496 + }, + { + "epoch": 7.389932081502197, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18497 + }, + { + "epoch": 7.390331602077507, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18498 + }, + { + "epoch": 7.3907311226528165, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18499 + }, + { + "epoch": 7.391130643228126, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18500 + }, + { + "epoch": 7.391530163803436, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18501 + }, + { + "epoch": 7.391929684378746, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3013, + "step": 18502 + }, + { + "epoch": 7.392329204954055, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18503 + }, + { + "epoch": 7.3927287255293646, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18504 + }, + { + "epoch": 7.393128246104674, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18505 + }, + { + "epoch": 7.393527766679984, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18506 + }, + { + "epoch": 7.393927287255294, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18507 + }, + { + "epoch": 7.394326807830604, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3284, + "step": 18508 + }, + { + "epoch": 7.394726328405913, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 18509 + }, + { + "epoch": 7.395125848981222, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18510 + }, + { + "epoch": 7.395525369556532, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18511 + }, + { + "epoch": 7.395924890131842, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18512 + }, + { + "epoch": 7.396324410707152, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18513 + }, + { + "epoch": 7.396723931282461, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.302, + "step": 18514 + }, + { + "epoch": 7.39712345185777, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18515 + }, + { + "epoch": 7.39752297243308, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18516 + }, + { + "epoch": 7.39792249300839, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18517 + }, + { + "epoch": 7.3983220135837, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18518 + }, + { + "epoch": 7.39872153415901, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18519 + }, + { + "epoch": 7.3991210547343185, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 18520 + }, + { + "epoch": 7.399520575309628, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18521 + }, + { + "epoch": 7.399920095884938, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3278, + "step": 18522 + }, + { + "epoch": 7.400319616460248, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 18523 + }, + { + "epoch": 7.400719137035558, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18524 + }, + { + "epoch": 7.401118657610867, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18525 + }, + { + "epoch": 7.401518178186176, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18526 + }, + { + "epoch": 7.401917698761486, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3252, + "step": 18527 + }, + { + "epoch": 7.402317219336796, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18528 + }, + { + "epoch": 7.402716739912106, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18529 + }, + { + "epoch": 7.4031162604874154, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18530 + }, + { + "epoch": 7.403515781062724, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3027, + "step": 18531 + }, + { + "epoch": 7.403915301638034, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 18532 + }, + { + "epoch": 7.404314822213344, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18533 + }, + { + "epoch": 7.404714342788654, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18534 + }, + { + "epoch": 7.4051138633639635, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 18535 + }, + { + "epoch": 7.405513383939273, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18536 + }, + { + "epoch": 7.405912904514582, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 18537 + }, + { + "epoch": 7.406312425089892, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18538 + }, + { + "epoch": 7.406711945665202, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18539 + }, + { + "epoch": 7.4071114662405115, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18540 + }, + { + "epoch": 7.407510986815821, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18541 + }, + { + "epoch": 7.407910507391131, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18542 + }, + { + "epoch": 7.40831002796644, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18543 + }, + { + "epoch": 7.40870954854175, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18544 + }, + { + "epoch": 7.40910906911706, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18545 + }, + { + "epoch": 7.409508589692369, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18546 + }, + { + "epoch": 7.409908110267679, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18547 + }, + { + "epoch": 7.410307630842988, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18548 + }, + { + "epoch": 7.410707151418298, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18549 + }, + { + "epoch": 7.411106671993608, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18550 + }, + { + "epoch": 7.411506192568917, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18551 + }, + { + "epoch": 7.411905713144227, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18552 + }, + { + "epoch": 7.412305233719536, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18553 + }, + { + "epoch": 7.412704754294846, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18554 + }, + { + "epoch": 7.413104274870156, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 18555 + }, + { + "epoch": 7.4135037954454654, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18556 + }, + { + "epoch": 7.413903316020775, + "grad_norm": 0.047119140625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18557 + }, + { + "epoch": 7.414302836596085, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18558 + }, + { + "epoch": 7.414702357171394, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 18559 + }, + { + "epoch": 7.415101877746704, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18560 + }, + { + "epoch": 7.4155013983220135, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 18561 + }, + { + "epoch": 7.415900918897323, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18562 + }, + { + "epoch": 7.416300439472633, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 18563 + }, + { + "epoch": 7.416699960047943, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18564 + }, + { + "epoch": 7.417099480623252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18565 + }, + { + "epoch": 7.4174990011985615, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.2993, + "step": 18566 + }, + { + "epoch": 7.417898521773871, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18567 + }, + { + "epoch": 7.418298042349181, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18568 + }, + { + "epoch": 7.418697562924491, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18569 + }, + { + "epoch": 7.4190970834998, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18570 + }, + { + "epoch": 7.41949660407511, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18571 + }, + { + "epoch": 7.419896124650419, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3285, + "step": 18572 + }, + { + "epoch": 7.420295645225729, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18573 + }, + { + "epoch": 7.420695165801039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18574 + }, + { + "epoch": 7.421094686376349, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3014, + "step": 18575 + }, + { + "epoch": 7.421494206951658, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18576 + }, + { + "epoch": 7.421893727526967, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3012, + "step": 18577 + }, + { + "epoch": 7.422293248102277, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18578 + }, + { + "epoch": 7.422692768677587, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18579 + }, + { + "epoch": 7.423092289252897, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18580 + }, + { + "epoch": 7.4234918098282066, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 18581 + }, + { + "epoch": 7.4238913304035155, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18582 + }, + { + "epoch": 7.424290850978825, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18583 + }, + { + "epoch": 7.424690371554135, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 18584 + }, + { + "epoch": 7.425089892129445, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 18585 + }, + { + "epoch": 7.425489412704755, + "grad_norm": 0.1103515625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18586 + }, + { + "epoch": 7.4258889332800635, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18587 + }, + { + "epoch": 7.426288453855373, + "grad_norm": 0.1171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18588 + }, + { + "epoch": 7.426687974430683, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18589 + }, + { + "epoch": 7.427087495005993, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18590 + }, + { + "epoch": 7.427487015581303, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18591 + }, + { + "epoch": 7.427886536156612, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18592 + }, + { + "epoch": 7.428286056731921, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 18593 + }, + { + "epoch": 7.428685577307231, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18594 + }, + { + "epoch": 7.429085097882541, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 18595 + }, + { + "epoch": 7.429484618457851, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18596 + }, + { + "epoch": 7.4298841390331605, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 18597 + }, + { + "epoch": 7.43028365960847, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18598 + }, + { + "epoch": 7.430683180183779, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18599 + }, + { + "epoch": 7.431082700759089, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18600 + }, + { + "epoch": 7.431482221334399, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18601 + }, + { + "epoch": 7.4318817419097085, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 18602 + }, + { + "epoch": 7.432281262485018, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18603 + }, + { + "epoch": 7.432680783060327, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18604 + }, + { + "epoch": 7.433080303635637, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18605 + }, + { + "epoch": 7.433479824210947, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18606 + }, + { + "epoch": 7.433879344786257, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18607 + }, + { + "epoch": 7.434278865361566, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18608 + }, + { + "epoch": 7.434678385936876, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18609 + }, + { + "epoch": 7.435077906512185, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 18610 + }, + { + "epoch": 7.435477427087495, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3022, + "step": 18611 + }, + { + "epoch": 7.435876947662805, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18612 + }, + { + "epoch": 7.436276468238114, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.2906, + "step": 18613 + }, + { + "epoch": 7.436675988813424, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18614 + }, + { + "epoch": 7.437075509388734, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18615 + }, + { + "epoch": 7.437475029964043, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18616 + }, + { + "epoch": 7.437874550539353, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18617 + }, + { + "epoch": 7.438274071114662, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 18618 + }, + { + "epoch": 7.438673591689972, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 18619 + }, + { + "epoch": 7.439073112265282, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18620 + }, + { + "epoch": 7.439472632840591, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18621 + }, + { + "epoch": 7.439872153415901, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18622 + }, + { + "epoch": 7.4402716739912105, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 18623 + }, + { + "epoch": 7.44067119456652, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18624 + }, + { + "epoch": 7.44107071514183, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3015, + "step": 18625 + }, + { + "epoch": 7.44147023571714, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18626 + }, + { + "epoch": 7.441869756292449, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18627 + }, + { + "epoch": 7.4422692768677585, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 18628 + }, + { + "epoch": 7.442668797443068, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18629 + }, + { + "epoch": 7.443068318018378, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18630 + }, + { + "epoch": 7.443467838593688, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18631 + }, + { + "epoch": 7.443867359168998, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 18632 + }, + { + "epoch": 7.444266879744307, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18633 + }, + { + "epoch": 7.444666400319616, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 18634 + }, + { + "epoch": 7.445065920894926, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 18635 + }, + { + "epoch": 7.445465441470236, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 18636 + }, + { + "epoch": 7.445864962045546, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18637 + }, + { + "epoch": 7.446264482620855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18638 + }, + { + "epoch": 7.446664003196164, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18639 + }, + { + "epoch": 7.447063523771474, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 18640 + }, + { + "epoch": 7.447463044346784, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18641 + }, + { + "epoch": 7.447862564922094, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18642 + }, + { + "epoch": 7.448262085497403, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 18643 + }, + { + "epoch": 7.448661606072712, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18644 + }, + { + "epoch": 7.449061126648022, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18645 + }, + { + "epoch": 7.449460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18646 + }, + { + "epoch": 7.449860167798642, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18647 + }, + { + "epoch": 7.450259688373952, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18648 + }, + { + "epoch": 7.4506592089492605, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18649 + }, + { + "epoch": 7.45105872952457, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18650 + }, + { + "epoch": 7.45145825009988, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18651 + }, + { + "epoch": 7.45185777067519, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18652 + }, + { + "epoch": 7.4522572912505, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18653 + }, + { + "epoch": 7.452656811825809, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18654 + }, + { + "epoch": 7.453056332401118, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18655 + }, + { + "epoch": 7.453455852976428, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18656 + }, + { + "epoch": 7.453855373551738, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3402, + "step": 18657 + }, + { + "epoch": 7.454254894127048, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18658 + }, + { + "epoch": 7.4546544147023575, + "grad_norm": 0.048095703125, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18659 + }, + { + "epoch": 7.455053935277666, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18660 + }, + { + "epoch": 7.455453455852976, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18661 + }, + { + "epoch": 7.455852976428286, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18662 + }, + { + "epoch": 7.456252497003596, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18663 + }, + { + "epoch": 7.4566520175789055, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 18664 + }, + { + "epoch": 7.457051538154215, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3067, + "step": 18665 + }, + { + "epoch": 7.457451058729524, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 18666 + }, + { + "epoch": 7.457850579304834, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18667 + }, + { + "epoch": 7.458250099880144, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18668 + }, + { + "epoch": 7.4586496204554535, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 18669 + }, + { + "epoch": 7.459049141030763, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18670 + }, + { + "epoch": 7.459448661606073, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 18671 + }, + { + "epoch": 7.459848182181382, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18672 + }, + { + "epoch": 7.460247702756692, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18673 + }, + { + "epoch": 7.460647223332002, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18674 + }, + { + "epoch": 7.461046743907311, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 18675 + }, + { + "epoch": 7.461446264482621, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18676 + }, + { + "epoch": 7.46184578505793, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18677 + }, + { + "epoch": 7.46224530563324, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18678 + }, + { + "epoch": 7.46264482620855, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18679 + }, + { + "epoch": 7.463044346783859, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18680 + }, + { + "epoch": 7.463443867359169, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18681 + }, + { + "epoch": 7.463843387934479, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18682 + }, + { + "epoch": 7.464242908509788, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18683 + }, + { + "epoch": 7.464642429085098, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18684 + }, + { + "epoch": 7.4650419496604075, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18685 + }, + { + "epoch": 7.465441470235717, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18686 + }, + { + "epoch": 7.465840990811027, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18687 + }, + { + "epoch": 7.466240511386337, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18688 + }, + { + "epoch": 7.466640031961646, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18689 + }, + { + "epoch": 7.4670395525369555, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 18690 + }, + { + "epoch": 7.467439073112265, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18691 + }, + { + "epoch": 7.467838593687575, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 18692 + }, + { + "epoch": 7.468238114262885, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 18693 + }, + { + "epoch": 7.468637634838194, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18694 + }, + { + "epoch": 7.4690371554135035, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18695 + }, + { + "epoch": 7.469436675988813, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 18696 + }, + { + "epoch": 7.469836196564123, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18697 + }, + { + "epoch": 7.470235717139433, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18698 + }, + { + "epoch": 7.470635237714743, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18699 + }, + { + "epoch": 7.471034758290052, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18700 + }, + { + "epoch": 7.471434278865361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 18701 + }, + { + "epoch": 7.471833799440671, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18702 + }, + { + "epoch": 7.472233320015981, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.2967, + "step": 18703 + }, + { + "epoch": 7.472632840591291, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18704 + }, + { + "epoch": 7.4730323611666005, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18705 + }, + { + "epoch": 7.473431881741909, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18706 + }, + { + "epoch": 7.473831402317219, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18707 + }, + { + "epoch": 7.474230922892529, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18708 + }, + { + "epoch": 7.474630443467839, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3268, + "step": 18709 + }, + { + "epoch": 7.475029964043149, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 18710 + }, + { + "epoch": 7.4754294846184575, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 18711 + }, + { + "epoch": 7.475829005193767, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18712 + }, + { + "epoch": 7.476228525769077, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18713 + }, + { + "epoch": 7.476628046344387, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 18714 + }, + { + "epoch": 7.477027566919697, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18715 + }, + { + "epoch": 7.477427087495006, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 18716 + }, + { + "epoch": 7.477826608070315, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18717 + }, + { + "epoch": 7.478226128645625, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 18718 + }, + { + "epoch": 7.478625649220935, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3294, + "step": 18719 + }, + { + "epoch": 7.479025169796245, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18720 + }, + { + "epoch": 7.479424690371554, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 18721 + }, + { + "epoch": 7.479824210946864, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 18722 + }, + { + "epoch": 7.480223731522173, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18723 + }, + { + "epoch": 7.480623252097483, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 18724 + }, + { + "epoch": 7.481022772672793, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18725 + }, + { + "epoch": 7.4814222932481025, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18726 + }, + { + "epoch": 7.481821813823412, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18727 + }, + { + "epoch": 7.482221334398721, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18728 + }, + { + "epoch": 7.482620854974031, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18729 + }, + { + "epoch": 7.483020375549341, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 18730 + }, + { + "epoch": 7.4834198961246505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 18731 + }, + { + "epoch": 7.48381941669996, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18732 + }, + { + "epoch": 7.484218937275269, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3323, + "step": 18733 + }, + { + "epoch": 7.484618457850579, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18734 + }, + { + "epoch": 7.485017978425889, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 18735 + }, + { + "epoch": 7.485417499001199, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18736 + }, + { + "epoch": 7.485817019576508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18737 + }, + { + "epoch": 7.486216540151818, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18738 + }, + { + "epoch": 7.486616060727127, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18739 + }, + { + "epoch": 7.487015581302437, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 18740 + }, + { + "epoch": 7.487415101877747, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18741 + }, + { + "epoch": 7.487814622453056, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18742 + }, + { + "epoch": 7.488214143028366, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18743 + }, + { + "epoch": 7.488613663603676, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 18744 + }, + { + "epoch": 7.489013184178985, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18745 + }, + { + "epoch": 7.489412704754295, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18746 + }, + { + "epoch": 7.4898122253296044, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18747 + }, + { + "epoch": 7.490211745904914, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 18748 + }, + { + "epoch": 7.490611266480224, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18749 + }, + { + "epoch": 7.491010787055533, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18750 + }, + { + "epoch": 7.491410307630843, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18751 + }, + { + "epoch": 7.4918098282061525, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 18752 + }, + { + "epoch": 7.492209348781462, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18753 + }, + { + "epoch": 7.492608869356772, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18754 + }, + { + "epoch": 7.493008389932082, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18755 + }, + { + "epoch": 7.493407910507391, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18756 + }, + { + "epoch": 7.4938074310827005, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18757 + }, + { + "epoch": 7.49420695165801, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18758 + }, + { + "epoch": 7.49460647223332, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 18759 + }, + { + "epoch": 7.49500599280863, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 18760 + }, + { + "epoch": 7.49540551338394, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18761 + }, + { + "epoch": 7.495805033959249, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3008, + "step": 18762 + }, + { + "epoch": 7.496204554534558, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18763 + }, + { + "epoch": 7.496604075109868, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 18764 + }, + { + "epoch": 7.497003595685178, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18765 + }, + { + "epoch": 7.497403116260488, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18766 + }, + { + "epoch": 7.497802636835797, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18767 + }, + { + "epoch": 7.498202157411106, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 18768 + }, + { + "epoch": 7.498601677986416, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18769 + }, + { + "epoch": 7.499001198561726, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 18770 + }, + { + "epoch": 7.499400719137036, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 18771 + }, + { + "epoch": 7.4998002397123456, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18772 + }, + { + "epoch": 7.5001997602876544, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18773 + }, + { + "epoch": 7.500599280862964, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 18774 + }, + { + "epoch": 7.500998801438274, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18775 + }, + { + "epoch": 7.501398322013584, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18776 + }, + { + "epoch": 7.501797842588894, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 18777 + }, + { + "epoch": 7.502197363164203, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18778 + }, + { + "epoch": 7.502596883739512, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18779 + }, + { + "epoch": 7.502996404314822, + "grad_norm": 0.095703125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18780 + }, + { + "epoch": 7.503395924890132, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18781 + }, + { + "epoch": 7.503795445465442, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 18782 + }, + { + "epoch": 7.504194966040751, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18783 + }, + { + "epoch": 7.50459448661606, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 18784 + }, + { + "epoch": 7.50499400719137, + "grad_norm": 0.11376953125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18785 + }, + { + "epoch": 7.50539352776668, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18786 + }, + { + "epoch": 7.50579304834199, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18787 + }, + { + "epoch": 7.5061925689172995, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18788 + }, + { + "epoch": 7.506592089492609, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18789 + }, + { + "epoch": 7.506991610067918, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18790 + }, + { + "epoch": 7.507391130643228, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18791 + }, + { + "epoch": 7.507790651218538, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 18792 + }, + { + "epoch": 7.5081901717938475, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18793 + }, + { + "epoch": 7.508589692369157, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18794 + }, + { + "epoch": 7.508989212944467, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 18795 + }, + { + "epoch": 7.509388733519776, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 18796 + }, + { + "epoch": 7.509788254095086, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18797 + }, + { + "epoch": 7.5101877746703956, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18798 + }, + { + "epoch": 7.510587295245705, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18799 + }, + { + "epoch": 7.510986815821015, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 18800 + }, + { + "epoch": 7.511386336396324, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 18801 + }, + { + "epoch": 7.511785856971634, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 18802 + }, + { + "epoch": 7.512185377546944, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 18803 + }, + { + "epoch": 7.512584898122253, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18804 + }, + { + "epoch": 7.512984418697563, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 18805 + }, + { + "epoch": 7.513383939272872, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18806 + }, + { + "epoch": 7.513783459848182, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 18807 + }, + { + "epoch": 7.514182980423492, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18808 + }, + { + "epoch": 7.514582500998801, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 18809 + }, + { + "epoch": 7.514982021574111, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 18810 + }, + { + "epoch": 7.515381542149421, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 18811 + }, + { + "epoch": 7.515781062724731, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 18812 + }, + { + "epoch": 7.51618058330004, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18813 + }, + { + "epoch": 7.5165801038753495, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 18814 + }, + { + "epoch": 7.516979624450659, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18815 + }, + { + "epoch": 7.517379145025969, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18816 + }, + { + "epoch": 7.517778665601279, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18817 + }, + { + "epoch": 7.518178186176588, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 18818 + }, + { + "epoch": 7.5185777067518975, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 18819 + }, + { + "epoch": 7.518977227327207, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 18820 + }, + { + "epoch": 7.519376747902517, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18821 + }, + { + "epoch": 7.519776268477827, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 18822 + }, + { + "epoch": 7.520175789053136, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18823 + }, + { + "epoch": 7.520575309628446, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18824 + }, + { + "epoch": 7.520974830203755, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 18825 + }, + { + "epoch": 7.521374350779065, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 18826 + }, + { + "epoch": 7.521773871354375, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18827 + }, + { + "epoch": 7.522173391929685, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18828 + }, + { + "epoch": 7.5225729125049945, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18829 + }, + { + "epoch": 7.522972433080303, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18830 + }, + { + "epoch": 7.523371953655613, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.328, + "step": 18831 + }, + { + "epoch": 7.523771474230923, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 18832 + }, + { + "epoch": 7.524170994806233, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 18833 + }, + { + "epoch": 7.5245705153815425, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 18834 + }, + { + "epoch": 7.524970035956851, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 18835 + }, + { + "epoch": 7.525369556532161, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18836 + }, + { + "epoch": 7.525769077107471, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 18837 + }, + { + "epoch": 7.526168597682781, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18838 + }, + { + "epoch": 7.526568118258091, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18839 + }, + { + "epoch": 7.5269676388333995, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 18840 + }, + { + "epoch": 7.527367159408709, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18841 + }, + { + "epoch": 7.527766679984019, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18842 + }, + { + "epoch": 7.528166200559329, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 18843 + }, + { + "epoch": 7.528565721134639, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 18844 + }, + { + "epoch": 7.528965241709948, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18845 + }, + { + "epoch": 7.529364762285257, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 18846 + }, + { + "epoch": 7.529764282860567, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 18847 + }, + { + "epoch": 7.530163803435877, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 18848 + }, + { + "epoch": 7.530563324011187, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18849 + }, + { + "epoch": 7.5309628445864965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18850 + }, + { + "epoch": 7.531362365161806, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 18851 + }, + { + "epoch": 7.531761885737115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18852 + }, + { + "epoch": 7.532161406312425, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18853 + }, + { + "epoch": 7.532560926887735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18854 + }, + { + "epoch": 7.5329604474630445, + "grad_norm": 0.154296875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18855 + }, + { + "epoch": 7.533359968038354, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 18856 + }, + { + "epoch": 7.533759488613663, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 18857 + }, + { + "epoch": 7.534159009188973, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 18858 + }, + { + "epoch": 7.534558529764283, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 18859 + }, + { + "epoch": 7.5349580503395925, + "grad_norm": 0.1201171875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 18860 + }, + { + "epoch": 7.535357570914902, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 18861 + }, + { + "epoch": 7.535757091490212, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 18862 + }, + { + "epoch": 7.536156612065521, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18863 + }, + { + "epoch": 7.536556132640831, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 18864 + }, + { + "epoch": 7.536955653216141, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 18865 + }, + { + "epoch": 7.53735517379145, + "grad_norm": 0.08837890625, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18866 + }, + { + "epoch": 7.53775469436676, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 18867 + }, + { + "epoch": 7.53815421494207, + "grad_norm": 0.107421875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18868 + }, + { + "epoch": 7.538553735517379, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18869 + }, + { + "epoch": 7.538953256092689, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18870 + }, + { + "epoch": 7.539352776667998, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 18871 + }, + { + "epoch": 7.539752297243308, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 18872 + }, + { + "epoch": 7.540151817818618, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 18873 + }, + { + "epoch": 7.540551338393927, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18874 + }, + { + "epoch": 7.540950858969237, + "grad_norm": 0.10595703125, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 18875 + }, + { + "epoch": 7.5413503795445465, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 18876 + }, + { + "epoch": 7.541749900119856, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3249, + "step": 18877 + }, + { + "epoch": 7.542149420695166, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18878 + }, + { + "epoch": 7.542548941270476, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 18879 + }, + { + "epoch": 7.542948461845785, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 18880 + }, + { + "epoch": 7.5433479824210945, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 18881 + }, + { + "epoch": 7.543747502996404, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 18882 + }, + { + "epoch": 7.544147023571714, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 18883 + }, + { + "epoch": 7.544546544147024, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18884 + }, + { + "epoch": 7.544946064722334, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18885 + }, + { + "epoch": 7.5453455852976425, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 18886 + }, + { + "epoch": 7.545745105872952, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18887 + }, + { + "epoch": 7.546144626448262, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18888 + }, + { + "epoch": 7.546544147023572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 18889 + }, + { + "epoch": 7.546943667598882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 18890 + }, + { + "epoch": 7.547343188174191, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 18891 + }, + { + "epoch": 7.5477427087495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18892 + }, + { + "epoch": 7.54814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 18893 + }, + { + "epoch": 7.54854174990012, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18894 + }, + { + "epoch": 7.54894127047543, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 18895 + }, + { + "epoch": 7.549340791050739, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18896 + }, + { + "epoch": 7.549740311626048, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 18897 + }, + { + "epoch": 7.550139832201358, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 18898 + }, + { + "epoch": 7.550539352776668, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18899 + }, + { + "epoch": 7.550938873351978, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 18900 + }, + { + "epoch": 7.551338393927288, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 18901 + }, + { + "epoch": 7.551737914502597, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3058, + "step": 18902 + }, + { + "epoch": 7.552137435077906, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18903 + }, + { + "epoch": 7.552536955653216, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 18904 + }, + { + "epoch": 7.552936476228526, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 18905 + }, + { + "epoch": 7.553335996803836, + "grad_norm": 0.0908203125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 18906 + }, + { + "epoch": 7.553735517379145, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18907 + }, + { + "epoch": 7.554135037954454, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 18908 + }, + { + "epoch": 7.554534558529764, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 18909 + }, + { + "epoch": 7.554934079105074, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 18910 + }, + { + "epoch": 7.555333599680384, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 18911 + }, + { + "epoch": 7.555733120255693, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18912 + }, + { + "epoch": 7.556132640831002, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.327, + "step": 18913 + }, + { + "epoch": 7.556532161406312, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18914 + }, + { + "epoch": 7.556931681981622, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 18915 + }, + { + "epoch": 7.557331202556932, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 18916 + }, + { + "epoch": 7.5577307231322415, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3029, + "step": 18917 + }, + { + "epoch": 7.558130243707551, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 18918 + }, + { + "epoch": 7.558529764282861, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 18919 + }, + { + "epoch": 7.55892928485817, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 18920 + }, + { + "epoch": 7.55932880543348, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18921 + }, + { + "epoch": 7.5597283260087895, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18922 + }, + { + "epoch": 7.560127846584099, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 18923 + }, + { + "epoch": 7.560527367159409, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18924 + }, + { + "epoch": 7.560926887734718, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18925 + }, + { + "epoch": 7.561326408310028, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 18926 + }, + { + "epoch": 7.561725928885338, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18927 + }, + { + "epoch": 7.562125449460647, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18928 + }, + { + "epoch": 7.562524970035957, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 18929 + }, + { + "epoch": 7.562924490611266, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 18930 + }, + { + "epoch": 7.563324011186576, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 18931 + }, + { + "epoch": 7.563723531761886, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 18932 + }, + { + "epoch": 7.564123052337195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18933 + }, + { + "epoch": 7.564522572912505, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 18934 + }, + { + "epoch": 7.564922093487815, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 18935 + }, + { + "epoch": 7.565321614063124, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 18936 + }, + { + "epoch": 7.565721134638434, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18937 + }, + { + "epoch": 7.566120655213743, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18938 + }, + { + "epoch": 7.566520175789053, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18939 + }, + { + "epoch": 7.566919696364363, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 18940 + }, + { + "epoch": 7.567319216939673, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 18941 + }, + { + "epoch": 7.567718737514982, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 18942 + }, + { + "epoch": 7.5681182580902915, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18943 + }, + { + "epoch": 7.568517778665601, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 18944 + }, + { + "epoch": 7.568917299240911, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18945 + }, + { + "epoch": 7.569316819816221, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 18946 + }, + { + "epoch": 7.56971634039153, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 18947 + }, + { + "epoch": 7.5701158609668395, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 18948 + }, + { + "epoch": 7.570515381542149, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 18949 + }, + { + "epoch": 7.570914902117459, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 18950 + }, + { + "epoch": 7.571314422692769, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 18951 + }, + { + "epoch": 7.571713943268079, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 18952 + }, + { + "epoch": 7.572113463843388, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 18953 + }, + { + "epoch": 7.572512984418697, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 18954 + }, + { + "epoch": 7.572912504994007, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 18955 + }, + { + "epoch": 7.573312025569317, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 18956 + }, + { + "epoch": 7.573711546144627, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 18957 + }, + { + "epoch": 7.5741110667199365, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18958 + }, + { + "epoch": 7.574510587295245, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 18959 + }, + { + "epoch": 7.574910107870555, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 18960 + }, + { + "epoch": 7.575309628445865, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 18961 + }, + { + "epoch": 7.575709149021175, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 18962 + }, + { + "epoch": 7.5761086695964845, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 18963 + }, + { + "epoch": 7.5765081901717934, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 18964 + }, + { + "epoch": 7.576907710747103, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 18965 + }, + { + "epoch": 7.577307231322413, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3021, + "step": 18966 + }, + { + "epoch": 7.577706751897723, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3307, + "step": 18967 + }, + { + "epoch": 7.578106272473033, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 18968 + }, + { + "epoch": 7.578505793048342, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3255, + "step": 18969 + }, + { + "epoch": 7.578905313623651, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 18970 + }, + { + "epoch": 7.579304834198961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 18971 + }, + { + "epoch": 7.579704354774271, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.332, + "step": 18972 + }, + { + "epoch": 7.580103875349581, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 18973 + }, + { + "epoch": 7.58050339592489, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 18974 + }, + { + "epoch": 7.5809029165002, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 18975 + }, + { + "epoch": 7.581302437075509, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 18976 + }, + { + "epoch": 7.581701957650819, + "grad_norm": 0.09716796875, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 18977 + }, + { + "epoch": 7.582101478226129, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 18978 + }, + { + "epoch": 7.5825009988014385, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18979 + }, + { + "epoch": 7.582900519376748, + "grad_norm": 0.119140625, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 18980 + }, + { + "epoch": 7.583300039952057, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.2999, + "step": 18981 + }, + { + "epoch": 7.583699560527367, + "grad_norm": 0.11181640625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 18982 + }, + { + "epoch": 7.584099081102677, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 18983 + }, + { + "epoch": 7.5844986016779865, + "grad_norm": 0.1064453125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 18984 + }, + { + "epoch": 7.584898122253296, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 18985 + }, + { + "epoch": 7.585297642828605, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 18986 + }, + { + "epoch": 7.585697163403915, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 18987 + }, + { + "epoch": 7.586096683979225, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 18988 + }, + { + "epoch": 7.5864962045545346, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 18989 + }, + { + "epoch": 7.586895725129844, + "grad_norm": 0.047607421875, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 18990 + }, + { + "epoch": 7.587295245705154, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 18991 + }, + { + "epoch": 7.587694766280464, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 18992 + }, + { + "epoch": 7.588094286855773, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 18993 + }, + { + "epoch": 7.588493807431083, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 18994 + }, + { + "epoch": 7.588893328006392, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 18995 + }, + { + "epoch": 7.589292848581702, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 18996 + }, + { + "epoch": 7.589692369157012, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 18997 + }, + { + "epoch": 7.590091889732321, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 18998 + }, + { + "epoch": 7.590491410307631, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 18999 + }, + { + "epoch": 7.59089093088294, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19000 + }, + { + "epoch": 7.59129045145825, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19001 + }, + { + "epoch": 7.59168997203356, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19002 + }, + { + "epoch": 7.592089492608869, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 19003 + }, + { + "epoch": 7.592489013184179, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19004 + }, + { + "epoch": 7.5928885337594885, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19005 + }, + { + "epoch": 7.593288054334798, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19006 + }, + { + "epoch": 7.593687574910108, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19007 + }, + { + "epoch": 7.594087095485418, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19008 + }, + { + "epoch": 7.594486616060728, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19009 + }, + { + "epoch": 7.5948861366360365, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19010 + }, + { + "epoch": 7.595285657211346, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19011 + }, + { + "epoch": 7.595685177786656, + "grad_norm": 0.1083984375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19012 + }, + { + "epoch": 7.596084698361966, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19013 + }, + { + "epoch": 7.596484218937276, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3281, + "step": 19014 + }, + { + "epoch": 7.5968837395125846, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19015 + }, + { + "epoch": 7.597283260087894, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19016 + }, + { + "epoch": 7.597682780663204, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19017 + }, + { + "epoch": 7.598082301238514, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19018 + }, + { + "epoch": 7.598481821813824, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19019 + }, + { + "epoch": 7.598881342389133, + "grad_norm": 0.103515625, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19020 + }, + { + "epoch": 7.599280862964442, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19021 + }, + { + "epoch": 7.599680383539752, + "grad_norm": 0.0927734375, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19022 + }, + { + "epoch": 7.600079904115062, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19023 + }, + { + "epoch": 7.600479424690372, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19024 + }, + { + "epoch": 7.6008789452656815, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19025 + }, + { + "epoch": 7.60127846584099, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19026 + }, + { + "epoch": 7.6016779864163, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19027 + }, + { + "epoch": 7.60207750699161, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19028 + }, + { + "epoch": 7.60247702756692, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19029 + }, + { + "epoch": 7.60287654814223, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19030 + }, + { + "epoch": 7.603276068717539, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19031 + }, + { + "epoch": 7.603675589292848, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19032 + }, + { + "epoch": 7.604075109868158, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19033 + }, + { + "epoch": 7.604474630443468, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19034 + }, + { + "epoch": 7.604874151018778, + "grad_norm": 0.11328125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19035 + }, + { + "epoch": 7.605273671594087, + "grad_norm": 0.1025390625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19036 + }, + { + "epoch": 7.605673192169396, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19037 + }, + { + "epoch": 7.606072712744706, + "grad_norm": 0.11474609375, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19038 + }, + { + "epoch": 7.606472233320016, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19039 + }, + { + "epoch": 7.606871753895326, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19040 + }, + { + "epoch": 7.6072712744706354, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19041 + }, + { + "epoch": 7.607670795045945, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19042 + }, + { + "epoch": 7.608070315621254, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19043 + }, + { + "epoch": 7.608469836196564, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19044 + }, + { + "epoch": 7.608869356771874, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19045 + }, + { + "epoch": 7.6092688773471835, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19046 + }, + { + "epoch": 7.609668397922493, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19047 + }, + { + "epoch": 7.610067918497803, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19048 + }, + { + "epoch": 7.610467439073112, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19049 + }, + { + "epoch": 7.610866959648422, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19050 + }, + { + "epoch": 7.6112664802237315, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19051 + }, + { + "epoch": 7.611666000799041, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3273, + "step": 19052 + }, + { + "epoch": 7.612065521374351, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19053 + }, + { + "epoch": 7.61246504194966, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19054 + }, + { + "epoch": 7.61286456252497, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19055 + }, + { + "epoch": 7.61326408310028, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19056 + }, + { + "epoch": 7.613663603675589, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19057 + }, + { + "epoch": 7.614063124250899, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19058 + }, + { + "epoch": 7.614462644826209, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19059 + }, + { + "epoch": 7.614862165401518, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19060 + }, + { + "epoch": 7.615261685976828, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19061 + }, + { + "epoch": 7.615661206552137, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19062 + }, + { + "epoch": 7.616060727127447, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19063 + }, + { + "epoch": 7.616460247702757, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19064 + }, + { + "epoch": 7.616859768278067, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19065 + }, + { + "epoch": 7.617259288853376, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19066 + }, + { + "epoch": 7.6176588094286855, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19067 + }, + { + "epoch": 7.618058330003995, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19068 + }, + { + "epoch": 7.618457850579305, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19069 + }, + { + "epoch": 7.618857371154615, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19070 + }, + { + "epoch": 7.619256891729924, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19071 + }, + { + "epoch": 7.6196564123052335, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19072 + }, + { + "epoch": 7.620055932880543, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19073 + }, + { + "epoch": 7.620455453455853, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19074 + }, + { + "epoch": 7.620854974031163, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19075 + }, + { + "epoch": 7.621254494606472, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19076 + }, + { + "epoch": 7.6216540151817815, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19077 + }, + { + "epoch": 7.622053535757091, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19078 + }, + { + "epoch": 7.622453056332401, + "grad_norm": 0.04833984375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19079 + }, + { + "epoch": 7.622852576907711, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19080 + }, + { + "epoch": 7.623252097483021, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19081 + }, + { + "epoch": 7.6236516180583305, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19082 + }, + { + "epoch": 7.624051138633639, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19083 + }, + { + "epoch": 7.624450659208949, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19084 + }, + { + "epoch": 7.624850179784259, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19085 + }, + { + "epoch": 7.625249700359569, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19086 + }, + { + "epoch": 7.6256492209348785, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19087 + }, + { + "epoch": 7.626048741510187, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19088 + }, + { + "epoch": 7.626448262085497, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19089 + }, + { + "epoch": 7.626847782660807, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19090 + }, + { + "epoch": 7.627247303236117, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19091 + }, + { + "epoch": 7.627646823811427, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19092 + }, + { + "epoch": 7.6280463443867355, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19093 + }, + { + "epoch": 7.628445864962045, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19094 + }, + { + "epoch": 7.628845385537355, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19095 + }, + { + "epoch": 7.629244906112665, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19096 + }, + { + "epoch": 7.629644426687975, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19097 + }, + { + "epoch": 7.630043947263284, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19098 + }, + { + "epoch": 7.630443467838594, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19099 + }, + { + "epoch": 7.630842988413903, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19100 + }, + { + "epoch": 7.631242508989213, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19101 + }, + { + "epoch": 7.631642029564523, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19102 + }, + { + "epoch": 7.632041550139832, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19103 + }, + { + "epoch": 7.632441070715142, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19104 + }, + { + "epoch": 7.632840591290451, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19105 + }, + { + "epoch": 7.633240111865761, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19106 + }, + { + "epoch": 7.633639632441071, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19107 + }, + { + "epoch": 7.6340391530163805, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19108 + }, + { + "epoch": 7.63443867359169, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19109 + }, + { + "epoch": 7.634838194166999, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19110 + }, + { + "epoch": 7.635237714742309, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19111 + }, + { + "epoch": 7.635637235317619, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19112 + }, + { + "epoch": 7.6360367558929285, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19113 + }, + { + "epoch": 7.636436276468238, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19114 + }, + { + "epoch": 7.636835797043548, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19115 + }, + { + "epoch": 7.637235317618857, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 19116 + }, + { + "epoch": 7.637634838194167, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19117 + }, + { + "epoch": 7.638034358769477, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19118 + }, + { + "epoch": 7.638433879344786, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19119 + }, + { + "epoch": 7.638833399920096, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19120 + }, + { + "epoch": 7.639232920495406, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19121 + }, + { + "epoch": 7.639632441070715, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19122 + }, + { + "epoch": 7.640031961646025, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19123 + }, + { + "epoch": 7.640431482221334, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19124 + }, + { + "epoch": 7.640831002796644, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19125 + }, + { + "epoch": 7.641230523371954, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19126 + }, + { + "epoch": 7.641630043947263, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3038, + "step": 19127 + }, + { + "epoch": 7.642029564522573, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19128 + }, + { + "epoch": 7.642429085097882, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19129 + }, + { + "epoch": 7.642828605673192, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19130 + }, + { + "epoch": 7.643228126248502, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19131 + }, + { + "epoch": 7.643627646823812, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19132 + }, + { + "epoch": 7.644027167399121, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3061, + "step": 19133 + }, + { + "epoch": 7.6444266879744305, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19134 + }, + { + "epoch": 7.64482620854974, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19135 + }, + { + "epoch": 7.64522572912505, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19136 + }, + { + "epoch": 7.64562524970036, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19137 + }, + { + "epoch": 7.64602477027567, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19138 + }, + { + "epoch": 7.6464242908509785, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19139 + }, + { + "epoch": 7.646823811426288, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 19140 + }, + { + "epoch": 7.647223332001598, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19141 + }, + { + "epoch": 7.647622852576908, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19142 + }, + { + "epoch": 7.648022373152218, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3057, + "step": 19143 + }, + { + "epoch": 7.648421893727527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19144 + }, + { + "epoch": 7.648821414302836, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19145 + }, + { + "epoch": 7.649220934878146, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19146 + }, + { + "epoch": 7.649620455453456, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19147 + }, + { + "epoch": 7.650019976028766, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19148 + }, + { + "epoch": 7.6504194966040755, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19149 + }, + { + "epoch": 7.650819017179384, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3207, + "step": 19150 + }, + { + "epoch": 7.651218537754694, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19151 + }, + { + "epoch": 7.651618058330004, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19152 + }, + { + "epoch": 7.652017578905314, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19153 + }, + { + "epoch": 7.6524170994806235, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19154 + }, + { + "epoch": 7.652816620055933, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19155 + }, + { + "epoch": 7.653216140631242, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19156 + }, + { + "epoch": 7.653615661206552, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19157 + }, + { + "epoch": 7.654015181781862, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19158 + }, + { + "epoch": 7.654414702357172, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19159 + }, + { + "epoch": 7.654814222932481, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19160 + }, + { + "epoch": 7.65521374350779, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19161 + }, + { + "epoch": 7.6556132640831, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19162 + }, + { + "epoch": 7.65601278465841, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.2987, + "step": 19163 + }, + { + "epoch": 7.65641230523372, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19164 + }, + { + "epoch": 7.656811825809029, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19165 + }, + { + "epoch": 7.657211346384338, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19166 + }, + { + "epoch": 7.657610866959648, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3031, + "step": 19167 + }, + { + "epoch": 7.658010387534958, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19168 + }, + { + "epoch": 7.658409908110268, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19169 + }, + { + "epoch": 7.6588094286855775, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19170 + }, + { + "epoch": 7.659208949260887, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19171 + }, + { + "epoch": 7.659608469836197, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19172 + }, + { + "epoch": 7.660007990411506, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19173 + }, + { + "epoch": 7.660407510986816, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19174 + }, + { + "epoch": 7.6608070315621255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19175 + }, + { + "epoch": 7.661206552137435, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19176 + }, + { + "epoch": 7.661606072712745, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19177 + }, + { + "epoch": 7.662005593288054, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19178 + }, + { + "epoch": 7.662405113863364, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19179 + }, + { + "epoch": 7.6628046344386735, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19180 + }, + { + "epoch": 7.663204155013983, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19181 + }, + { + "epoch": 7.663603675589293, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19182 + }, + { + "epoch": 7.664003196164602, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19183 + }, + { + "epoch": 7.664402716739912, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19184 + }, + { + "epoch": 7.664802237315222, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3261, + "step": 19185 + }, + { + "epoch": 7.665201757890531, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19186 + }, + { + "epoch": 7.665601278465841, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3274, + "step": 19187 + }, + { + "epoch": 7.666000799041151, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19188 + }, + { + "epoch": 7.666400319616461, + "grad_norm": 0.08935546875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19189 + }, + { + "epoch": 7.66679984019177, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19190 + }, + { + "epoch": 7.667199360767079, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19191 + }, + { + "epoch": 7.667598881342389, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19192 + }, + { + "epoch": 7.667998401917699, + "grad_norm": 0.099609375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19193 + }, + { + "epoch": 7.668397922493009, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19194 + }, + { + "epoch": 7.668797443068318, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19195 + }, + { + "epoch": 7.6691969636436275, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19196 + }, + { + "epoch": 7.669596484218937, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19197 + }, + { + "epoch": 7.669996004794247, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19198 + }, + { + "epoch": 7.670395525369557, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19199 + }, + { + "epoch": 7.670795045944866, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19200 + }, + { + "epoch": 7.6711945665201755, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19201 + }, + { + "epoch": 7.671594087095485, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19202 + }, + { + "epoch": 7.671993607670795, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19203 + }, + { + "epoch": 7.672393128246105, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.2982, + "step": 19204 + }, + { + "epoch": 7.672792648821415, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 19205 + }, + { + "epoch": 7.6731921693967235, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19206 + }, + { + "epoch": 7.673591689972033, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19207 + }, + { + "epoch": 7.673991210547343, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19208 + }, + { + "epoch": 7.674390731122653, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3034, + "step": 19209 + }, + { + "epoch": 7.674790251697963, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19210 + }, + { + "epoch": 7.6751897722732725, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19211 + }, + { + "epoch": 7.675589292848581, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19212 + }, + { + "epoch": 7.675988813423891, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19213 + }, + { + "epoch": 7.676388333999201, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19214 + }, + { + "epoch": 7.676787854574511, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19215 + }, + { + "epoch": 7.6771873751498205, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19216 + }, + { + "epoch": 7.677586895725129, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 19217 + }, + { + "epoch": 7.677986416300439, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3282, + "step": 19218 + }, + { + "epoch": 7.678385936875749, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19219 + }, + { + "epoch": 7.678785457451059, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19220 + }, + { + "epoch": 7.679184978026369, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19221 + }, + { + "epoch": 7.679584498601678, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19222 + }, + { + "epoch": 7.679984019176987, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19223 + }, + { + "epoch": 7.680383539752297, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19224 + }, + { + "epoch": 7.680783060327607, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19225 + }, + { + "epoch": 7.681182580902917, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19226 + }, + { + "epoch": 7.681582101478226, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19227 + }, + { + "epoch": 7.681981622053536, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19228 + }, + { + "epoch": 7.682381142628845, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19229 + }, + { + "epoch": 7.682780663204155, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19230 + }, + { + "epoch": 7.683180183779465, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19231 + }, + { + "epoch": 7.683579704354774, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19232 + }, + { + "epoch": 7.683979224930084, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19233 + }, + { + "epoch": 7.684378745505393, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19234 + }, + { + "epoch": 7.684778266080703, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19235 + }, + { + "epoch": 7.685177786656013, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19236 + }, + { + "epoch": 7.6855773072313225, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19237 + }, + { + "epoch": 7.685976827806632, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19238 + }, + { + "epoch": 7.686376348381942, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19239 + }, + { + "epoch": 7.686775868957251, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19240 + }, + { + "epoch": 7.687175389532561, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19241 + }, + { + "epoch": 7.6875749101078705, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19242 + }, + { + "epoch": 7.68797443068318, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19243 + }, + { + "epoch": 7.68837395125849, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19244 + }, + { + "epoch": 7.6887734718338, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19245 + }, + { + "epoch": 7.689172992409109, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19246 + }, + { + "epoch": 7.689572512984419, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19247 + }, + { + "epoch": 7.689972033559728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2992, + "step": 19248 + }, + { + "epoch": 7.690371554135038, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19249 + }, + { + "epoch": 7.690771074710348, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19250 + }, + { + "epoch": 7.691170595285657, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19251 + }, + { + "epoch": 7.691570115860967, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19252 + }, + { + "epoch": 7.691969636436276, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19253 + }, + { + "epoch": 7.692369157011586, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19254 + }, + { + "epoch": 7.692768677586896, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19255 + }, + { + "epoch": 7.693168198162205, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19256 + }, + { + "epoch": 7.693567718737515, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19257 + }, + { + "epoch": 7.6939672393128244, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19258 + }, + { + "epoch": 7.694366759888134, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19259 + }, + { + "epoch": 7.694766280463444, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19260 + }, + { + "epoch": 7.695165801038754, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19261 + }, + { + "epoch": 7.695565321614064, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19262 + }, + { + "epoch": 7.6959648421893725, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19263 + }, + { + "epoch": 7.696364362764682, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 19264 + }, + { + "epoch": 7.696763883339992, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 19265 + }, + { + "epoch": 7.697163403915302, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19266 + }, + { + "epoch": 7.697562924490612, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19267 + }, + { + "epoch": 7.6979624450659205, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3254, + "step": 19268 + }, + { + "epoch": 7.69836196564123, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19269 + }, + { + "epoch": 7.69876148621654, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19270 + }, + { + "epoch": 7.69916100679185, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19271 + }, + { + "epoch": 7.69956052736716, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19272 + }, + { + "epoch": 7.699960047942469, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19273 + }, + { + "epoch": 7.700359568517778, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19274 + }, + { + "epoch": 7.700759089093088, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3208, + "step": 19275 + }, + { + "epoch": 7.701158609668398, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3037, + "step": 19276 + }, + { + "epoch": 7.701558130243708, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19277 + }, + { + "epoch": 7.7019576508190175, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19278 + }, + { + "epoch": 7.702357171394327, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19279 + }, + { + "epoch": 7.702756691969636, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19280 + }, + { + "epoch": 7.703156212544946, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19281 + }, + { + "epoch": 7.703555733120256, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19282 + }, + { + "epoch": 7.7039552536955656, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19283 + }, + { + "epoch": 7.704354774270875, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19284 + }, + { + "epoch": 7.704754294846184, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19285 + }, + { + "epoch": 7.705153815421494, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19286 + }, + { + "epoch": 7.705553335996804, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3221, + "step": 19287 + }, + { + "epoch": 7.705952856572114, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19288 + }, + { + "epoch": 7.706352377147423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19289 + }, + { + "epoch": 7.706751897722732, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19290 + }, + { + "epoch": 7.707151418298042, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19291 + }, + { + "epoch": 7.707550938873352, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19292 + }, + { + "epoch": 7.707950459448662, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3099, + "step": 19293 + }, + { + "epoch": 7.708349980023971, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19294 + }, + { + "epoch": 7.708749500599281, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3213, + "step": 19295 + }, + { + "epoch": 7.70914902117459, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19296 + }, + { + "epoch": 7.7095485417499, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19297 + }, + { + "epoch": 7.70994806232521, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19298 + }, + { + "epoch": 7.7103475829005195, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19299 + }, + { + "epoch": 7.710747103475829, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19300 + }, + { + "epoch": 7.711146624051139, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19301 + }, + { + "epoch": 7.711546144626448, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19302 + }, + { + "epoch": 7.711945665201758, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19303 + }, + { + "epoch": 7.7123451857770675, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19304 + }, + { + "epoch": 7.712744706352377, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19305 + }, + { + "epoch": 7.713144226927687, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19306 + }, + { + "epoch": 7.713543747502996, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19307 + }, + { + "epoch": 7.713943268078306, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19308 + }, + { + "epoch": 7.714342788653616, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19309 + }, + { + "epoch": 7.714742309228925, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19310 + }, + { + "epoch": 7.715141829804235, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19311 + }, + { + "epoch": 7.715541350379545, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3104, + "step": 19312 + }, + { + "epoch": 7.715940870954854, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3113, + "step": 19313 + }, + { + "epoch": 7.716340391530164, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19314 + }, + { + "epoch": 7.716739912105473, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3235, + "step": 19315 + }, + { + "epoch": 7.717139432680783, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19316 + }, + { + "epoch": 7.717538953256093, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19317 + }, + { + "epoch": 7.717938473831403, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19318 + }, + { + "epoch": 7.718337994406712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19319 + }, + { + "epoch": 7.718737514982021, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19320 + }, + { + "epoch": 7.719137035557331, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19321 + }, + { + "epoch": 7.719536556132641, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19322 + }, + { + "epoch": 7.719936076707951, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19323 + }, + { + "epoch": 7.72033559728326, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19324 + }, + { + "epoch": 7.7207351178585695, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19325 + }, + { + "epoch": 7.721134638433879, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19326 + }, + { + "epoch": 7.721534159009189, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19327 + }, + { + "epoch": 7.721933679584499, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19328 + }, + { + "epoch": 7.722333200159808, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19329 + }, + { + "epoch": 7.7227327207351175, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19330 + }, + { + "epoch": 7.723132241310427, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19331 + }, + { + "epoch": 7.723531761885737, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 19332 + }, + { + "epoch": 7.723931282461047, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19333 + }, + { + "epoch": 7.724330803036357, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19334 + }, + { + "epoch": 7.7247303236116664, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19335 + }, + { + "epoch": 7.725129844186975, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19336 + }, + { + "epoch": 7.725529364762285, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19337 + }, + { + "epoch": 7.725928885337595, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3077, + "step": 19338 + }, + { + "epoch": 7.726328405912905, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19339 + }, + { + "epoch": 7.7267279264882145, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19340 + }, + { + "epoch": 7.727127447063523, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19341 + }, + { + "epoch": 7.727526967638833, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19342 + }, + { + "epoch": 7.727926488214143, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19343 + }, + { + "epoch": 7.728326008789453, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19344 + }, + { + "epoch": 7.7287255293647625, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3291, + "step": 19345 + }, + { + "epoch": 7.729125049940071, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19346 + }, + { + "epoch": 7.729524570515381, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19347 + }, + { + "epoch": 7.729924091090691, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19348 + }, + { + "epoch": 7.730323611666001, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19349 + }, + { + "epoch": 7.730723132241311, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19350 + }, + { + "epoch": 7.73112265281662, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19351 + }, + { + "epoch": 7.73152217339193, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19352 + }, + { + "epoch": 7.731921693967239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 19353 + }, + { + "epoch": 7.732321214542549, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19354 + }, + { + "epoch": 7.732720735117859, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 19355 + }, + { + "epoch": 7.733120255693168, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19356 + }, + { + "epoch": 7.733519776268478, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19357 + }, + { + "epoch": 7.733919296843787, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3317, + "step": 19358 + }, + { + "epoch": 7.734318817419097, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19359 + }, + { + "epoch": 7.734718337994407, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19360 + }, + { + "epoch": 7.7351178585697165, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19361 + }, + { + "epoch": 7.735517379145026, + "grad_norm": 0.048828125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19362 + }, + { + "epoch": 7.735916899720335, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19363 + }, + { + "epoch": 7.736316420295645, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19364 + }, + { + "epoch": 7.736715940870955, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19365 + }, + { + "epoch": 7.7371154614462645, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3247, + "step": 19366 + }, + { + "epoch": 7.737514982021574, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19367 + }, + { + "epoch": 7.737914502596884, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19368 + }, + { + "epoch": 7.738314023172194, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19369 + }, + { + "epoch": 7.738713543747503, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19370 + }, + { + "epoch": 7.7391130643228125, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19371 + }, + { + "epoch": 7.739512584898122, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19372 + }, + { + "epoch": 7.739912105473432, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19373 + }, + { + "epoch": 7.740311626048742, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19374 + }, + { + "epoch": 7.740711146624051, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19375 + }, + { + "epoch": 7.741110667199361, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19376 + }, + { + "epoch": 7.74151018777467, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19377 + }, + { + "epoch": 7.74190970834998, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19378 + }, + { + "epoch": 7.74230922892529, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19379 + }, + { + "epoch": 7.742708749500599, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19380 + }, + { + "epoch": 7.743108270075909, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19381 + }, + { + "epoch": 7.743507790651218, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19382 + }, + { + "epoch": 7.743907311226528, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19383 + }, + { + "epoch": 7.744306831801838, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19384 + }, + { + "epoch": 7.744706352377148, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19385 + }, + { + "epoch": 7.745105872952457, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19386 + }, + { + "epoch": 7.7455053935277665, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19387 + }, + { + "epoch": 7.745904914103076, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19388 + }, + { + "epoch": 7.746304434678386, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3001, + "step": 19389 + }, + { + "epoch": 7.746703955253696, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3272, + "step": 19390 + }, + { + "epoch": 7.747103475829006, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19391 + }, + { + "epoch": 7.7475029964043145, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19392 + }, + { + "epoch": 7.747902516979624, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19393 + }, + { + "epoch": 7.748302037554934, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19394 + }, + { + "epoch": 7.748701558130244, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19395 + }, + { + "epoch": 7.749101078705554, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3302, + "step": 19396 + }, + { + "epoch": 7.7495005992808625, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19397 + }, + { + "epoch": 7.749900119856172, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19398 + }, + { + "epoch": 7.750299640431482, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19399 + }, + { + "epoch": 7.750699161006792, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3279, + "step": 19400 + }, + { + "epoch": 7.751098681582102, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 19401 + }, + { + "epoch": 7.7514982021574115, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 19402 + }, + { + "epoch": 7.75189772273272, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19403 + }, + { + "epoch": 7.75229724330803, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19404 + }, + { + "epoch": 7.75269676388334, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19405 + }, + { + "epoch": 7.75309628445865, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19406 + }, + { + "epoch": 7.7534958050339595, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19407 + }, + { + "epoch": 7.753895325609269, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3019, + "step": 19408 + }, + { + "epoch": 7.754294846184578, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19409 + }, + { + "epoch": 7.754694366759888, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19410 + }, + { + "epoch": 7.755093887335198, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19411 + }, + { + "epoch": 7.755493407910508, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.2998, + "step": 19412 + }, + { + "epoch": 7.755892928485817, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19413 + }, + { + "epoch": 7.756292449061126, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19414 + }, + { + "epoch": 7.756691969636436, + "grad_norm": 0.051513671875, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19415 + }, + { + "epoch": 7.757091490211746, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19416 + }, + { + "epoch": 7.757491010787056, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19417 + }, + { + "epoch": 7.757890531362365, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19418 + }, + { + "epoch": 7.758290051937674, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19419 + }, + { + "epoch": 7.758689572512984, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3256, + "step": 19420 + }, + { + "epoch": 7.759089093088294, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3267, + "step": 19421 + }, + { + "epoch": 7.759488613663604, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19422 + }, + { + "epoch": 7.759888134238913, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19423 + }, + { + "epoch": 7.760287654814223, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19424 + }, + { + "epoch": 7.760687175389533, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3296, + "step": 19425 + }, + { + "epoch": 7.761086695964842, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19426 + }, + { + "epoch": 7.761486216540152, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 19427 + }, + { + "epoch": 7.7618857371154615, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19428 + }, + { + "epoch": 7.762285257690771, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19429 + }, + { + "epoch": 7.762684778266081, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19430 + }, + { + "epoch": 7.76308429884139, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19431 + }, + { + "epoch": 7.7634838194167, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19432 + }, + { + "epoch": 7.7638833399920095, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19433 + }, + { + "epoch": 7.764282860567319, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19434 + }, + { + "epoch": 7.764682381142629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19435 + }, + { + "epoch": 7.765081901717938, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19436 + }, + { + "epoch": 7.765481422293248, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 19437 + }, + { + "epoch": 7.765880942868558, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19438 + }, + { + "epoch": 7.766280463443867, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19439 + }, + { + "epoch": 7.766679984019177, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19440 + }, + { + "epoch": 7.767079504594487, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19441 + }, + { + "epoch": 7.767479025169797, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19442 + }, + { + "epoch": 7.767878545745106, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19443 + }, + { + "epoch": 7.768278066320415, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19444 + }, + { + "epoch": 7.768677586895725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19445 + }, + { + "epoch": 7.769077107471035, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3298, + "step": 19446 + }, + { + "epoch": 7.769476628046345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19447 + }, + { + "epoch": 7.769876148621654, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19448 + }, + { + "epoch": 7.770275669196963, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19449 + }, + { + "epoch": 7.770675189772273, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3265, + "step": 19450 + }, + { + "epoch": 7.771074710347583, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3257, + "step": 19451 + }, + { + "epoch": 7.771474230922893, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19452 + }, + { + "epoch": 7.771873751498202, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19453 + }, + { + "epoch": 7.7722732720735115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3248, + "step": 19454 + }, + { + "epoch": 7.772672792648821, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19455 + }, + { + "epoch": 7.773072313224131, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19456 + }, + { + "epoch": 7.773471833799441, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19457 + }, + { + "epoch": 7.773871354374751, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3244, + "step": 19458 + }, + { + "epoch": 7.77427087495006, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19459 + }, + { + "epoch": 7.774670395525369, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19460 + }, + { + "epoch": 7.775069916100679, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19461 + }, + { + "epoch": 7.775469436675989, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19462 + }, + { + "epoch": 7.775868957251299, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19463 + }, + { + "epoch": 7.7762684778266085, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19464 + }, + { + "epoch": 7.776667998401917, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19465 + }, + { + "epoch": 7.777067518977227, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 19466 + }, + { + "epoch": 7.777467039552537, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19467 + }, + { + "epoch": 7.777866560127847, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19468 + }, + { + "epoch": 7.7782660807031565, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19469 + }, + { + "epoch": 7.778665601278465, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19470 + }, + { + "epoch": 7.779065121853775, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 19471 + }, + { + "epoch": 7.779464642429085, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.314, + "step": 19472 + }, + { + "epoch": 7.779864163004395, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19473 + }, + { + "epoch": 7.7802636835797045, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19474 + }, + { + "epoch": 7.780663204155014, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19475 + }, + { + "epoch": 7.781062724730323, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19476 + }, + { + "epoch": 7.781462245305633, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19477 + }, + { + "epoch": 7.781861765880943, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19478 + }, + { + "epoch": 7.782261286456253, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19479 + }, + { + "epoch": 7.782660807031562, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19480 + }, + { + "epoch": 7.783060327606872, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19481 + }, + { + "epoch": 7.783459848182181, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19482 + }, + { + "epoch": 7.783859368757491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19483 + }, + { + "epoch": 7.784258889332801, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19484 + }, + { + "epoch": 7.78465840990811, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19485 + }, + { + "epoch": 7.78505793048342, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19486 + }, + { + "epoch": 7.785457451058729, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19487 + }, + { + "epoch": 7.785856971634039, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3263, + "step": 19488 + }, + { + "epoch": 7.786256492209349, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19489 + }, + { + "epoch": 7.7866560127846585, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19490 + }, + { + "epoch": 7.787055533359968, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19491 + }, + { + "epoch": 7.787455053935278, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19492 + }, + { + "epoch": 7.787854574510587, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3325, + "step": 19493 + }, + { + "epoch": 7.788254095085897, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19494 + }, + { + "epoch": 7.7886536156612065, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19495 + }, + { + "epoch": 7.789053136236516, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19496 + }, + { + "epoch": 7.789452656811826, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19497 + }, + { + "epoch": 7.789852177387136, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19498 + }, + { + "epoch": 7.790251697962445, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3041, + "step": 19499 + }, + { + "epoch": 7.7906512185377546, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19500 + }, + { + "epoch": 7.791050739113064, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19501 + }, + { + "epoch": 7.791450259688374, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19502 + }, + { + "epoch": 7.791849780263684, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19503 + }, + { + "epoch": 7.792249300838993, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19504 + }, + { + "epoch": 7.792648821414303, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19505 + }, + { + "epoch": 7.793048341989612, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 19506 + }, + { + "epoch": 7.793447862564922, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19507 + }, + { + "epoch": 7.793847383140232, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.325, + "step": 19508 + }, + { + "epoch": 7.794246903715541, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19509 + }, + { + "epoch": 7.794646424290851, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19510 + }, + { + "epoch": 7.79504594486616, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19511 + }, + { + "epoch": 7.79544546544147, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19512 + }, + { + "epoch": 7.79584498601678, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19513 + }, + { + "epoch": 7.79624450659209, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19514 + }, + { + "epoch": 7.7966440271674, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19515 + }, + { + "epoch": 7.7970435477427085, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19516 + }, + { + "epoch": 7.797443068318018, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3238, + "step": 19517 + }, + { + "epoch": 7.797842588893328, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 19518 + }, + { + "epoch": 7.798242109468638, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19519 + }, + { + "epoch": 7.798641630043948, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19520 + }, + { + "epoch": 7.7990411506192565, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19521 + }, + { + "epoch": 7.799440671194566, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19522 + }, + { + "epoch": 7.799840191769876, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19523 + }, + { + "epoch": 7.800239712345186, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19524 + }, + { + "epoch": 7.800639232920496, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3225, + "step": 19525 + }, + { + "epoch": 7.801038753495805, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19526 + }, + { + "epoch": 7.801438274071114, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19527 + }, + { + "epoch": 7.801837794646424, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19528 + }, + { + "epoch": 7.802237315221734, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19529 + }, + { + "epoch": 7.802636835797044, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19530 + }, + { + "epoch": 7.8030363563723535, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19531 + }, + { + "epoch": 7.803435876947663, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19532 + }, + { + "epoch": 7.803835397522972, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19533 + }, + { + "epoch": 7.804234918098282, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19534 + }, + { + "epoch": 7.804634438673592, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19535 + }, + { + "epoch": 7.8050339592489015, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19536 + }, + { + "epoch": 7.805433479824211, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.312, + "step": 19537 + }, + { + "epoch": 7.80583300039952, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19538 + }, + { + "epoch": 7.80623252097483, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19539 + }, + { + "epoch": 7.80663204155014, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19540 + }, + { + "epoch": 7.80703156212545, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19541 + }, + { + "epoch": 7.807431082700759, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3043, + "step": 19542 + }, + { + "epoch": 7.807830603276068, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19543 + }, + { + "epoch": 7.808230123851378, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 19544 + }, + { + "epoch": 7.808629644426688, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19545 + }, + { + "epoch": 7.809029165001998, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19546 + }, + { + "epoch": 7.809428685577307, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19547 + }, + { + "epoch": 7.809828206152617, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.326, + "step": 19548 + }, + { + "epoch": 7.810227726727927, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19549 + }, + { + "epoch": 7.810627247303236, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3035, + "step": 19550 + }, + { + "epoch": 7.811026767878546, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19551 + }, + { + "epoch": 7.8114262884538554, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 19552 + }, + { + "epoch": 7.811825809029165, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 19553 + }, + { + "epoch": 7.812225329604475, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 19554 + }, + { + "epoch": 7.812624850179784, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3326, + "step": 19555 + }, + { + "epoch": 7.813024370755094, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3194, + "step": 19556 + }, + { + "epoch": 7.8134238913304035, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19557 + }, + { + "epoch": 7.813823411905713, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 19558 + }, + { + "epoch": 7.814222932481023, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19559 + }, + { + "epoch": 7.814622453056332, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19560 + }, + { + "epoch": 7.815021973631642, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19561 + }, + { + "epoch": 7.8154214942069515, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19562 + }, + { + "epoch": 7.815821014782261, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3083, + "step": 19563 + }, + { + "epoch": 7.816220535357571, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19564 + }, + { + "epoch": 7.816620055932881, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19565 + }, + { + "epoch": 7.81701957650819, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19566 + }, + { + "epoch": 7.8174190970835, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19567 + }, + { + "epoch": 7.817818617658809, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19568 + }, + { + "epoch": 7.818218138234119, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3171, + "step": 19569 + }, + { + "epoch": 7.818617658809429, + "grad_norm": 0.050537109375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19570 + }, + { + "epoch": 7.819017179384739, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19571 + }, + { + "epoch": 7.819416699960048, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19572 + }, + { + "epoch": 7.819816220535357, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19573 + }, + { + "epoch": 7.820215741110667, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3087, + "step": 19574 + }, + { + "epoch": 7.820615261685977, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19575 + }, + { + "epoch": 7.821014782261287, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19576 + }, + { + "epoch": 7.821414302836596, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19577 + }, + { + "epoch": 7.8218138234119055, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19578 + }, + { + "epoch": 7.822213343987215, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19579 + }, + { + "epoch": 7.822612864562525, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19580 + }, + { + "epoch": 7.823012385137835, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19581 + }, + { + "epoch": 7.823411905713145, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3287, + "step": 19582 + }, + { + "epoch": 7.8238114262884535, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19583 + }, + { + "epoch": 7.824210946863763, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3292, + "step": 19584 + }, + { + "epoch": 7.824610467439073, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19585 + }, + { + "epoch": 7.825009988014383, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19586 + }, + { + "epoch": 7.825409508589693, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3056, + "step": 19587 + }, + { + "epoch": 7.825809029165002, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3223, + "step": 19588 + }, + { + "epoch": 7.826208549740311, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19589 + }, + { + "epoch": 7.826608070315621, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19590 + }, + { + "epoch": 7.827007590890931, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19591 + }, + { + "epoch": 7.827407111466241, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19592 + }, + { + "epoch": 7.8278066320415505, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19593 + }, + { + "epoch": 7.828206152616859, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 19594 + }, + { + "epoch": 7.828605673192169, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3166, + "step": 19595 + }, + { + "epoch": 7.829005193767479, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19596 + }, + { + "epoch": 7.829404714342789, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19597 + }, + { + "epoch": 7.8298042349180985, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19598 + }, + { + "epoch": 7.830203755493407, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19599 + }, + { + "epoch": 7.830603276068717, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19600 + }, + { + "epoch": 7.831002796644027, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19601 + }, + { + "epoch": 7.831402317219337, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19602 + }, + { + "epoch": 7.831801837794647, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19603 + }, + { + "epoch": 7.832201358369956, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.2977, + "step": 19604 + }, + { + "epoch": 7.832600878945266, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19605 + }, + { + "epoch": 7.833000399520575, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19606 + }, + { + "epoch": 7.833399920095885, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19607 + }, + { + "epoch": 7.833799440671195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19608 + }, + { + "epoch": 7.834198961246504, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3271, + "step": 19609 + }, + { + "epoch": 7.834598481821814, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19610 + }, + { + "epoch": 7.834998002397123, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19611 + }, + { + "epoch": 7.835397522972433, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19612 + }, + { + "epoch": 7.835797043547743, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19613 + }, + { + "epoch": 7.836196564123052, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19614 + }, + { + "epoch": 7.836596084698362, + "grad_norm": 0.09912109375, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19615 + }, + { + "epoch": 7.836995605273671, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3088, + "step": 19616 + }, + { + "epoch": 7.837395125848981, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3243, + "step": 19617 + }, + { + "epoch": 7.837794646424291, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19618 + }, + { + "epoch": 7.8381941669996005, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3151, + "step": 19619 + }, + { + "epoch": 7.83859368757491, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.306, + "step": 19620 + }, + { + "epoch": 7.83899320815022, + "grad_norm": 0.0859375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19621 + }, + { + "epoch": 7.83939272872553, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19622 + }, + { + "epoch": 7.839792249300839, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19623 + }, + { + "epoch": 7.8401917698761485, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19624 + }, + { + "epoch": 7.840591290451458, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19625 + }, + { + "epoch": 7.840990811026768, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19626 + }, + { + "epoch": 7.841390331602078, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19627 + }, + { + "epoch": 7.841789852177387, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19628 + }, + { + "epoch": 7.842189372752697, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19629 + }, + { + "epoch": 7.842588893328006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3246, + "step": 19630 + }, + { + "epoch": 7.842988413903316, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19631 + }, + { + "epoch": 7.843387934478626, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 19632 + }, + { + "epoch": 7.843787455053935, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19633 + }, + { + "epoch": 7.844186975629245, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 19634 + }, + { + "epoch": 7.844586496204554, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19635 + }, + { + "epoch": 7.844986016779864, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19636 + }, + { + "epoch": 7.845385537355174, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3025, + "step": 19637 + }, + { + "epoch": 7.845785057930484, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3206, + "step": 19638 + }, + { + "epoch": 7.8461845785057935, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19639 + }, + { + "epoch": 7.846584099081102, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19640 + }, + { + "epoch": 7.846983619656412, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19641 + }, + { + "epoch": 7.847383140231722, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19642 + }, + { + "epoch": 7.847782660807032, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19643 + }, + { + "epoch": 7.848182181382342, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3189, + "step": 19644 + }, + { + "epoch": 7.8485817019576505, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19645 + }, + { + "epoch": 7.84898122253296, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3064, + "step": 19646 + }, + { + "epoch": 7.84938074310827, + "grad_norm": 0.052001953125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19647 + }, + { + "epoch": 7.84978026368358, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3192, + "step": 19648 + }, + { + "epoch": 7.85017978425889, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19649 + }, + { + "epoch": 7.8505793048341985, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3069, + "step": 19650 + }, + { + "epoch": 7.850978825409508, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19651 + }, + { + "epoch": 7.851378345984818, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19652 + }, + { + "epoch": 7.851777866560128, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19653 + }, + { + "epoch": 7.852177387135438, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 19654 + }, + { + "epoch": 7.8525769077107475, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19655 + }, + { + "epoch": 7.852976428286056, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3046, + "step": 19656 + }, + { + "epoch": 7.853375948861366, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19657 + }, + { + "epoch": 7.853775469436676, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19658 + }, + { + "epoch": 7.854174990011986, + "grad_norm": 0.083984375, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19659 + }, + { + "epoch": 7.8545745105872955, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19660 + }, + { + "epoch": 7.854974031162605, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19661 + }, + { + "epoch": 7.855373551737914, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.311, + "step": 19662 + }, + { + "epoch": 7.855773072313224, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19663 + }, + { + "epoch": 7.856172592888534, + "grad_norm": 0.09521484375, + "learning_rate": 0.0002, + "loss": 1.3237, + "step": 19664 + }, + { + "epoch": 7.8565721134638435, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19665 + }, + { + "epoch": 7.856971634039153, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19666 + }, + { + "epoch": 7.857371154614462, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 19667 + }, + { + "epoch": 7.857770675189772, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 19668 + }, + { + "epoch": 7.858170195765082, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19669 + }, + { + "epoch": 7.858569716340392, + "grad_norm": 0.06396484375, + "learning_rate": 0.0002, + "loss": 1.3033, + "step": 19670 + }, + { + "epoch": 7.858969236915701, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19671 + }, + { + "epoch": 7.859368757491011, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3178, + "step": 19672 + }, + { + "epoch": 7.85976827806632, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19673 + }, + { + "epoch": 7.86016779864163, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19674 + }, + { + "epoch": 7.86056731921694, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19675 + }, + { + "epoch": 7.860966839792249, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3045, + "step": 19676 + }, + { + "epoch": 7.861366360367559, + "grad_norm": 0.07568359375, + "learning_rate": 0.0002, + "loss": 1.3201, + "step": 19677 + }, + { + "epoch": 7.861765880942869, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3242, + "step": 19678 + }, + { + "epoch": 7.862165401518178, + "grad_norm": 0.1142578125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 19679 + }, + { + "epoch": 7.862564922093488, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19680 + }, + { + "epoch": 7.8629644426687975, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19681 + }, + { + "epoch": 7.863363963244107, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 19682 + }, + { + "epoch": 7.863763483819417, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19683 + }, + { + "epoch": 7.864163004394726, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19684 + }, + { + "epoch": 7.864562524970036, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19685 + }, + { + "epoch": 7.8649620455453455, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19686 + }, + { + "epoch": 7.865361566120655, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19687 + }, + { + "epoch": 7.865761086695965, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19688 + }, + { + "epoch": 7.866160607271274, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3106, + "step": 19689 + }, + { + "epoch": 7.866560127846584, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19690 + }, + { + "epoch": 7.8669596484218935, + "grad_norm": 0.0478515625, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19691 + }, + { + "epoch": 7.867359168997203, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3097, + "step": 19692 + }, + { + "epoch": 7.867758689572513, + "grad_norm": 0.076171875, + "learning_rate": 0.0002, + "loss": 1.3023, + "step": 19693 + }, + { + "epoch": 7.868158210147823, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3239, + "step": 19694 + }, + { + "epoch": 7.868557730723133, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19695 + }, + { + "epoch": 7.868957251298442, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19696 + }, + { + "epoch": 7.869356771873751, + "grad_norm": 0.1162109375, + "learning_rate": 0.0002, + "loss": 1.3276, + "step": 19697 + }, + { + "epoch": 7.869756292449061, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19698 + }, + { + "epoch": 7.870155813024371, + "grad_norm": 0.08984375, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19699 + }, + { + "epoch": 7.870555333599681, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19700 + }, + { + "epoch": 7.87095485417499, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19701 + }, + { + "epoch": 7.871354374750299, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19702 + }, + { + "epoch": 7.871753895325609, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19703 + }, + { + "epoch": 7.872153415900919, + "grad_norm": 0.130859375, + "learning_rate": 0.0002, + "loss": 1.3218, + "step": 19704 + }, + { + "epoch": 7.872552936476229, + "grad_norm": 0.10400390625, + "learning_rate": 0.0002, + "loss": 1.3229, + "step": 19705 + }, + { + "epoch": 7.872952457051538, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19706 + }, + { + "epoch": 7.8733519776268475, + "grad_norm": 0.11962890625, + "learning_rate": 0.0002, + "loss": 1.324, + "step": 19707 + }, + { + "epoch": 7.873751498202157, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19708 + }, + { + "epoch": 7.874151018777467, + "grad_norm": 0.10302734375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19709 + }, + { + "epoch": 7.874550539352777, + "grad_norm": 0.10205078125, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19710 + }, + { + "epoch": 7.874950059928087, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3122, + "step": 19711 + }, + { + "epoch": 7.875349580503396, + "grad_norm": 0.11572265625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19712 + }, + { + "epoch": 7.875749101078705, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3084, + "step": 19713 + }, + { + "epoch": 7.876148621654015, + "grad_norm": 0.09130859375, + "learning_rate": 0.0002, + "loss": 1.3062, + "step": 19714 + }, + { + "epoch": 7.876548142229325, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19715 + }, + { + "epoch": 7.876947662804635, + "grad_norm": 0.0546875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19716 + }, + { + "epoch": 7.877347183379944, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19717 + }, + { + "epoch": 7.877746703955253, + "grad_norm": 0.1015625, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19718 + }, + { + "epoch": 7.878146224530563, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19719 + }, + { + "epoch": 7.878545745105873, + "grad_norm": 0.1259765625, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 19720 + }, + { + "epoch": 7.878945265681183, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3137, + "step": 19721 + }, + { + "epoch": 7.8793447862564925, + "grad_norm": 0.11279296875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19722 + }, + { + "epoch": 7.879744306831801, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3277, + "step": 19723 + }, + { + "epoch": 7.880143827407111, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19724 + }, + { + "epoch": 7.880543347982421, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3245, + "step": 19725 + }, + { + "epoch": 7.880942868557731, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3093, + "step": 19726 + }, + { + "epoch": 7.8813423891330405, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19727 + }, + { + "epoch": 7.88174190970835, + "grad_norm": 0.0947265625, + "learning_rate": 0.0002, + "loss": 1.3114, + "step": 19728 + }, + { + "epoch": 7.88214143028366, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19729 + }, + { + "epoch": 7.882540950858969, + "grad_norm": 0.1044921875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19730 + }, + { + "epoch": 7.882940471434279, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19731 + }, + { + "epoch": 7.883339992009589, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19732 + }, + { + "epoch": 7.883739512584898, + "grad_norm": 0.0966796875, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19733 + }, + { + "epoch": 7.884139033160208, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3259, + "step": 19734 + }, + { + "epoch": 7.884538553735517, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19735 + }, + { + "epoch": 7.884938074310827, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19736 + }, + { + "epoch": 7.885337594886137, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19737 + }, + { + "epoch": 7.885737115461446, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.3076, + "step": 19738 + }, + { + "epoch": 7.886136636036756, + "grad_norm": 0.08642578125, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19739 + }, + { + "epoch": 7.886536156612065, + "grad_norm": 0.09765625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19740 + }, + { + "epoch": 7.886935677187375, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.3024, + "step": 19741 + }, + { + "epoch": 7.887335197762685, + "grad_norm": 0.0703125, + "learning_rate": 0.0002, + "loss": 1.3028, + "step": 19742 + }, + { + "epoch": 7.887734718337994, + "grad_norm": 0.08251953125, + "learning_rate": 0.0002, + "loss": 1.3205, + "step": 19743 + }, + { + "epoch": 7.888134238913304, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19744 + }, + { + "epoch": 7.888533759488614, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19745 + }, + { + "epoch": 7.888933280063923, + "grad_norm": 0.09326171875, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19746 + }, + { + "epoch": 7.889332800639233, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3136, + "step": 19747 + }, + { + "epoch": 7.8897323212145425, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19748 + }, + { + "epoch": 7.890131841789852, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19749 + }, + { + "epoch": 7.890531362365162, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19750 + }, + { + "epoch": 7.890930882940472, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19751 + }, + { + "epoch": 7.891330403515781, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19752 + }, + { + "epoch": 7.8917299240910905, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.321, + "step": 19753 + }, + { + "epoch": 7.8921294446664, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3251, + "step": 19754 + }, + { + "epoch": 7.89252896524171, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19755 + }, + { + "epoch": 7.89292848581702, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.316, + "step": 19756 + }, + { + "epoch": 7.893328006392329, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19757 + }, + { + "epoch": 7.893727526967639, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19758 + }, + { + "epoch": 7.894127047542948, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3065, + "step": 19759 + }, + { + "epoch": 7.894526568118258, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 19760 + }, + { + "epoch": 7.894926088693568, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19761 + }, + { + "epoch": 7.895325609268878, + "grad_norm": 0.07373046875, + "learning_rate": 0.0002, + "loss": 1.3228, + "step": 19762 + }, + { + "epoch": 7.895725129844187, + "grad_norm": 0.058349609375, + "learning_rate": 0.0002, + "loss": 1.3095, + "step": 19763 + }, + { + "epoch": 7.896124650419496, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3115, + "step": 19764 + }, + { + "epoch": 7.896524170994806, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19765 + }, + { + "epoch": 7.896923691570116, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19766 + }, + { + "epoch": 7.897323212145426, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3053, + "step": 19767 + }, + { + "epoch": 7.8977227327207356, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.307, + "step": 19768 + }, + { + "epoch": 7.8981222532960444, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19769 + }, + { + "epoch": 7.898521773871354, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19770 + }, + { + "epoch": 7.898921294446664, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19771 + }, + { + "epoch": 7.899320815021974, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 19772 + }, + { + "epoch": 7.899720335597284, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19773 + }, + { + "epoch": 7.9001198561725925, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19774 + }, + { + "epoch": 7.900519376747902, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.313, + "step": 19775 + }, + { + "epoch": 7.900918897323212, + "grad_norm": 0.0849609375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19776 + }, + { + "epoch": 7.901318417898522, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19777 + }, + { + "epoch": 7.901717938473832, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19778 + }, + { + "epoch": 7.9021174590491405, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19779 + }, + { + "epoch": 7.90251697962445, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19780 + }, + { + "epoch": 7.90291650019976, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19781 + }, + { + "epoch": 7.90331602077507, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19782 + }, + { + "epoch": 7.90371554135038, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3124, + "step": 19783 + }, + { + "epoch": 7.9041150619256895, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3144, + "step": 19784 + }, + { + "epoch": 7.904514582500999, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3297, + "step": 19785 + }, + { + "epoch": 7.904914103076308, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19786 + }, + { + "epoch": 7.905313623651618, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3258, + "step": 19787 + }, + { + "epoch": 7.905713144226928, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3126, + "step": 19788 + }, + { + "epoch": 7.9061126648022375, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19789 + }, + { + "epoch": 7.906512185377547, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3072, + "step": 19790 + }, + { + "epoch": 7.906911705952856, + "grad_norm": 0.06494140625, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19791 + }, + { + "epoch": 7.907311226528166, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19792 + }, + { + "epoch": 7.907710747103476, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.308, + "step": 19793 + }, + { + "epoch": 7.9081102676787856, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19794 + }, + { + "epoch": 7.908509788254095, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3073, + "step": 19795 + }, + { + "epoch": 7.908909308829404, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 19796 + }, + { + "epoch": 7.909308829404714, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 19797 + }, + { + "epoch": 7.909708349980024, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19798 + }, + { + "epoch": 7.910107870555334, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3148, + "step": 19799 + }, + { + "epoch": 7.910507391130643, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3059, + "step": 19800 + }, + { + "epoch": 7.910906911705953, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19801 + }, + { + "epoch": 7.911306432281263, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3133, + "step": 19802 + }, + { + "epoch": 7.911705952856572, + "grad_norm": 0.057373046875, + "learning_rate": 0.0002, + "loss": 1.3158, + "step": 19803 + }, + { + "epoch": 7.912105473431882, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 19804 + }, + { + "epoch": 7.912504994007191, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19805 + }, + { + "epoch": 7.912904514582501, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19806 + }, + { + "epoch": 7.913304035157811, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3111, + "step": 19807 + }, + { + "epoch": 7.91370355573312, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 19808 + }, + { + "epoch": 7.91410307630843, + "grad_norm": 0.07666015625, + "learning_rate": 0.0002, + "loss": 1.3094, + "step": 19809 + }, + { + "epoch": 7.9145025968837395, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3039, + "step": 19810 + }, + { + "epoch": 7.914902117459049, + "grad_norm": 0.07421875, + "learning_rate": 0.0002, + "loss": 1.3066, + "step": 19811 + }, + { + "epoch": 7.915301638034359, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19812 + }, + { + "epoch": 7.915701158609668, + "grad_norm": 0.091796875, + "learning_rate": 0.0002, + "loss": 1.3212, + "step": 19813 + }, + { + "epoch": 7.916100679184978, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.2976, + "step": 19814 + }, + { + "epoch": 7.9165001997602875, + "grad_norm": 0.056396484375, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19815 + }, + { + "epoch": 7.916899720335597, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3222, + "step": 19816 + }, + { + "epoch": 7.917299240910907, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19817 + }, + { + "epoch": 7.917698761486217, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.2984, + "step": 19818 + }, + { + "epoch": 7.918098282061527, + "grad_norm": 0.08203125, + "learning_rate": 0.0002, + "loss": 1.3185, + "step": 19819 + }, + { + "epoch": 7.918497802636836, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19820 + }, + { + "epoch": 7.918897323212145, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3173, + "step": 19821 + }, + { + "epoch": 7.919296843787455, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3219, + "step": 19822 + }, + { + "epoch": 7.919696364362765, + "grad_norm": 0.08056640625, + "learning_rate": 0.0002, + "loss": 1.315, + "step": 19823 + }, + { + "epoch": 7.920095884938075, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3155, + "step": 19824 + }, + { + "epoch": 7.920495405513384, + "grad_norm": 0.0791015625, + "learning_rate": 0.0002, + "loss": 1.3092, + "step": 19825 + }, + { + "epoch": 7.920894926088693, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3098, + "step": 19826 + }, + { + "epoch": 7.921294446664003, + "grad_norm": 0.0888671875, + "learning_rate": 0.0002, + "loss": 1.3127, + "step": 19827 + }, + { + "epoch": 7.921693967239313, + "grad_norm": 0.08447265625, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19828 + }, + { + "epoch": 7.922093487814623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3226, + "step": 19829 + }, + { + "epoch": 7.922493008389932, + "grad_norm": 0.09423828125, + "learning_rate": 0.0002, + "loss": 1.3183, + "step": 19830 + }, + { + "epoch": 7.922892528965241, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19831 + }, + { + "epoch": 7.923292049540551, + "grad_norm": 0.07958984375, + "learning_rate": 0.0002, + "loss": 1.3141, + "step": 19832 + }, + { + "epoch": 7.923691570115861, + "grad_norm": 0.06884765625, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19833 + }, + { + "epoch": 7.924091090691171, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19834 + }, + { + "epoch": 7.924490611266481, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19835 + }, + { + "epoch": 7.9248901318417895, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 19836 + }, + { + "epoch": 7.925289652417099, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19837 + }, + { + "epoch": 7.925689172992409, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3149, + "step": 19838 + }, + { + "epoch": 7.926088693567719, + "grad_norm": 0.1240234375, + "learning_rate": 0.0002, + "loss": 1.3052, + "step": 19839 + }, + { + "epoch": 7.926488214143029, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19840 + }, + { + "epoch": 7.926887734718338, + "grad_norm": 0.1298828125, + "learning_rate": 0.0002, + "loss": 1.3216, + "step": 19841 + }, + { + "epoch": 7.927287255293647, + "grad_norm": 0.06201171875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 19842 + }, + { + "epoch": 7.927686775868957, + "grad_norm": 0.0615234375, + "learning_rate": 0.0002, + "loss": 1.3217, + "step": 19843 + }, + { + "epoch": 7.928086296444267, + "grad_norm": 0.072265625, + "learning_rate": 0.0002, + "loss": 1.319, + "step": 19844 + }, + { + "epoch": 7.928485817019577, + "grad_norm": 0.08740234375, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19845 + }, + { + "epoch": 7.9288853375948865, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3048, + "step": 19846 + }, + { + "epoch": 7.929284858170195, + "grad_norm": 0.1220703125, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19847 + }, + { + "epoch": 7.929684378745505, + "grad_norm": 0.0751953125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19848 + }, + { + "epoch": 7.930083899320815, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.323, + "step": 19849 + }, + { + "epoch": 7.930483419896125, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3203, + "step": 19850 + }, + { + "epoch": 7.9308829404714345, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19851 + }, + { + "epoch": 7.931282461046744, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3075, + "step": 19852 + }, + { + "epoch": 7.931681981622053, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19853 + }, + { + "epoch": 7.932081502197363, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19854 + }, + { + "epoch": 7.932481022772673, + "grad_norm": 0.09033203125, + "learning_rate": 0.0002, + "loss": 1.2921, + "step": 19855 + }, + { + "epoch": 7.9328805433479825, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3071, + "step": 19856 + }, + { + "epoch": 7.933280063923292, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19857 + }, + { + "epoch": 7.933679584498602, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3168, + "step": 19858 + }, + { + "epoch": 7.934079105073911, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.3177, + "step": 19859 + }, + { + "epoch": 7.934478625649221, + "grad_norm": 0.08544921875, + "learning_rate": 0.0002, + "loss": 1.3163, + "step": 19860 + }, + { + "epoch": 7.934878146224531, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19861 + }, + { + "epoch": 7.93527766679984, + "grad_norm": 0.078125, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19862 + }, + { + "epoch": 7.93567718737515, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.3167, + "step": 19863 + }, + { + "epoch": 7.936076707950459, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19864 + }, + { + "epoch": 7.936476228525769, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19865 + }, + { + "epoch": 7.936875749101079, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19866 + }, + { + "epoch": 7.937275269676388, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19867 + }, + { + "epoch": 7.937674790251698, + "grad_norm": 0.0712890625, + "learning_rate": 0.0002, + "loss": 1.305, + "step": 19868 + }, + { + "epoch": 7.938074310827007, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19869 + }, + { + "epoch": 7.938473831402317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0002, + "loss": 1.3074, + "step": 19870 + }, + { + "epoch": 7.938873351977627, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.2995, + "step": 19871 + }, + { + "epoch": 7.9392728725529365, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3172, + "step": 19872 + }, + { + "epoch": 7.939672393128246, + "grad_norm": 0.09619140625, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19873 + }, + { + "epoch": 7.940071913703556, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3026, + "step": 19874 + }, + { + "epoch": 7.940471434278866, + "grad_norm": 0.080078125, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19875 + }, + { + "epoch": 7.940870954854175, + "grad_norm": 0.09814453125, + "learning_rate": 0.0002, + "loss": 1.3032, + "step": 19876 + }, + { + "epoch": 7.9412704754294845, + "grad_norm": 0.059326171875, + "learning_rate": 0.0002, + "loss": 1.3164, + "step": 19877 + }, + { + "epoch": 7.941669996004794, + "grad_norm": 0.09228515625, + "learning_rate": 0.0002, + "loss": 1.3128, + "step": 19878 + }, + { + "epoch": 7.942069516580104, + "grad_norm": 0.07275390625, + "learning_rate": 0.0002, + "loss": 1.3182, + "step": 19879 + }, + { + "epoch": 7.942469037155414, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19880 + }, + { + "epoch": 7.942868557730723, + "grad_norm": 0.10791015625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19881 + }, + { + "epoch": 7.9432680783060325, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.3, + "step": 19882 + }, + { + "epoch": 7.943667598881342, + "grad_norm": 0.0810546875, + "learning_rate": 0.0002, + "loss": 1.3145, + "step": 19883 + }, + { + "epoch": 7.944067119456652, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3107, + "step": 19884 + }, + { + "epoch": 7.944466640031962, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3139, + "step": 19885 + }, + { + "epoch": 7.944866160607271, + "grad_norm": 0.087890625, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19886 + }, + { + "epoch": 7.945265681182581, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19887 + }, + { + "epoch": 7.94566520175789, + "grad_norm": 0.07763671875, + "learning_rate": 0.0002, + "loss": 1.3132, + "step": 19888 + }, + { + "epoch": 7.9460647223332, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 19889 + }, + { + "epoch": 7.94646424290851, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.309, + "step": 19890 + }, + { + "epoch": 7.94686376348382, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3224, + "step": 19891 + }, + { + "epoch": 7.9472632840591295, + "grad_norm": 0.054443359375, + "learning_rate": 0.0002, + "loss": 1.31, + "step": 19892 + }, + { + "epoch": 7.947662804634438, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3184, + "step": 19893 + }, + { + "epoch": 7.948062325209748, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3096, + "step": 19894 + }, + { + "epoch": 7.948461845785058, + "grad_norm": 0.07080078125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19895 + }, + { + "epoch": 7.948861366360368, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3138, + "step": 19896 + }, + { + "epoch": 7.949260886935678, + "grad_norm": 0.062255859375, + "learning_rate": 0.0002, + "loss": 1.3044, + "step": 19897 + }, + { + "epoch": 7.9496604075109865, + "grad_norm": 0.0693359375, + "learning_rate": 0.0002, + "loss": 1.3234, + "step": 19898 + }, + { + "epoch": 7.950059928086296, + "grad_norm": 0.0771484375, + "learning_rate": 0.0002, + "loss": 1.3047, + "step": 19899 + }, + { + "epoch": 7.950459448661606, + "grad_norm": 0.10888671875, + "learning_rate": 0.0002, + "loss": 1.3179, + "step": 19900 + }, + { + "epoch": 7.950858969236916, + "grad_norm": 0.06005859375, + "learning_rate": 0.0002, + "loss": 1.3161, + "step": 19901 + }, + { + "epoch": 7.951258489812226, + "grad_norm": 0.08349609375, + "learning_rate": 0.0002, + "loss": 1.3063, + "step": 19902 + }, + { + "epoch": 7.9516580103875345, + "grad_norm": 0.0869140625, + "learning_rate": 0.0002, + "loss": 1.3195, + "step": 19903 + }, + { + "epoch": 7.952057530962844, + "grad_norm": 0.06787109375, + "learning_rate": 0.0002, + "loss": 1.3109, + "step": 19904 + }, + { + "epoch": 7.952457051538154, + "grad_norm": 0.07861328125, + "learning_rate": 0.0002, + "loss": 1.3295, + "step": 19905 + }, + { + "epoch": 7.952856572113464, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3197, + "step": 19906 + }, + { + "epoch": 7.953256092688774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0002, + "loss": 1.3131, + "step": 19907 + }, + { + "epoch": 7.953655613264083, + "grad_norm": 0.0830078125, + "learning_rate": 0.0002, + "loss": 1.3191, + "step": 19908 + }, + { + "epoch": 7.954055133839393, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3068, + "step": 19909 + }, + { + "epoch": 7.954454654414702, + "grad_norm": 0.06298828125, + "learning_rate": 0.0002, + "loss": 1.3288, + "step": 19910 + }, + { + "epoch": 7.954854174990012, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3082, + "step": 19911 + }, + { + "epoch": 7.955253695565322, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3081, + "step": 19912 + }, + { + "epoch": 7.9556532161406315, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3188, + "step": 19913 + }, + { + "epoch": 7.956052736715941, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19914 + }, + { + "epoch": 7.95645225729125, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19915 + }, + { + "epoch": 7.95685177786656, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3157, + "step": 19916 + }, + { + "epoch": 7.95725129844187, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3079, + "step": 19917 + }, + { + "epoch": 7.9576508190171795, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3078, + "step": 19918 + }, + { + "epoch": 7.958050339592489, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3233, + "step": 19919 + }, + { + "epoch": 7.958449860167798, + "grad_norm": 0.05029296875, + "learning_rate": 0.0002, + "loss": 1.3214, + "step": 19920 + }, + { + "epoch": 7.958849380743108, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3156, + "step": 19921 + }, + { + "epoch": 7.959248901318418, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19922 + }, + { + "epoch": 7.959648421893728, + "grad_norm": 0.05078125, + "learning_rate": 0.0002, + "loss": 1.3262, + "step": 19923 + }, + { + "epoch": 7.960047942469037, + "grad_norm": 0.051025390625, + "learning_rate": 0.0002, + "loss": 1.3007, + "step": 19924 + }, + { + "epoch": 7.960447463044347, + "grad_norm": 0.061767578125, + "learning_rate": 0.0002, + "loss": 1.3153, + "step": 19925 + }, + { + "epoch": 7.960846983619656, + "grad_norm": 0.0498046875, + "learning_rate": 0.0002, + "loss": 1.3176, + "step": 19926 + }, + { + "epoch": 7.961246504194966, + "grad_norm": 0.08154296875, + "learning_rate": 0.0002, + "loss": 1.3152, + "step": 19927 + }, + { + "epoch": 7.961646024770276, + "grad_norm": 0.052978515625, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19928 + }, + { + "epoch": 7.962045545345585, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3174, + "step": 19929 + }, + { + "epoch": 7.962445065920895, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3231, + "step": 19930 + }, + { + "epoch": 7.962844586496205, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19931 + }, + { + "epoch": 7.963244107071514, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3204, + "step": 19932 + }, + { + "epoch": 7.963643627646824, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19933 + }, + { + "epoch": 7.964043148222133, + "grad_norm": 0.060546875, + "learning_rate": 0.0002, + "loss": 1.3089, + "step": 19934 + }, + { + "epoch": 7.964442668797443, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3086, + "step": 19935 + }, + { + "epoch": 7.964842189372753, + "grad_norm": 0.054931640625, + "learning_rate": 0.0002, + "loss": 1.3198, + "step": 19936 + }, + { + "epoch": 7.965241709948062, + "grad_norm": 0.0537109375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 19937 + }, + { + "epoch": 7.965641230523372, + "grad_norm": 0.058837890625, + "learning_rate": 0.0002, + "loss": 1.3119, + "step": 19938 + }, + { + "epoch": 7.9660407510986815, + "grad_norm": 0.055908203125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19939 + }, + { + "epoch": 7.966440271673991, + "grad_norm": 0.06591796875, + "learning_rate": 0.0002, + "loss": 1.3091, + "step": 19940 + }, + { + "epoch": 7.966839792249301, + "grad_norm": 0.05517578125, + "learning_rate": 0.0002, + "loss": 1.3123, + "step": 19941 + }, + { + "epoch": 7.967239312824611, + "grad_norm": 0.053955078125, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 19942 + }, + { + "epoch": 7.96763883339992, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19943 + }, + { + "epoch": 7.9680383539752295, + "grad_norm": 0.09375, + "learning_rate": 0.0002, + "loss": 1.3253, + "step": 19944 + }, + { + "epoch": 7.968437874550539, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3162, + "step": 19945 + }, + { + "epoch": 7.968837395125849, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3154, + "step": 19946 + }, + { + "epoch": 7.969236915701159, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3118, + "step": 19947 + }, + { + "epoch": 7.969636436276469, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19948 + }, + { + "epoch": 7.970035956851778, + "grad_norm": 0.0517578125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19949 + }, + { + "epoch": 7.970435477427087, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3135, + "step": 19950 + }, + { + "epoch": 7.970834998002397, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3036, + "step": 19951 + }, + { + "epoch": 7.971234518577707, + "grad_norm": 0.060302734375, + "learning_rate": 0.0002, + "loss": 1.3102, + "step": 19952 + }, + { + "epoch": 7.971634039153017, + "grad_norm": 0.049560546875, + "learning_rate": 0.0002, + "loss": 1.3232, + "step": 19953 + }, + { + "epoch": 7.972033559728326, + "grad_norm": 0.059814453125, + "learning_rate": 0.0002, + "loss": 1.3108, + "step": 19954 + }, + { + "epoch": 7.972433080303635, + "grad_norm": 0.07177734375, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19955 + }, + { + "epoch": 7.972832600878945, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3196, + "step": 19956 + }, + { + "epoch": 7.973232121454255, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.3125, + "step": 19957 + }, + { + "epoch": 7.973631642029565, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3227, + "step": 19958 + }, + { + "epoch": 7.974031162604874, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.3241, + "step": 19959 + }, + { + "epoch": 7.974430683180183, + "grad_norm": 0.05615234375, + "learning_rate": 0.0002, + "loss": 1.32, + "step": 19960 + }, + { + "epoch": 7.974830203755493, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3181, + "step": 19961 + }, + { + "epoch": 7.975229724330803, + "grad_norm": 0.057861328125, + "learning_rate": 0.0002, + "loss": 1.33, + "step": 19962 + }, + { + "epoch": 7.975629244906113, + "grad_norm": 0.06103515625, + "learning_rate": 0.0002, + "loss": 1.3159, + "step": 19963 + }, + { + "epoch": 7.976028765481423, + "grad_norm": 0.053466796875, + "learning_rate": 0.0002, + "loss": 1.3146, + "step": 19964 + }, + { + "epoch": 7.976428286056732, + "grad_norm": 0.068359375, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 19965 + }, + { + "epoch": 7.976827806632041, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3169, + "step": 19966 + }, + { + "epoch": 7.977227327207351, + "grad_norm": 0.07470703125, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19967 + }, + { + "epoch": 7.977626847782661, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3117, + "step": 19968 + }, + { + "epoch": 7.978026368357971, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3085, + "step": 19969 + }, + { + "epoch": 7.97842588893328, + "grad_norm": 0.0576171875, + "learning_rate": 0.0002, + "loss": 1.3211, + "step": 19970 + }, + { + "epoch": 7.978825409508589, + "grad_norm": 0.064453125, + "learning_rate": 0.0002, + "loss": 1.318, + "step": 19971 + }, + { + "epoch": 7.979224930083899, + "grad_norm": 0.05712890625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19972 + }, + { + "epoch": 7.979624450659209, + "grad_norm": 0.056640625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19973 + }, + { + "epoch": 7.980023971234519, + "grad_norm": 0.049072265625, + "learning_rate": 0.0002, + "loss": 1.3186, + "step": 19974 + }, + { + "epoch": 7.9804234918098285, + "grad_norm": 0.06640625, + "learning_rate": 0.0002, + "loss": 1.3055, + "step": 19975 + }, + { + "epoch": 7.980823012385137, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3042, + "step": 19976 + }, + { + "epoch": 7.981222532960447, + "grad_norm": 0.0595703125, + "learning_rate": 0.0002, + "loss": 1.317, + "step": 19977 + }, + { + "epoch": 7.981622053535757, + "grad_norm": 0.0634765625, + "learning_rate": 0.0002, + "loss": 1.3116, + "step": 19978 + }, + { + "epoch": 7.982021574111067, + "grad_norm": 0.052490234375, + "learning_rate": 0.0002, + "loss": 1.3101, + "step": 19979 + }, + { + "epoch": 7.9824210946863765, + "grad_norm": 0.050048828125, + "learning_rate": 0.0002, + "loss": 1.3165, + "step": 19980 + }, + { + "epoch": 7.982820615261686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0002, + "loss": 1.3199, + "step": 19981 + }, + { + "epoch": 7.983220135836996, + "grad_norm": 0.05322265625, + "learning_rate": 0.0002, + "loss": 1.3112, + "step": 19982 + }, + { + "epoch": 7.983619656412305, + "grad_norm": 0.05859375, + "learning_rate": 0.0002, + "loss": 1.3175, + "step": 19983 + }, + { + "epoch": 7.984019176987615, + "grad_norm": 0.05908203125, + "learning_rate": 0.0002, + "loss": 1.3147, + "step": 19984 + }, + { + "epoch": 7.9844186975629245, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3143, + "step": 19985 + }, + { + "epoch": 7.984818218138234, + "grad_norm": 0.0556640625, + "learning_rate": 0.0002, + "loss": 1.3129, + "step": 19986 + }, + { + "epoch": 7.985217738713544, + "grad_norm": 0.0625, + "learning_rate": 0.0002, + "loss": 1.3236, + "step": 19987 + }, + { + "epoch": 7.985617259288853, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3142, + "step": 19988 + }, + { + "epoch": 7.986016779864163, + "grad_norm": 0.055419921875, + "learning_rate": 0.0002, + "loss": 1.3215, + "step": 19989 + }, + { + "epoch": 7.986416300439473, + "grad_norm": 0.05126953125, + "learning_rate": 0.0002, + "loss": 1.3193, + "step": 19990 + }, + { + "epoch": 7.986815821014782, + "grad_norm": 0.052734375, + "learning_rate": 0.0002, + "loss": 1.3202, + "step": 19991 + }, + { + "epoch": 7.987215341590092, + "grad_norm": 0.05419921875, + "learning_rate": 0.0002, + "loss": 1.3209, + "step": 19992 + }, + { + "epoch": 7.987614862165401, + "grad_norm": 0.0654296875, + "learning_rate": 0.0002, + "loss": 1.3103, + "step": 19993 + }, + { + "epoch": 7.988014382740711, + "grad_norm": 0.061279296875, + "learning_rate": 0.0002, + "loss": 1.3121, + "step": 19994 + }, + { + "epoch": 7.988413903316021, + "grad_norm": 0.05224609375, + "learning_rate": 0.0002, + "loss": 1.322, + "step": 19995 + }, + { + "epoch": 7.98881342389133, + "grad_norm": 0.0673828125, + "learning_rate": 0.0002, + "loss": 1.3134, + "step": 19996 + }, + { + "epoch": 7.98921294446664, + "grad_norm": 0.05810546875, + "learning_rate": 0.0002, + "loss": 1.3105, + "step": 19997 + }, + { + "epoch": 7.98961246504195, + "grad_norm": 0.06689453125, + "learning_rate": 0.0002, + "loss": 1.304, + "step": 19998 + }, + { + "epoch": 7.99001198561726, + "grad_norm": 0.04931640625, + "learning_rate": 0.0002, + "loss": 1.3275, + "step": 19999 + }, + { + "epoch": 7.990411506192569, + "grad_norm": 0.056884765625, + "learning_rate": 0.0002, + "loss": 1.3187, + "step": 20000 + } + ], + "logging_steps": 1.0, + "max_steps": 250300, + "num_input_tokens_seen": 0, + "num_train_epochs": 100, + "save_steps": 250, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 4.020101347939451e+18, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-20000/training_args.bin b/checkpoint-20000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2bc6e8c6a7b45b79e23a7e4b282dcb7f0e746bc4 --- /dev/null +++ b/checkpoint-20000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9307aa53030dab795f2b0f5373487f7b4270f37adff792b3db36f71bcb8dae52 +size 5816